ISBN 978-1-9433-6438-1 ABSTRACT With support from the United States Agency for International Development, and in partnership with the South Africa Department of Basic Education, MEASURE Evaluation is conducting an impact evaluation of the implementation of scripted lesson plans (SLP) and supporting activities developed to increase the rigor and uniformity of a life skills program for in-school youth. The evaluation aims to describe the effect of the SLP on herpes simplex virus 2 incidence or change in prevalence of pregnancy among a cohort of Grade-8 female learners as they progress to Grade 10. Secondarily, this evaluation aims to describe the program’s effects on knowledge, attitudes, school retention, self-reported risk behavior, HIV testing, and completed referrals for health services among the cohort of Grade-8 female learners and cross-sections of male learners in Grades 8, 9, and 10. The evaluation employed a two-arm, stratified, cluster-randomized design. Grade-8 females, Grade-8 males, and Grade-10 females (n=10,133) enrolled in 106 schools were surveyed. Baseline data were collected in August–October 2016. Sampling weights were applied to compute descriptive statistics and frequency distributions of all variables. Few differences were observed between the intervention and control groups. Most young people valued and appreciated what they were learning in their Life Orientation curriculum, but gaps emerged in their knowledge of HIV and AIDS and their self-efficacy for healthy behaviors. Of those surveyed, 68 percent to 90 percent had never had sex at the time of the survey. Among those who were sexually active, 26 percent to 46 percent did not use condoms consistently. 4 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 5 ACKNOWLEDGMENTS The evaluation team wishes to thank the following people and organizations for their support and involvement in the study. First, we greatly appreciate the effort of the learners in Grade 8 and Grade 10 who gave their time to be part of the baseline survey. Second, the school principals, Life Orientation educators, school administrators, school governing bodies, and the parents also provided us with support to collect these data. At each phase of the project, we had the support of the National Department of Basic Education, the United States Agency for International Development (USAID) in both South Africa and Washington, and the Education Development Center. Without their high-level engagement, this project would never have taken flight. Equally important was the engagement of the provincial and district Departments of Education in KwaZulu-Natal and Mpumalanga. The team would also like to thank the USAID-funded MEASURE Evaluation–Strategic Information for South Africa (MEval-SIFSA) project team, led by John Snow, Inc., which supported us throughout the process. Enterprises at the University of Pretoria—through a subcontract led by the local principal investigator, Professor Andy Beke—collected the data for this baseline report. We appreciate the tireless efforts of the entire University of Pretoria team, including the data-collection supervisors and interviewers. Cover photo: Senior secondary school learners in KwaZulu-Natal, a province in South Africa with the highest prevalence of HIV and AIDS. © 2000 Patrick Coleman, courtesy of Photoshare 6 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa CONTENTS ABSTRACT.................................................................................................................................................................. 3 ACKNOWLEDGMENTS ........................................................................................................................................ 5 FIGURES...................................................................................................................................................................... 8 TABLES ........................................................................................................................................................................ 9 ABBREVIATIONS................................................................................................................................................... 11 EXECUTIVE SUMMARY...................................................................................................................................... 12 Background ............................................................................................................................................................ 12 Methods.................................................................................................................................................................. 12 Findings .................................................................................................................................................................. 13 Conclusion.............................................................................................................................................................. 14 INTRODUCTION................................................................................................................................................... 15 Background ............................................................................................................................................................ 15 Aims of the Impact Evaluation........................................................................................................................... 17 METHODS ................................................................................................................................................................ 19 Evaluation Design................................................................................................................................................. 19 Target Population and Program Assignment ................................................................................................... 19 Statistical Power/Sample Size ............................................................................................................................. 20 Sampling ................................................................................................................................................................. 20 Inclusion and Exclusion Criteria ........................................................................................................................ 21 Selection Bias......................................................................................................................................................... 21 Survey Instrument................................................................................................................................................. 21 Field Implementation ........................................................................................................................................... 22 Pilot Study ......................................................................................................................................................... 22 Recruitment of Fieldworkers.......................................................................................................................... 22 Fieldworker Training ....................................................................................................................................... 22 Stakeholder Engagement and School Permission....................................................................................... 22 Parental Informed Consent............................................................................................................................ 23 Learner Informed Consent............................................................................................................................. 23 Questionnaire Completion and Dried Blood Spots Collection................................................................ 23 Ethics....................................................................................................................................................................... 24 Response Rates...................................................................................................................................................... 24 FINDINGS................................................................................................................................................................. 27 Demographics and School Attendance ............................................................................................................. 27 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 7 Demographics (Appendix Tables A1a–A1c)............................................................................................... 27 School Attendance and Performance (Appendix Tables A2a–A2c)........................................................ 28 Knowledge and Attitudes toward Learning about HIV and AIDS in Life Orientation............................ 29 Participation in and Perception of the Life Orientation Curriculum (Appendix Tables A3a–A3c)... 29 Sources of Information on and Knowledge of STIs, including HIV (Appendix Tables A4a-–A4c). 30 Knowledge of STIs, including HIV.............................................................................................................. 30 Sources of Information on HIV/AIDS and STIs...................................................................................... 32 Expectations about the Future, Gender Norms, and Self-Efficacy.............................................................. 34 Expectations about the Future....................................................................................................................... 34 Gender Norm Attitudes.................................................................................................................................. 35 Self-Efficacy (Appendix Tables A8a–A8c) .................................................................................................. 37 Sexual Behavior, Risk Perception, and HIV Testing and Counseling .......................................................... 38 Sexual Behavior ................................................................................................................................................ 38 Risk Perception................................................................................................................................................. 39 Risky Sexual Behaviors among Sexually Experienced Learners............................................................... 41 Pregnancy Experience ..................................................................................................................................... 42 HIV Testing and Counseling.......................................................................................................................... 43 DISCUSSION............................................................................................................................................................ 44 CONCLUSION......................................................................................................................................................... 45 REFERENCES.......................................................................................................................................................... 46 APPENDIX 1. BASELINE FINDINGS............................................................................................................. 48 APPENDIX 2. ADDITIONAL FINDINGS...................................................................................................... 93 APPENDIX 3. BASELINE QUESTIONNAIRE.............................................................................................. 99 APPENDIX 4. PROTOCOL................................................................................................................................116 APPENDIX 5. EVALUATION TEAM MEMBERS......................................................................................134 8 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa FIGURES Figure 1. Logic model of the DBE sexuality program......................................................................................... 16 Figure 2. Percentage of learners who reported losing one or both parents*.................................................... 28 Figure 3. Reported school attendance (percentage of learners who reported missing school)..................... 31 Figure 4. Sources of information about HIV and AIDS and other STIs among Grade-8 girls in MP and KZN (%) ............................................................................................................................................................. 33 Figure 5. Respondents learned most about sex, sexuality, and HIV at school (percentage of Grade-10 girls).............................................................................................................................................................................. 34 Figure 6. Percentage of Grade-8 female and male learners who “agreed” or “somewhat agreed” with gender attitudes related to sexual behaviors.......................................................................................................... 35 Figure 7. Self-efficacy around sexual behaviors among learners from KZN, by grade.................................. 37 Figure 8. Percentage of learners who report ever having had sexual intercourse, by grade, sex, and province ....................................................................................................................................................................................... 39 Figure 9. Levels of concern among learners about risk for HIV or AIDS during their lifetime, by grade, sex, and province...............................................................................................................................................................35 Figure 10. Levels of concern among learners about risk for HIV or AIDS during their lifetime, by grade, sex, and prior sexual experience..............................................................................................................................36 Figure 11. Percentage of sexually experienced female learners who had ever been pregnant ......................37 Figure 12. Percentage of learners who have ever been tested for HIV, by grade, sex, and province .......... 43 Figure A.1. Percentage of learners who went any of the past three days without food in their home (food￾insecure), by orphanhood status .............................................................................................................................93 Figure A.2. Percentage of learners who missed school, by grade and orphan status .....................................94 Figure A.3. Percentage of learners who had ever had sex, by grade and orphan status.................................94 Figure A.4. Percentage of learners who had ever had sex, by grade and by household food insecurity (i.e., went any days without food in the past 3 days vs. had food all those days)....................................................95 Figure A.5. Percentage of learners who had ever been pregnant (or gotten a partner pregnant) by grade and by household food insecurity for a) all learners; and b) learners who had ever had sex ........................95 Figure A.6. Percentage of learners who had ever been tested for HIV, by sexual experience and grade ...96 Figure A.7. Learners’ perceived chance (no chance, some chance, high chance) of acquiring HIV, by sexual experience and grade (%) .............................................................................................................................96 Figure A.8. Learners’ perceived chance (no chance, some chance, high chance) of getting HIV, by household food insecurity and grade (%)..............................................................................................................97 Figure A.9. Learners’ perceived chance (no chance, some chance, high chance) of getting pregnant or getting a partner pregnant while in high school, by sexual experience and grade (%)...................................97 Figure A.10. Learners’ perceived chance of getting pregnant or getting a partner pregnant while in high school, by household food insecurity and grade (%)...........................................................................................98 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 9 TABLES Table 1. Response rates, by province and study group........................................................................................ 19 Table 2. Response rates, by district and study group ........................................................................................... 26 Table 3. Sample sizes (unweighted and weighted)................................................................................................ 27 Table 4. Learners’ Life Orientation experience and perceived relevance, and parental attitudes toward and learners’ involvement in Life Orientation among Grade-8 boys from KZN and MP (%).................... 29 Table 5. Percentage of learners who correctly identified the following statements as true or false (correct response in parentheses)............................................................................................................................ 32 Table 6. Percentage of learners who “agree a lot” or “somewhat agree” with statements on gender roles and gender equity issues, by sex and grade groups..................................................................................... 29 Table 7. Risky sexual behaviors among learners who reported ever having sex, by age and grade.............. 39 Table Set A1. Demographics characteristics.......................................................................................................... 48 Table A1b. Demographic characteristics by province and intervention group, Grade-8 boys* ................... 49 Table A1c. Demographic characteristics by province and intervention group, Grade-10 girls* .................. 50 Table Set A2. School attendance, performance, and relationships.................................................................... 51 Table A2c. School attendance, performance, and relationships by province and intervention group, Grade-10 girls* ........................................................................................................................................................... 53 Table Set A3. Participation in and perceptions of Life Orientation Curriculum............................................. 54 Table A3c. Participation and perceptions of Life Orientation Curriculum, by province and intervention group, Grade-10 girls*¥............................................................................................................................................. 56 Table A4c. Knowledge of STIs/HIV, by province and intervention group, Grade-10 girls ........................ 61 Table A5c. Confidence and expectations about the future, by province and intervention group, Grade-10 girls............................................................................................................................................................. 65 Table Set A6. Gender norm attitudes..................................................................................................................... 66 Table A6c. Gender norm attitudes by province and intervention group, Grade-10 girls.............................. 72 Table Set A7. Risk perception.................................................................................................................................. 75 Table A7b. Risk perception by province and intervention group, Grade-8 boys............................................ 76 Table A7c. Risk perception, by province and intervention group, Grade-10 girls.......................................... 77 Table Set A8. Self-efficacy ........................................................................................................................................ 78 Table A8b. Self-efficacy, by province and intervention group, Grade-8 boys................................................. 79 Table A8c. Self-efficacy, by province and intervention group, Grade-10 girls................................................ 80 Table Set A9. Sexual behavior.................................................................................................................................. 81 Table A9b. Sexual behavior, by province and intervention group, Grade-8 boys........................................... 83 Table A9c. Sexual behavior, by province and intervention group, Grade-10 girls.......................................... 85 Table Set A10. HIV testing and counseling and other services.......................................................................... 87 10 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table A10b. HIV testing and counseling and other services, by province and intervention group, Grade-8 boys .............................................................................................................................................................. 89 Table A10c. HIV testing and counseling and other services, by province and intervention group, Grade-10 girls............................................................................................................................................................. 91 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 11 ABBREVIATIONS CSE Comprehensive Sexuality Education DBE Department of Basic Education DBS dried blood spot(s) DREAMS Determined, Resilient, Empowered, AIDS-free, Mentored, and Safe EDC Education Development Center HSV herpes simplex virus KZN KwaZulu-Natal LO Life Orientation MP Mpumalanga ODK Open Data Kit PEPFAR United States President’s Emergency Plan for AIDS Relief SACMEQ III Southern and Eastern Africa Consortium for Monitoring Educational Quality III SES socioeconomic status SLP scripted lesson plan STATSSA Statistics South Africa STI sexually transmitted infection USAID United States Agency for International Development 12 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa EXECUTIVE SUMMARY Background Gaining the knowledge and skills necessary to make healthy choices about sexual behavior as adolescent learners transition to young adulthood is key to controlling the potentially devastating effects of the HIV/AIDS epidemic. The South Africa Department of Basic Education (DBE) has attempted to attenuate the impact of HIV/AIDS on the nation through its Life Orientation (LO) curriculum, which has been in place since 2000. Previous studies found exposure to the curriculum was associated with better knowledge, attitudes, and some practices, but there was uneven implementation among educators and schools, and challenges with the curriculum (DBE, 2010). In 2015 DBE and the Education Development Center (EDC), with support from United States Agency for International Development (USAID) through the United States President’s Emergency Plan for AIDS Relief (PEPFAR), invested in developing and piloting scripted lesson plans (SLPs) and supporting activities to increase the rigor and uniformity of the life skills program. Before rolling out the program nationally, DBE and USAID tasked MEASURE Evaluation to conduct an impact evaluation of the activity in two South African provinces. The evaluation aims to describe the effect of the SLP on herpes simplex virus 2 (HSV-2) incidence or change in prevalence of pregnancy among a cohort of Grade-8 female learners in 2016 as they progress to Grade 10 in 2018. HIV incidence was initially considered but ultimately not selected as a primary outcome, because of ethical concerns. International guidelines recommend and South Africa policy requires that if HIV testing is done, then precounseling and postcounseling procedures must be implemented and test results must be given immediately or, if that is not possible, within five days of testing (Republic of South Africa Department of Health, 2015). Counseling and provision of results would potentially have a greater effect on learner behavior than the curriculum, so HSV-2 incidence was selected instead as a primary outcome. Because the incidence and prevalence of HSV-2 in young people are higher than for HIV, HSV-2 is often the biomarker of choice for these types of evaluation studies and permits a more accurate assessment of sexual behaviors than self-reported sexual behaviors. This evaluation has secondary outcomes. First, the evaluation will describe the effects of the program on knowledge, attitudes, school retention, and self-reported risk behavior, HIV counseling and testing, and completed referrals for health services among the cohort of Grade-8 female learners in 2016 as they progress to Grade 10 in 2018. Second, we will compare these secondary outcomes between Grade-10 female learners in 2016 with those of Grade-10 female learners in 2018. Third, using the cross-sectional sample of males in Grade 8 in 2016, Grade 9 in 2017, and Grade 10 in 2018, we will examine the secondary outcomes between learners in intervention and control schools. Secondary outcomes among the cross-sections of male learners are the same as those among the female learners. This report describes the methods of the impact evaluation of the South African school-based sexuality education program and documents findings from baseline data collection undertaken between August and October 2016. Methods The evaluation employed a two-arm, stratified, cluster-randomized design, where a secondary school is a cluster and a student learner is the unit of observation/analysis. Schools were sampled from five education districts in two provinces of South Africa—Bohlabela and Gert Sibande Districts, in Mpumalanga Province, and King Cetshwayo, Pinetown, and Umlazi Districts, in KwaZulu-Natal Province. In total, 58 intervention schools and 57 control schools (115 schools in total) were randomly Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 13 selected, and all Grade-8 females, Grade-8 males, and Grade-10 females enrolled in the selected schools in 2016 were eligible for interview upon receipt of their parent’s or guardian’s consent. Most learners were within the appropriate ages for grade (e.g., Grade 8, ages 13–15, and Grade 10, ages 15–17). Out of the 115 schools selected, 106 schools (55 intervention; 51 control) were included in the final sample, in which 3,606 Grade-8 females, 2,788 Grade-8 males, and 3,739 Grade-10 females were interviewed. The nine schools that were dropped either refused to participate (n=4), no longer had Grade 8-10 classes (n=2), or were not accessible because the schools were in a high-risk crime area (n=3). Baseline data were collected in the third term in the 2016 school year (from August–October). Learners were asked to take consent forms home for parents and guardians; assent was obtained from all learners along with parental/guardian consent. Learners who were 18 years or older provided consent to participate. Baseline data collection involved the collection of dried blood spots (DBS) by trained nurses; audio-assisted, self-administered survey data from Grade-8 female learners; and survey data (only) from Grade-8 male learners and Grade-10 female learners. At baseline, biomarker testing for HSV-2 was collected through DBS for the Grade-8 female learners. DBS from baseline are being stored until 2018 to permit incidence testing between baseline and 2018 for those female learners who test positive in 2018. At baseline, all participating Grade-8 female learners provided contact information to permit tracking at each follow-up round. Participants will be followed, even if they drop out of school. Institutional review boards at the University of North Carolina at Chapel Hill and University of Pretoria (in South Africa) provided ethical approval of all study procedures. Data were analyzed, using STATA statistical software (Stata Corp LP, College Station, Texas), by applying sampling weights to account for the sampling design and nonresponse. Descriptive statistics and frequency distribution were computed for all variables analyzed. Findings Within each province, there were few observed differences between the intervention and control group, which indicates that randomization within the evaluation study was successful. Therefore, an intention to treat analysis will be appropriate after end line data are collected. Baseline data demonstrate that just under half of the learners in the study sample were orphans: about one-third were single orphans and about one-tenth were double orphans. The mother was the primary caregiver for most participants, followed in frequency by a grandparent. Many of the young people in the sample valued and appreciated what they were learning in their LO classes. The baseline data indicate gaps in young people’s HIV and AIDS knowledge, experience, and self-efficacy that can be addressed through the LO program. Most of the young people surveyed had never had sex at the time of the survey. Among the Grade-8 boys, a quarter in KwaZulu-Natal (KZN) and one-fifth in Mpumalanga (MP) had ever had sex. One￾tenth of Grade-8 girls in each province and among Grade-10 girls one-quarter in KZN and one-third in MP had ever had sex. While three-quarters of sexually experienced boys reported using a condom the last time they had sex, and two-thirds to three-quarters reported using a condom every time they had sex, there remain gaps in condom use that place these young people at risk of acquiring HIV. Furthermore, by Grade 10, one-quarter of the female learners were sexually active, and about 59 percent of females who had ever had sex used a condom at last sex or consistently. 14 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Conclusion The EDC/DBE scripted LO lesson plans with supplemental activities program is seeking to address young people’s sexual behaviors as well as young people’s knowledge, self-efficacy, and awareness about HIV/AIDS in South Africa. This evaluation will illuminate gaps in the program and ways to strengthen it to meet the needs of all learners. The EDC team can use this baseline report to determine if they are currently meeting the needs of young people, and over time, the midline and end line data will make it possible to understand whether there are improvements in knowledge, attitudes, self-efficacy, behaviors and health status related to program exposure. This large-scale evaluation of the LO program, prior to national-level scale-up, is valuable because of the information it will provide about current efforts and potential for improvements. Evaluation results may also be used to inform the design and implementation of life skill curriculum and related HIV prevention activities for young people in other developing countries. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 15 INTRODUCTION Background Young people in South Africa, especially young women, are at high risk of acquiring HIV. In a 2012 South African national survey, HIV rates rose from 2.4 percent among girls under age 14, to 5.6 percent among adolescent girls ages 15–19. In 2013, 5.4 percent of girls 14–19 years old were pregnant (Stats SA, 2014). Among female learners in 14 high schools in KwaZulu-Natal Province, the prevalence of HIV was 6.4 percent; the prevalence of HSV-2 was 10.7 percent; and the prevalence of teenage pregnancy was 3.6 percent (Abdool Karim, et al., 2014). The authors of the KwaZulu-Natal study concluded, “The high prevalence of HIV, HSV-2, and pregnancy underscore the need for school-based sexual and reproductive health services.” South Africa’s DBE has created a national HIV and AIDS strategy with the goals of reducing the incidence of HIV, other sexually transmitted infections (STIs), and pregnancy among learners in Grades 7–9 in public schools in all nine provinces (DBE, 2010). The need to intervene with young women to reduce HIV incidence is underscored by the objectives of the new Determined, Resilient, Empowered, AIDS-free, Mentored, and Safe (DREAMS) initiative of the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR). Gaining the knowledge and skills necessary to make healthy choices about sexual behavior as learners transition to young adulthood is key to controlling the HIV/AIDS epidemic and attenuating the potentially devastating effects of the epidemic. Additionally, it is important to identify HIV-positive young people in schools (those who acquire the virus perinatally and behaviorally) and link these youths to HIV testing and counseling and to care, treatment, wellness services, and positive prevention. Through its life skills curriculum, in place since 2000, the DBE has attempted to address some of these issues. Previous studies found exposure to the curriculum to be associated with better knowledge, attitudes, and some practices, but there is uneven implementation among educators and schools and challenges with the curriculum (DBE, 2010). Because of these problems, DBE, with support from USAID, has invested in developing SLPs and supporting activities to increase the depth of information provided as well as the rigor of the life skills program—the LO curriculum—based on recommendations from the DBE report and the Southern and Eastern Africa Consortium for Monitoring Educational Quality III (SACMEQ III) project in South Africa. The main component of the new program is the sexuality and HIV education program for secondary school learners, which life skills educators deliver. These educators have been trained to use SLPs developed from an extensive review of existing life skills curricula and enhanced with interactive pedagogies. The SLPs aim to delay sexual debut, reduce unprotected sex, increase male and female condom use, reduce the number of sexual partners, and reduce violence and risk. Supportive activities are mobilization and engagement of parents, school management teams, and school governing bodies; focusing on sexual and reproductive health and gender-based violence; strengthening of referrals to health and social services; and linkages and increased access to youth-friendly sexual and reproductive health care services. The new SLPs are expected to be mandatory. A timetable for their implementation has been developed, and they have been assessed in order to institutionalize comprehensive sexuality education (CSE) in schools. To ensure implementation and the rollout of the new SLPs in select areas, USAID is providing technical assistance (through a five-year contract that was awarded to EDC and its partners the HIV & AIDS Research Division at the University of KwaZulu-Natal, the Society for Family Health, and Mott McDonald, in 2015) to support the DBE with in-service educator training activities in targeted provinces and districts to strengthen the DBE’s ability to implement and monitor the new sexuality and HIV- 16 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa prevention education program. EDC trained LO educators before the second quarter of the 2016 school year and subsequently began piloting the SLPs in intervention schools. In addition, the DBE is revising current policies to promote access to HIV-prevention services in schools, using biomedical, behavioral, social, and structural interventions to reduce transmission and vulnerability to HIV. DBE is also planning to roll out CSE. The first years of program implementation focus on Grades 7–9; development and implementation of the curriculum for other grades is in process. Figure 1. Logic model of the DBE sexuality program Most studies of similar programs have relied on self-reported risk behaviors, levels of HIV knowledge, and other behavioral data for assessing effectiveness. Of 83 studies of school-based sexuality programs, 65 percent found a significant impact on at least one sexual behavior but did not include biomarkers (Kirby, et al., 2006). A more recent review of adolescent pregnancy prevention programs in developing countries demonstrated that school-based interventions can be effective for increasing contraceptive use and reducing pregnancy among young people; however, the findings were mixed for some of the studies (McQueston, et al., 2013). Other strategies that were effective at reducing pregnancy risk and increasing contraceptive use were cash transfer programs, communication and health services/counseling, and peer education implemented in conjunction with school-based programming (McQueston, et al., 2013). Further, a randomized trial assessing the impact of a school-based program in 328 primary schools in Kenya found that the program had a significant effect on risky HIV-related behaviors and a 65 percent decrease in pregnancies with adult fathers (Duflo, et al., 2006; Dupas, 2006). In another large randomized trial in Tanzania, a teacher-led and peer-assisted sexual health education program with a community component was found to have had a significant impact on knowledge, attitudes, reported sexually transmitted infection symptoms, and sexual behavioral outcomes (Ross, et al., 2007). McCoy, et al., (2010) reviewed the results from studies of nine behavioral interventions conducted in Africa, India, Thailand, and Mexico and found limited evidence, overall, that behavioral interventions successfully change risk behavior and prevent HIV infection for women and girls in low- and middle-income countries. Locally, Jewkes, et al., (2008) evaluated the Stepping Stones program, an intensive 50-hour sexual health education activity aimed at changing gender norms and HIV risk behaviors, with a large (n = 2,776) randomized cluster trial of South African adolescents and young adults ages 15–26. The evaluation found that the program had prompted improvement in HIV testing rates; HIV-related communication; and HIV-related sexual risk behaviors, including condom use. The evaluation also found that, compared to men in the control group, fewer men who participated in the program reported perpetration of intimate partner violence at 24 months (6% vs. 10%, p = 0.054). There was also a 33 percent decrease in incidence of HSV-2 in the intervention group compared to the control group (Jewkes, et al., 2008). Because the Inputs •Resources •Schools •DBE curriculum with scripted lessons and supporting activities Program activities •Engage schools •Train teachers •Implement lessons •Implement supportive activities •Site visits to schools Outputs •Schools implementing the program •Teachers trained •Lessons administered •Site visits undertaken •Supportive activites delivered Outcomes •Improvement in: •Knowledge & attitudes •School retention •Risk behavior •HIV testing •Referrals for & uptake of health services Impact •Incidence reduction among females in: •HIV (not measured) •HSV-2, or •Pregnancy Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 17 incidence and prevalence of HSV-2 in young people are higher than for HIV, HSV-2 is often the biomarker of choice for these types of evaluation studies. Furthermore, including a biomarker for HSV-2 permits a better indication of sexual behaviors and is a marker of unprotected sex; self-reporting may miss this relevant information because young people may give socially desirable (and inaccurate) responses about sexual experience and protected sex. Before rolling out the new program nationwide, the USAID Mission in South Africa and the DBE needed to evaluate the program to see if it is having the expected impact. By contributing to the country’s evidence base of effective HIV programming for youth, findings from this study will help to ensure that young people in South Africa receive high-quality sexuality and HIV prevention education while in school. It will help to institutionalize prevention education in the South African school system, therefore helping to ensure that programs aimed at preparing the country’s youth to address sexual health and HIV challenges are sustainable. Aims of the Impact Evaluation The goal of the impact evaluation is to assess the impact of a school-based sexuality and HIV prevention education activity on learners over a period of two years. The evaluation aims to provide the South African DBE and PEPFAR South Africa with evidence of the effectiveness of the sexuality and HIV education program, by estimating the impact of the SLPs and supporting activities on the incidence of HSV-2 or prevalence of pregnancy among a cohort of girls in secondary school in two provinces of South Africa. The primary evaluation question is this: What is the effect of the SLPs and supporting activities on the incidence of HSV-2 or pregnancy, at the end of Grade 10, among a cohort of girls enrolled in Grade 8 at the beginning of 2016 in intervention schools, compared to a cohort of girls in Grade 8 at control schools that provide the current life skills program (i.e., the standard of practice)? The secondary evaluation questions are: • What is the effect of the SLPs and supporting activities on knowledge, attitudes, school retention, and self-reported risk behavior, HIV and other STI testing, and completed referrals for health services at the end of eighth, ninth, and tenth grades among a cohort of girls first interviewed in Grade 8 and among a cross section of boys interviewed in Grade 8, Grade 9, and Grade 10? • If there is a reduction in the primary outcome, does the intervention work by increasing school retention, or is the effect independent of school retention? • Do effects differ by sex or by rural versus urban schools? (Note: After baseline data collection, the realization is that there are not enough urban schools to answer this regional question.) The primary hypothesis is that the SLPs and supporting activities result in a reduction of the incidence of HSV-2 and/or pregnancy over a period of two years among a cohort of girls enrolled in Grade 8 at the beginning of 2016 compared to girls in schools with the current life skills program (without SLPs). The secondary hypothesis is that the SLPs and supporting activities increase school retention, knowledge, attitudes, self-reported risk and health behavior (i.e., condom use at last sex and delayed sexual debut), uptake of HIV and other STI testing and counseling, and completed referrals for reproductive health and family planning services among girls and boys in intervention schools in comparison with girls and boys 18 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa in schools with the current life skills program (without SLPs). Changes in these secondary outcome indicators, in turn, are expected to lead to reduced incidence of HSV-2, pregnancy, and HIV. The primary outcome measures for this impact evaluation are the incidence of HSV-2, measured through DBS or self-reported pregnancy. HIV incidence, measured through DBS, was initially considered but ultimately not selected as a primary outcome, because of ethical concerns. International guidelines suggest and South Africa policy requires that if HIV testing is done, then precounseling and postcounseling procedures must be implemented and test results must be given immediately or, if that is not possible, within five days of testing (Republic of South Africa Department of Health, 2015). Additionally, if DBS for HIV were collected, the institutional review boards at the University of North Carolina at Chapel Hill and the University of Pretoria would have required study staff to provide learners with their test results and counseling. Because counseling and provision of test results would potentially have a greater effect on learner behavior than the curriculum, HSV-2 incidence was selected instead as a primary outcome. Dried blood spots for HSV-2 testing were collected for Grade-8 female learners and will be stored for two years. DBS will be collected and analyzed in Grade 10; for any girl who tests positive for HSV-2, we will analyze her Grade-8 DBS to determine if she has an incident infection. Secondary outcome measures are school retention, knowledge, attitudes, self-reported risk behavior (i.e., condom use at last sex and delayed sexual debut), uptake of STI and HIV testing and counseling, and completed referrals for reproductive health and family planning services. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 19 METHODS Evaluation Design The evaluation employs a two-arm, stratified cluster-randomized trial, where a secondary school is a cluster and a learner is the unit of observation/analysis. A sample of schools was randomly selected and assigned to intervention and control arms for a longitudinal observation of Grade-8 female learners (data collected at baseline in year 2016, midline in 2017, and end line in 2018). In the intervention arm, educators were trained to provide sexuality and HIV-prevention education based on the new SLPs to learners enrolled in Grade 8 at the beginning of 2016, while in the control arm, schools follow the existing LO curriculum. The impact (at the end of 2018) of the new program will be measured by comparing the incidence of HSV-2, or pregnancy, among the cohort of Grade-8 female learners enrolled in the selected schools in 2016, in the intervention and control arms. The cohort of female learners in the intervention and control arms will take part in an annual panel survey from Grade 8 through Grade 10. Biomarkers for HSV-2 were collected from the cohort of girls in Grade 8 (2016) and will be collected again from the cohort of girls in 2018. All biomarker samples collected in 2018 (in Grade 10) will be tested for HSV-2. Biomarkers collected at Grade 8 are stored at -80 degrees Celsius and will only be tested, for baseline comparison, if the corresponding Grade-10 biomarker result is positive for HSV-2. This permits assessment of whether HSV-2 infection has occurred since the baseline. If approved by the IRB as part of end line data collection, biomarker testing of HIV may also be conducted at end line to estimate HIV prevalence (but not incidence). Pregnancy self-report was obtained during the baseline and will be collected at each survey time. Learners will be followed even if they drop out of school, to ensure a complete picture of intervention effects. Additionally, a cross-sectional sample of male learners in the same schools participated in the baseline survey in 2016 and will participate in the annual follow-up surveys in 2017 and 2018; no effort will be made to link the male learners over time. Knowledge, attitudes, and behaviors of male learners will be compared between intervention and control schools for each time period. Finally, a cross-sectional sample of Grade-10 girls in the same schools was also surveyed during baseline data collection in 2016 (when the cohort was in Grade 8). Grade-10 girls will be used as a comparison group for the Grade-10 longitudinal sample at end line, so that, in 2018, we can examine the knowledge, attitudes, and behaviors of Grade-10 girls exposed to the program and those not exposed to the program and compare their outcomes to those of the Grade-10 girls in 2016, prior to program launch with this grade. In addition, surveying the boys in the same grade each year as the target girls permits comparing differences between girls and boys as well as relative changes in each group with and without the program. The evaluation includes a qualitative component at midline. The qualitative study will examine the perspectives of learners, parents, educators, and other stakeholders regarding the sexuality portion of the LO curriculum and supportive activities and how accepting and comfortable educators are with the sexuality education lessons of the LO curriculum. The qualitative study will also examine contextual factors that may influence whether changes in primary and secondary outcomes are observed. Results from the qualitative study are intended to be used by the stakeholders to refine program implementation before the completion of the impact evaluation. Target Population and Program Assignment The coverage of the program for the evaluation included five education districts in two provinces with a high prevalence of STIs and pregnancy, as identified by the USAID mission in South Africa (USAID/SA) and the DBE. The districts are Bohlabela and Gert Sibande, in Mpumalanga Province, and 20 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa King Cetshwayo, Pinetown, and Umlazi, in KwaZulu-Natal Province. Target schools are those in the three lowest socioeconomic-status (SES) quintiles and those that include learners in Grades 8–10. A random sample of target schools was selected and randomly assigned either to the intervention or control arm. The population covered is a cohort of female learners enrolled in Grade 8 at the beginning of 2016 and a cross-section of female learners in Grade 10 and male learners in Grade 8 at baseline. Statistical Power/Sample Size The goal of the sample size calculation was to power the statistical analysis on the primary outcome, i.e., the incidence of HSV-2 or pregnancy among a cohort of Grade-8 female learners. The sampling plan was designed to recruit 2,500 female learners in Grade 8 in each of the two arms (5,000 female learners in total) from 115 schools. Sample size calculations were based on assumptions and specifications of sampling parameters. First, we specified the minimum detectable change in the primary outcomes based on the assumed incidence rate of HSV-2 or pregnancy of 0.04 in the intervention arm versus 0.08 in the control arm over two school years at the significance level (α) of 0.05 (two-sided). Next, we adjusted the sample size for the following: (1) design effect, to account for elevated standard errors in a cluster sample design; (2) baseline prevalence of the primary outcome, to account for loss of units available to estimate the incidence rate1 ; and (3) nonresponse of schools and female learners. The design effect from clustering is approximated as 1+ICC×(𝑀−1), where ICC is intra-cluster correlation and 𝑀 is the average cluster size (Kish, 1965).2 We assumed an ICC of 0.03 and an average of 50 female learners per school, implying a design effect of 2.47. Next, we assumed the prevalence at 1 percent of HSV-2 and 0 percent of pregnancy at baseline. Finally, we accounted for potential nonresponse in schools (15 out of 115 schools) and assumed a response rate of 70 percent for Grade-8 female learners at baseline. With a total sample size in both arms together of 3,500 female learners successfully interviewed at baseline (or 5,000 female learners recruited for interviews with an assumed response rate of 70%), we estimated a statistical power (1-β) of 88 percent. Sampling The study applies a stratified cluster sampling approach to obtain a random sample of learners in the intervention and control arms, respectively. The sampling frame was constructed from a list of schools provided by the DBE, which had information on the schools’ locations and measures of the SES of the catchment learners. Eligible schools were those in the three lowest SES quintiles and those that include Grades 8–10 learners. Most learners in Grade 8 will be ages 13–15; most in Grade 10 will be ages 15–17. The frame was then stratified by the five education districts in two provinces: Bohlabela and Gert Sibande Districts, in Mpumalanga Province, and King Cetshwayo, Pinetown, and Umlazi Districts, in KwaZulu-Natal Province. A stratified random sample of schools was selected whereby the number of selected schools within each district was proportional to the number of eligible schools in the district within each province to reflect the composition of the target population in each province. Then, within each district, the selected schools were randomly assigned either to the intervention or control arm. In total, 115 schools were randomly selected and assigned to 58 intervention schools (23 in MP and 35 in KZN) and 57 control schools (22 in MP and 35 in KZN) as presented in Table 1. All Grade 8 female, Grade 8 male, and Grade 10 female learners in January 2016 in the selected schools were invited to participate in the baseline study upon receipt of consent of a parent or guardian. 1 For the estimation of the incidence rate of HSV-2, we need to exclude from data analysis those who are HSV-2 positive at baseline. The HSV-2 status at baseline will be assessed for those whose biomarker testing results are positive at end line. Those who contracted HSV-2 prior to study enrollment will be excluded from data analysis at end line. 2 Kish, L. (1965). Survey sampling. New York: John Wiley & Sons. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 21 Inclusion and Exclusion Criteria Schools were eligible for the study if they met all of the following criteria: • Located in PEPFAR priority provinces: Bohlabela and Gert Sibande Districts in Mpumalanga Province, and King Cetshwayo, Pinetown and Umlazi Districts in KwaZulu-Natal Province, • Secondary schools with at least Grades 8–10 present • Schools in the Quintiles 1–3 range Individual learners were eligible for the study if they met the following criteria: • Female learner enrolled in Grade 8 in a selected school at time of baseline; followed longitudinally those who enrolled in the study • Male learner enrolled in Grade 8 in a selected school at time of baseline • Female learner enrolled in Grade 10 in a selected school at the time of baseline • Received parental consent to participate Selection Bias We randomized schools to minimize selection bias of schools and learners between the intervention and control arms. However, selection bias may still be a concern if unobserved characteristics of learners in the intervention arm systematically differ from unobserved characteristics of learners in the control arm, and these differences are related to the study outcome(s). First, some learners will not be able to participate in some or all surveys throughout the study. Learners who did not have a parental/guardian consent were not interviewed for the baseline study. Second, there may be attrition of learners when they transfer or drop out of schools. We will monitor learners’ transfer or dropout and their exposure to the LO program throughout the study. We will minimize attrition by retrieving contact information from Grade-8 female learners and their schools, and following up with those at home who have left school at midline and end line. We will also apply an intention-to-treat analysis and analyze data from learners based on their initial assignment of the LO program. Using data from the program monitoring system, we will classify schools based on their level of program exposure. This information will be used to include additional variables in the analysis to detect the minimum level of exposure needed to observe changes. Survey Instrument The survey instrument was developed in consultation with stakeholders from the DBE, EDC and the consortium partners, and USAID. It covers the following topics that measure questions for the evaluation: • Demographics and household composition • Connectivity to caregivers • School attendance and performance • Aspirations and expectations about the future • Risk perception • Sexual behavior • Participation in and perceptions of LO curriculum 22 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa The surveys were translated to local languages by professional translators. The languages available for this survey were English, Sepedi, SiSwati, xiTsonga, and IsiZulu. The surveys were uploaded onto tablets using Open Data Kit (ODK) and each question could be read or listened to. Field Implementation Pilot Study A pilot study was conducted in three schools in MP (Witbank area) that were not included in the sample. Learners in Grades 8 and 10, representing all language groups, participated in the pilot. The purpose of the pilot study was to test the data-collection procedure and pretest the questionnaire. The study procedure and questionnaire were revised on the basis of feedback from participating learners and educators. Recruitment of Fieldworkers Field teams consisting of a provincial coordinator, team leads, data collectors, and nurses were recruited to conduct data collection in each of the two provinces. Most of the data collectors were recruited from the Statistics South Africa (STATSSA) fieldworkers’ database—namely those with previous fieldwork experience for large-scale evaluations and proficiency in select local languages. The field team was required to approach the selected schools, gain permission from the principals and/or school governing bodies, follow consent and assent procedures, guide learners through self-administered surveys (on tablets), collect blood samples (conducted by nurses), and collect contact information for follow-up with Grade-8 female learners for the subsequent data-collection points. Fieldworker Training All recruited fieldworkers participated in a three-day training within their province. Training included research ethics, study-specific procedures for data collection and use of data collection tools, gaining permission from sampled schools, obtaining parental consent and learner assent, managing classroom dynamics, and developing approaches for collecting blood samples (for the nurses). The training emphasized the primary ethical standards: informed consent, anonymity, confidentiality and privacy, and doing no harm. All fieldworkers signed statements agreeing to protect the security and confidentiality of identifiable information. In addition to training, a comprehensive standard of operation manual was developed to provide guidance on pre-visit preparations, the use of all data-collection tools (including the tablet-based self-administered survey), the process for nurses to administer a health screening tool and collect DBS from Grade-8 female learners, the flow of onsite procedures, and quality assurance measures to ensure accurate data collection. Stakeholder Engagement and School Permission A series of stakeholder engagement meetings were conducted before data collection at all levels: the national, provincial, district, school, and community levels. The purpose of stakeholder engagement meetings was to introduce and discuss the imminent survey and process requirements for their support. Letters of support and endorsement were obtained from the DBE and each Provincial Department of Education. Once letters of endorsement were obtained, meetings were held with the principals of the selected schools at the beginning of the data collection period. The purpose of the meetings was to provide the principals and members of school staff with information about the study and what would be required of Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 23 the school, school staff, and learners in the selected grades. Permission to conduct the study was gained from the principals. Parental Informed Consent At the school level, the field teams, with the support of school staff, informed learners in the participating classes in each school about the study and what would be required of them. It was explained that parents or guardians would first need to agree for learners to participate, and learners were given a study information sheet to take home to their parents or guardians. The sheet contained a reply space where parents or guardians were asked to indicate if they consented or not for their child or ward to participate in the study. Learners were asked to return the signed forms to the school within three days. The information sheet also contained contact details for the principal investigator and the ethics committee, so they could be contacted by any parents/guardians who had questions regarding the research. Learner Informed Consent All learners whose parents consented to study participation were eligible to participate in the survey (see Table 1 for study response rates). On the day of data collection, the field team summarized the information sheet with the learners and answered all their questions. Potential study participants were clearly told that they were free at all times not to participate in the research and that refusal to participate would in no way affect their academic enrollment or performance at their school. Participants were also encouraged to share with researchers any concern they might have regarding their research participation. Learners who wanted to participate were asked to complete the assent form attached to the information sheet. All retuned parental consent and learner assent forms were collected and stored in accordance with ethics requirements. Questionnaire Completion and Dried Blood Spots Collection Survey completion and DBS collection, in most cases, occurred within three days after parental informed consent. On the day of data collection, field team members provided instructions to learners before they began completing the ODK survey and DBS collection (for Grade-8 girls). The instructions emphasized the following: • To learners, that participation was voluntary, and learners were allowed to withdraw from the study at any time • To learners, to seek assistance from field team members and not from another learner, should they require it • To learners, the need to abstain from discussing the contents of the questionnaire with other learners • To the nurses, the need to provide simple explanations to Grade-8 girls of the testing procedures for data collection The learners completed ODK audio-assisted, self-administered survey interviews. Biomarker data collection—DBS for HSV-2 testing were collected from Grade 8 female learners at baseline and stored for future testing at end line. Biomarkers will be collected again at end line and tested for HSV-2; biomarkers from baseline are currently being stored at -80 degrees Celsius, and the corresponding baseline samples will be tested only for comparison with those samples that are HSV-2 positive at end line. The later testing of the baseline biomarkers will permit a determination of whether an individual with HSV-2 at end line had an incident infection between baseline and end line or if she was already HSV-2 positive at baseline. All positive biomarker tests will be subject to confirmatory tests to 24 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa minimize measurement error. At the end of the study in 2018, when the cohort is in Grade 10, learners will receive the results of the HSV-2 testing and will receive a referral for HIV testing. The trained nurses also provided a health screening for the cohort of Grade-8 female learners. The health screening involved going through a standardized checklist of the algorithm for syndromic diagnosis of STI. Any learner who presented a health issue, including symptoms of HSV-2, was referred to the local public health clinic for further management. No information from the health screen was stored as part of this study. Anonymous data-collection strategies were followed to maintain a high level of confidentiality. No questionnaires or specimen-collection forms contained names or other explicit identifiers. Learners were not required to enter their names in the tablet-based survey that uses the ODK mobile data collection application. Instead, all participating learners were assigned a unique identifying number for follow-up. Corresponding names and codes are kept separately from any data, are password-protected, and are accessible only by authorized members of the evaluation team. Furthermore, all Grade-8 girls received a thank you letter in a sealed envelope after the biomarker testing. Those who were referred had an additional piece of paper in the sealed envelope; this was done to protect the privacy of those who needed referral. Ethics This study adheres to the three Belmont principles of ethics that guide researchers in conducting safe research: respect for persons, beneficence, and justice. Ethical clearance was obtained from the University of Pretoria Faculty of Health Sciences Research Ethics Committee (Ref No 153/2016) and the University of North Carolina at Chapel Hill Institutional Review Board. Response Rates Tables 1–2 present the results of the interviews with schools and learners, by province and district for the intervention and control arms. A total of 115 schools were selected, of which 106 (92.2%) were successfully interviewed: 55 intervention and 51 control. There were two schools that were determined to be ineligible after selection, because they no longer included Grades 8–10. The principal reasons for nonresponse (n=7) among the remaining eligible schools were refusal (n=4) and inaccessibility owing to security concerns (n=3). In the interviewed schools, 3,606 female learners in Grade 8 (74.0% response rate), 2,788 male learners in Grade 8 (54.0% response rate), and 3,739 female learners in Grade 10 (60.4% response rate) were successfully interviewed; the final total sample size of learners was 10,133. The response rates are calculated based on the number of eligible learners in a class/grade provided by the school or estimated for the small number of schools where this information was not available. The following were the principle reasons given for nonresponse among eligible individual learners: lack of parental consent, learner’s refusal, and learner’s absence. 3 The number (weighted and unweighted) of interviewed learners by district, study arm, grade, and sex is shown in Table 3. The average response rates were applied to three schools missing the number of school enrollees. Sampling weights were obtained by accounting for the sampling design, and response rates were calculated by combination of districts, intervention arms, sex, and grades. 3 From the data available, it is not possible to distinguish refusals from those were absent or did not have parental consent. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 25 Table 1. Response rates, by province and study group KwaZulu-Natal Mpumalanga Intervention Control Intervention Control Schools sampled 35 35 23 22 Schools eligible 35 33 23 22 Schools interviewed 32 30 23 21 School response rate (%)1 91.4 90.9 100.0 95.5 Grade-8 girls interviewed 1064 924 833 785 Grade-8 girls’ response rate (%)2 67.03 75.54 75.9 77.6 Grade-8 boys interviewed 736 699 654 699 Grade-8 boys’ response rate (%)2 43.23 48.34 58.3 66.1 Grade-10 girls interviewed 1033 975 886 845 Grade-10 girls’ response rate (%)2 62.03 62.84 57.2 59.1 1 Defined as the number of interviewed schools divided by the number of eligible schools 2 Defined as the number of respondents divided by the number of enrollees 3 The number of enrollees was missing for two schools. The response rate is calculated excluding the two schools. 4 The number of enrollees was missing for one school. The response rate is calculated excluding the school. 26 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table 2. Response rates, by district and study group* KwaZulu-Natal Mpumalanga Pinetown Umlazi King Cetshwayo Bohlabela Gert Sibande I* C* I C I C I C I C School response rate (%)1 100.0 85.7 80.0 100.0 90.9 90.5 100.0 100.0 100.0 85.7 Grade-8 girls’ response rate (%)2 60.8 76.2 44.23 81.64 77.7 73.0 86.1 83.6 68.8 63.3 Grade-8 boys’ response rate (%)2 43.6 38.9 26.63 59.84 46.1 49.7 78.5 72.9 43.8 51.1 Grade-10 girls’ response rate (%)2 57.6 59.1 34.63 82.04 70.8 59.6 78.5 73.8 39.2 37.5 *I = intervention group; C = control group 1 Defined as the number of interviewed schools divided by the number of eligible schools 2 Defined as the number of respondents divided by the number of enrollees 3 The number of enrollees was missing for two schools. The response rate is calculated excluding the two schools. 4 The number of enrollees was missing for one school. The response rate is calculated excluding the school. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 27 Table 3. Sample sizes (unweighted and weighted) KwaZulu Natal Mpumalanga Intervention Control Total Intervention Control Total Grade-8 girls Unweighted Weighted 1064 1398.0 924 983.0 1988 2381.0 833 351.9 785 340.7 1618 692.6 Grade-8 boys Unweighted Weighted 736 1502.0 699 1148.0 1435 2650.0 654 359.6 699 356.6 1353 716.2 Grade-10 girls Unweighted Weighted 1033 1463.0 975 1244.0 2008 2707.0 886 496.9 845 489.3 1731 986.3 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 27 FINDINGS This section presents the baseline results from the survey of the study participants. Frequencies were computed for each variable by province and intervention and control groups. Separate tables are shown by grade and sex (i.e., for each table, we have a comparable table for Grade-8 girls, Grade-8 boys, and Grade￾10 girls.) The full set of tables can be found in the appendix. Here, we present a sample of the characteristics, behaviors, and experiences of the in-school learners. Though not presented here, the tables in the appendix demonstrate that randomization of the schools was successful and generally, across all measures including knowledge, self-efficacy, sexual behaviors, and HIV testing, the intervention and control learners (by grade, sex, and province) are comparable. Demographics and School Attendance Demographics (Appendix Tables A1a–A1c) The mean age for each group of respondents was 14.3 years for Grade-8 boys, 13.6 years for Grade-8 girls, and 16.0 for Grade-10 girls. The age range for all three groups was 11–25 years; however, most learners were within the appropriate ages for grade (e.g., Grade 8, ages 13–15, and Grade 10, ages 15–17). Mean ages were the same in both provinces. Grade-8 boys were slightly older than Grade-8 girls by approximately eight months. In both provinces, “Mother” was reported as the primary caregiver by more than two-thirds of Grade-8 boys and approximately three-quarters of Grade-8 girls and Grade-10 girls. The second most frequently reported primary caregiver was “Grandparent,” by more than 10 percent of Grade-8 boys, Grade-8 girls, and Grade-10 girls, in both provinces. A greater percentage of Grade-8 boys reported their father as the primary caregiver (about 12.5%) than did Grade-8 girls (5%–6%) and Grade-10 girls (6%–7%). Children were defined as orphans if they reported a parent was dead or did not know if a parent was alive. It is noteworthy that just under half of all learners were orphans: approximately one-third of all learners who responded had lost one biological parent and more than one-tenth had lost both parents (see Figure 2). There were higher percentages of learners reporting the loss of one or both parents in KZN than in MP. For example, among Grade-8 girls, 31.2 percent in KZN and 28.8 percent in MP had lost one parent; 12.9 percent in KZN and 9.7 percent in MP had lost both parents. Among Grade-10 girls, 35.4 percent in KZN and 29.8 percent in MP had lost one parent; 14.7 percent in KZN and 10.7 percent in MP had lost both parents. A similar pattern was observed among Grade-8 boys. 28 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Figure 2. Percentage of learners who reported losing one or both parents* *Reported parent was dead or did not know if parent was alive School Attendance and Performance (Appendix Tables A2a–A2c) There was a high percentage of reported school attendance, with almost two-thirds of all respondents in both provinces reporting having never missed school. Additionally, approximately 20 percent of the respondents reported having missed only one to two days of school, and only around 15 percent reported having missed three or more days of school. (See Figure 3.) Among Grade-8 girls, about 25 percent of double orphans missed three or more days of schools, while about 17 percent of single orphans and 15 percent of nonorphans missed three or more days. Among Grade-8 boys, about 28% of double orphans missed three or more days of school, while the proportion for single and nonorphans was 18 percent. (See Table A.2). Overall, learners in both provinces had high optimism of being promoted to the next grade, though optimism was slightly higher in KZN. Ninety percent of Grade-8 boys, 91 percent of Grade-8 girls, and 95 percent of Grade-10 girls reported that they had some chance or a high chance of passing to the next grade in KZN, compared to 82 percent of Grade-8 boys, 83 percent of Grade-8 girls, and 87 percent of Grade-10 girls in MP. Equally positive is that 90 percent of learners, consistent across all groups and in both provinces, felt that they belong in school. 34 30 31 29 35 30 12 11 13 10 15 11 KZN (n=1435) MP (n=1353) KZN (n=1988) MP (n=1618) KZN (n=2008) MP (n=1731) Grade-8 boys Grade-8 girls Grade-10 girls Single Orphan Double Orphan Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 29 Figure 3. Reported school attendance (percentage of learners who reported missing school) Knowledge and Attitudes toward Learning about HIV and AIDS in Life Orientation Participation in and Perception of the Life Orientation Curriculum (Appendix Tables A3a–A3c) Learners were asked how much they have learned about sexuality and HIV-related topics in LO class. The majority answered that they have learned a lot. (See Table 4 for results for Grade-8 girls, by province; similar results were found for Grade-8 boys and Grade-10 girls.) Comparing the provinces, more female Grade-8 learners from MP (80%) indicated that they have learned a lot about sexuality and HIV-related topics in LO class, compared to female Grade-8 learners in KZN (75%). In both provinces, more Grade-10 girls (between 82%–84%) reported having learned a lot compared to their Grade-8 female and male peers (see Appendix Tables A3a–A3c). Notably, the Grade-10 learners in both intervention and control schools are getting the standard LO curriculum and not the upgraded SLP version. The study also asked learners whether they think their parents or caregivers believe that learning about HIV and AIDS in school is a good thing. More than three-quarters of learners in all grades reported that their parents/caregivers consider learning about HIV and AIDS in school to be a good thing. Slightly more than half of the learners reported having talked to their parents/caregivers about the HIV-related topics they had learned about in their LO class; fewer of the Grade-8 girl learners from KZN (51%) indicated talking with parents/caregivers compared to their peers from MP (62%). A lower percentage of Grade-8 girls and a higher percentage of Grade-10 female learners talked to their parents/caregivers (see Tables A3a–A3c). 58 62 57 63 65 68 23 21 27 18 19 19 19 17 17 19 15 13 Grade-8 boys (n=1435) Grade-8 girls (n=1988) Grade-10 girls (n=2008) Grade-8 boys (n=1353) Grade-8 girls (n=1618) Grade-10 girls (n=1731) KZN MP Missed 0 days of school Missed 1–2 days of school Missed 3 or more days of school 30 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table 4. Learners’ Life Orientation experience and perceived relevance, and parental/caregiver attitudes toward and learners’ involvement in Life Orientation among Grade-8 girls and boys in KZN and MP (percentage of learners indicating the statements were very or mostly true compared to a little or not true)* KZN Grade-8 girls (n=1988) MP Grade-8 girls (n=1618) I have learned a lot about sexuality and HIV-related topics in my Life Orientation class. 75.0 80.4 The things we learn about gender roles, sexuality, and HIV in the Life Orientation class are similar to what I experience in my life. 56.7 63.1 I am able to apply some of the things I have learned about gender roles, sexuality, and HIV in the Life Orientation class to my personal life. 66.1 70.2 I talk to my parent/caregiver about the sexual and HIV-related topics I learn about in the Life Orientation class. 51.0 61.6 My parents/caregiver think it is a good thing I am learning about HIV/AIDS in school. 77.2 86.8 *Unweighted Ns and weighted percentages shown; see Table 3 for weighted Ns. Perceived Relevance of the Life Orientation Class Table 4 above also reveals the degree to which learners perceive the information they learn about gender roles, sexuality, and HIV in the LO class is relevant to their lives. More than half of Grade-8 girls indicated that what they had learned about gender roles, sexuality, and HIV-related topics in LO class was similar to what they experienced in life. Following the same logic, learners were asked whether they found LO to be applicable to their personal lives since knowledge gained from LO is intended to be applicable. More than two-thirds of female Grade-8 learners indicated that they had applied the topics from LO to their personal lives. Similar results were found for Grade-8 boys and Grade-10 girls. These findings can be examined in more depth in the midline qualitative data collection that will focus on the importance of the LO program in the lives of the young people. See Tables A3a–A3c for some distinctions by province. Sources of Information on and Knowledge of STIs, including HIV (Appendix Tables A4a-–A4c) To assess learners’ knowledge of HIV and other STIs, the study focused on facts and myths about HIV and other STIs, general sources of information, and school-specific sources of information on HIV and other STIs. This subsection presents findings from the study about learners’ knowledge of HIV and other STIs. Knowledge of STIs, including HIV Table 5 shows the percentage of learners who correctly responded that statements about HIV and other STIs were true or false. Generally, the study revealed low knowledge of HIV and other STIs among learners. For example, in MP, about 20 percent of Grade-8 boys, 13 percent of Grade-8 girls, and 14 percent of Grade-10 girls correctly identified as false the statement that one would definitely know if he or she has an STI because he or she would see or feel symptoms. The ability to identify this statement as false was also low in KZN at about 19 percent, 18 percent, and 15 percent for Grade-8 boys, Grade-8 girls, and Grade-10 girls, respectively. In general, learners in KZN had lower levels of knowledge than in MP. For example, about 14 percent of Grade-8 girls in KZN correctly identified as true the statement that not all STIs are curable, in comparison with 45 percent in MP. Similarly, 28 percent of Grade-8 girls in KZN compared to 43 percent of Grade-8 girls in MP correctly identified as false that oral sex poses no risk for STIs. Finally, about 53 percent of Grade-8 girls in KZN correctly identified as false the statement that one Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 31 can get HIV from kissing a person who is HIV-positive, in comparison with 65 percent in MP. There was no observed pattern of knowledge of STIs and HIV by orphanhood status. In comparison with Grade-8 girls, Grade-10 girls were more knowledgeable about HIV and other STIs. Most Grade-10 girls correctly identified as true that not all STIs are curable (54% in KZN; 57% in MP). Additionally, most Grade-10 girls identified as false that one can get HIV from kissing a person who is HIV-positive (72% in KZN; 78% in MP) and that a woman who is pregnant can do nothing to prevent her baby from being born with HIV (62% in KZN; 66% in MP). 32 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table 5. Percentage of learners who correctly identified the following statements as true or false (correct response in parentheses)* *Unweighted Ns and weighted percentages shown; see Table 3 for weighted Ns. Sources of Information on HIV/AIDS and STIs In KZN, television shows (35%) and radio commercials and public service announcements (34%) were reported as the most important source of HIV information among the Grade-8 girls and boys and the Grade-10 girls (see Figure 4 for sources among Grade-8 girls, by province). In MP, more than half of the learners cited television shows as the most important source for HIV information. Among Grade-8 girls in MP, family members (32%) and friends (29%) also featured highly as sources of HIV information. In both provinces, street performance and cinema were the least-mentioned sources of information about HIV, AIDS, or other STIs. Except for a significant majority of Grade-10 KZN girls citing radio as their most important source, the same pattern was observed among Grade-8 boys and Grade-10 girls in each province (Table A4a–A4c). Of note, school was not a response option for this question; the question that follows asks details on HIV information learned at school. Grade-8 girls Grade-8 boys Grade-10 girls KZN (n=1988) MP (n=1618) KZN (n=1435) MP (n=1353) KZN (n=2008) MP (n=1731) You can usually tell if someone has HIV/AIDS by the way they look. (FALSE) 36.0 41.8 36.1 42.8 47.8 53.2 If you have a STI you will definitely know because you will see/feel symptoms (FALSE) 19.4 19.8 17.7 12.9 15.1 14.3 Not all sexually transmitted infections are curable. (TRUE) 49.0 51.2 13.4 44.8 54.3 57.3 Oral sex poses no risk for STIs (FALSE) 32.6 46.0 27.7 42.8 37.2 50.6 If a mosquito bites you, it can infect you with HIV. (FALSE) 33.1 39.4 33.7 44.2 39.9 43.9 You can get HIV from kissing a person who is HIV-positive. (FALSE) 52.1 60.9 53.0 64.5 72.4 77.8 A woman who is pregnant can do nothing to prevent her baby from being born with HIV. (FALSE) 40.8 44.5 41.4 46.3 62.0 66.2 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 33 Figure 4. Sources of information about HIV and AIDS and other STIs among Grade-8 girls in MP and KZN (%) Figure 5 presents responses to the question concerning activities and people at school from whom learners have received information about HIV, AIDS, and other STIs among Grade-10 girls. In both provinces, more than three-quarters of learners reported the LO educator as the most useful source of information for education about sex, sexuality, and HIV. Fewer Grade-10 girls (77%) in MP than in KZN (87%) cited an LO educator as the main source of HIV, AIDS, and other STI information. Relatively more MP Grade-10 learners cited external people who visit the school (e.g., representatives of nongovernmental organizations and nurses) as the sources from whom they learned the most about sex, sexuality, and HIV. In both provinces, school libraries, computers, and friends or other learners were the least important sources of information in school about HIV, AIDS, and other STIs. See Tables A4a–A4c for a depiction of all learners by grade and province. 34 26 26 35 14 6 15 23 18 29 26 25 54 16 11 29 32 21 0 10 20 30 40 50 60 Radio commercials or public services announcements Radio shows/programs T.V. commercials/public service announcements T.V. shows Cinema/Movies Street performances Friends Family members Library MP (n=1618) KZN (n=1988) 34 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Figure 5. Respondents learned most about sex, sexuality, and HIV at school (percentage of Grade-10 girls) Expectations about the Future, Gender Norms, and Self-Efficacy Expectations about the Future An assessment of how confident youth are in their ability to negotiate their current life situations can be valuable information about future life planning, and the extent of future life planning is related to the level of risky behaviors, including sexual risk behaviors, youth will take on. Respondents were asked to assess their degree of confidence in solving their problems. The study revealed that about 56 percent of the learners across all grades, in both provinces, have high confidence in their ability to solve problems even when others want to quit while fewer respondents felt that they are capable of coming up with many solutions to a problem. This may indicate that most learners feel able to solve problems, even if they come up with only one or two solutions. A slightly greater share of learners from MP reported that when they have a problem they can find ways to solve it, although this still remained below half—with the exception of Grade-8 female learners in MP (52%). A little more than half of the learners felt that even when others want to quit, they can find a solution. Across all of these items, no significant differences were observed between learners in the two provinces. The findings of the research showed that the overwhelming majority of learners were optimistic about their future in the upcoming year and the upcoming five years. Across all groups, more than 80 percent of the learners believed that their life will get better one year from now. Even more learners believed that their life will get better in the next five years. Only 5 percent of the learners thought that their life will get worse in the next year and five years. MP learners were somewhat more optimistic than learners in KZN about life getting better in the next year. However, no difference in their expectations of life getting better in the next five years was observed among the groups in the two provinces. 87 3 7 1 KZN (n=2008) The LO educator Other educators at the school External people who visit the school (e.g., NGOs, nurses, etc.) School library books or computers Friends or other learners at school Other sources Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 35 Full information on aspirations and expectations about the future by the respondents’ gender, grade, and residence are presented in Appendix Tables A5a–A5c. Gender Norm Attitudes This subsection presents findings on learner’s perceptions of gender norms in sexual relationships (sexual communication, sexual pleasure, condom and contraceptive use, and gender roles). Power imbalances in relationships in favor of men mean that women have little or no ability to protect themselves from HIV. In terms of gender attitudes related to preventing HIV, more learners in MP “agreed a lot” or “somewhat agreed” with behaviors that are associated with male power in relationships, in comparison with their counterparts in KZN. For example, a higher proportion of MP female and male Grade-8 learners agreed that a man needs other women, even if things with his wife or partner are fine, compared to female and male Grade-8 learners in KZN who hold a similar opinion (see Figure 6). Similarly, a greater percentage of learners in MP than in KZN agreed that a woman should not initiate sex. Numerous gender-related questions were included as part of the survey; results of these findings by grade, gender, and province can be found in Appendix Tables A6a–A6c. Figure 6. Percentage of Grade-8 female and male learners who “agreed” or “somewhat agreed” with gender attitudes related to sexual behaviors In both provinces, most learners were open-minded regarding gender roles as they relate to reproduction. Most learners agreed that it is important for a father to be present in the lives of his children, even if the mother and father are separated. Most learners also agreed that if a man gets a woman pregnant, then the child is the responsibility of both. Even though most learners reported that the couple should decide together if they want to have children, most learners also thought that it is only the woman’s responsibility to avoid getting pregnant (see Appendix Tables A6a–A6c). 57% 66% 61% 56% 64% 73% 67% 67% 32% 54% 50% 42% 47% 75% 67% 58% 0% 10% 20% 30% 40% 50% 60% 70% 80% It is the man who decides when to have sex. Men need sex more than women do. A man needs other women even if things with his wife/partner are fine. A woman should not initiate sex. Grade-8 girls MP Grade-8 girls KZN Grade-8 boys MP Grade-8 boys KZN 36 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa At the same time, we found large differences by gender in the opinions given on gender equity-related issues (see Table 6). For example, a higher percentage of Grade-8 male learners than Grade-8 and Grade-10 female learners reported the following beliefs: you don’t talk about sex, you just have it; a woman or man who has sex before marriage does not deserve respect; and a woman who carries condoms is “loose.” Grade-8 and Grade-10 female learners provided similar responses. There was little difference between respondents’ opinions across the two provinces (KZN and MP) in the statements presented in Table 6. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 37 Table 6. Percentage of learners who “agree a lot” or “somewhat agree” with statements on gender roles and gender equity issues, by sex and grade* *All percentages are weighted. Self-Efficacy (Appendix Tables A8a–A8c) Learners were asked to score their level of self-efficacy on sexual behaviors, condom use, and ability to refuse sexual coercion. Across the grades in both provinces, a higher percentage of Grade-10 girls reported high confidence in their capacity to talk about condoms and refuse to be pressured into sexual activities. Figure 7 presents the self-efficacy measures for learners from KZN. For all the self-efficacy questions in Figure 8, female Grade-10 learners indicated higher confidence than their Grade-8 peers. Male Grade-8 learners reported high confidence in their ability to talk about condoms with a partner if the partner were asking to have sex. Overall, few learners reported a high degree of confidence that they could refuse to start having sex if they did not want to, even if friends were having sex with their boyfriends or girlfriends. Few learners felt that they could refuse to have sex, even if someone offered them a meal, gifts, money, or favor. Little difference was observed between the reported confidence of Grade-8 girls and Grade-8 boys to refuse sexual coercion when offered something. Similar distinctions by grade and sex were found for MP and can be seen in Appendix Tables A8a–A8c. Figure 7. Self-efficacy around sexual behaviors among learners from KZN, by grade Grade-8 boys Grade-8 girls Grade-10 girls KZN MP KZN MP KZN MP You don’t talk about sex, you just do it 43.3 48.5 21.2 25.6 24.1 30.4 A woman who has sex before she is married does not deserve respect 54.8 56.7 43.3 52.9 37.7 37.7 A man who has sex before he is married does not deserve respect 51.4 55.1 43.2 52.1 35.7 37.8 Women who carry condoms on them are loose 55.0 60.0 44.6 50.7 39.8 42.3 57 57.5 69.6 51.9 57.3 63.6 61.6 49.4 64.1 71.1 76.8 60.1 0 10 20 30 40 50 60 70 80 90 Could refuse to start having sex if did not want to, even if friends were having sex Could refuse to have sex with partner, even if partner was pressuring Able to talk about condoms with partner if partner was asking to have sex Could refuse to have sex even if someone offered a meal, gifts, money, or favor Grade-10 girls Grade-8 girls Grade-8 boys 38 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Encouragingly, most learners (80% or more) reported knowing where to get tested for HIV. The highest number of learners who reported knowing where to get tested for HIV was among Grade-10 female learners. Figure 8. Percentage of learners reporting they know where to get tested for HIV, by sex and grade Sexual Behavior, Risk Perception, and HIV Testing and Counseling Sexual Behavior About 49 percent of Grade-8 boys, 37 percent of Grade-8 girls, and 68 percent of Grade-10 girls in MP reported having a boyfriend or girlfriend at the time of the survey (see Appendix Tables 9a–9c). A smaller proportion of learners in KZN reported having a girlfriend or boyfriend at the time of the survey, but the pattern there, based on grade and sex, was similar to the pattern in MP (KZN values: 43% of Grade-8 males, 28% of Grade-8 females, and 56% of Grade-10 females). In KZN, about 24 percent of Grade-8 boys, 10 percent of Grade-8 girls, and 26 percent of Grade-10 girls reported ever having had sexual intercourse. Learners in MP followed a similar pattern, with the highest and lowest proportions of learners reporting that they ever had sex among Grade-10 girls and Grade-8 girls, respectively (see Figure 8). Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 39 Figure 9. Percentage of learners who report ever having had sexual intercourse, by grade, sex, and province Risk Perception The degree of perceived personal risk is a crucial construct that is associated with risky behavior. People who perceive themselves to be at high risk for acquiring HIV will be more likely to use safe sexual practices. In terms of level of risk perception in relation to HIV, the study revealed that more than half of learners regarded themselves as being at low risk for acquiring HIV (see Figure 9). Comparing the two provinces, more KZN learners than MP learners considered themselves at no risk of acquiring HIV. For example, 65 percent of Grade-8 female learners from KZN perceived themselves to have no chance of acquiring HIV compared to 58 percent of MP girls in the same grade. (See below for analyses of risk perception by sexual experience.) Within the province, little difference was observed among the groups. A similar pattern emerged in terms of the learners’ perceived chance of already being HIV-positive. The overwhelming majority of learners believed that they did not have HIV, with more KZN learners than MP learners reporting “no chance.” See details by province, grade, and intervention or control group in Appendix Tables A7a–A7c. 24 10 26 20 10 32 Grade-8 boys Grade-8 girls Grade-10 girls KZN MP 40 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Figure 10. Levels of concern among learners about risk for HIV or AIDS during their lifetime, by grade, sex, and province It is likely that those learners who had ever had sex had different perceptions of HIV risk compared to those who had never had sex. In Figure 10, we provide a similar graphic to the one above, by grade/sex and whether or not the young person was sexually experienced (both provinces are combined). This figure shows that among Grade-8 girls and Grade-10 girls, those who had ever had sex reported significantly higher perceived risk of HIV than those who had never had sex (i.e., greater reporting of some or a high risk of HIV acquisition). Among Grade-8 boys, there was no significant difference in perceived HIV risk by sexual experience. 61 55 65 58 60 57 28 30 21 29 31 29 12 15 14 13 9 14 KZN MP KZN MP KZN MP Grade-8 boys Grade-8 girls Grade-10 girls High chance Some chance No chance Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 41 *Significant difference, p<0.001 Figure 11. Levels of concern among learners about risk for HIV or AIDS during their lifetime, by grade, sex, and prior sexual experience Risky Sexual Behaviors among Sexually Experienced Learners Among those who reported ever having had sexual intercourse in MP, the mean number of reported sex partners in the past 12 months was highest among Grade-8 boys, at about 6; lower among Grade-8 girls, at about 5; and lowest among Grade-10 girls, at about 2. Learners in KZN followed a similar pattern (see Table 7). Approximately one-half of Grade-8 boys and girls reported their most recent or current sex partner attended the same school as the learners; among Grade-10 girls, this figure was about 40 percent. Reported condom use at last sex varied widely by learners’ grade, sex, and province: about 76 percent of Grade-8 boys in each province reported using a condom at last sex; about 62 percent and 80 percent of Grade-8 girls in KZN and MP, respectively, reported using a condom at last sex; and 62 percent and 73 percent of Grade-10 girls in KZN and MP, respectively, reported using a condom at last sex. There was no observed pattern of condom use at last sex, by orphanhood status. Reports of consistent condom use were higher among learners in MP than in KZN. For example, among Grade-8 girls who had ever had sex, about 73 percent in MP, compared to about 54 percent in KZN, reported using a condom every time they had sex in the past three months. Consistent condom use was generally lowest among double orphans, with the lowest among Grade-8 girls in KZN, at 33 percent. Notably, among Grade-8 girls in MP, 92 percent of double orphans reported using a condom every time they had sex in the past three months, compared to 64 percent among single orphans and 70 percent among nonorphans; one caveat of these analyses is that the sample of sexually experienced by orphanhood status is small. In MP, reports of giving or receiving money, gifts, or favors in exchange for sex at last sex were highest among Grade-8 boys (46% and 43% for giving and receiving, respectively), lower among Grade-8 girls (33% and 39%, respectively), and lowest among Grade-10 girls (17% and 23%, respectively). Learners in KZN followed a similar pattern, though the proportions were lower in each grade/sex group of learners. See Table A9 for details of sexual behaviors by province, grade, sex, and intervention and control group. 60 56 65 52 63 51 28 30 22 32 28 37 12 14 13 16 9 12 Never sex Ever sex Never sex Ever sex Never sex Ever sex Grade-8 boys Grade-8 girls* Grade-10 girls* A high chance Some chance No chance 42 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table 7. Risky sexual behaviors among learners who reported ever having sex, by age and grade* KwaZulu-Natal Mpumalanga Grade-8 boys (n=335) Grade-8 girls (n=183) Grade￾10 girls (n=521) Grade-8 boys (n=286) Grade-8 girls (n=163) Grade-10 girls (n=578) Mean number of sex partners in past 12 months 4.4 3.6 2.0 5.8 4.5 1.8 Percentage who used a condom at last sex 76.2 62.0 61.8 76.2 80.3 72.8 Percentage who used a condom every time in past 3 months 67.1 54.4 56.2 73.7 73.3 67.3 Percentage who, at last sex, gave or money, gifts, or favors in exchange 31.6 30.1 15.4 45.8 32.9 17.4 Percentage who, at last sex, received money, gifts, or favors in exchange 33.2 31.0 15.9 42.8 39.1 22.7 Unweighted Ns and weighted percentages shown; see Appendix Tables A9a-A9c for weighted Ns. Pregnancy Experience Female learners who reported that they had ever had sex were asked if they had ever been pregnant. Given the small number of learners who had ever had sex, we present the pregnancy experience by grade only, with a focus on those learners within two years of their age for grade. Figure 11 shows that there is little difference in pregnancy experience between the Grade-8 and Grade-10 female learners. In particular, one￾fifth of sexually experienced female learners have ever been pregnant. Notably, there were more Grade-10 ever-pregnant girls, because the percentage of girls who had ever had sex was higher in Grade 10 than in Grade 8. Figure 12. Percentage of sexually experienced female learners who had ever been pregnant 24 27 76 73 Grade 8 Grade 10 Yes No Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 43 HIV Testing and Counseling About 48 percent of Grade-8 boys, 34 percent of Grade-8 girls, and 40 percent of Grade-10 girls in KZN reported having been to a clinic or hospital in the past 12 months to get information about sex-related issues. Learners in MP followed the same pattern (see Tables A10a–A10c for details). More learners in KZN than MP had ever been tested for HIV. For example, 42 percent of Grade-8 girls in KZN compared to 35 percent of Grade-8 girls in MP had ever been tested for HIV (see Figure 12). Similar patterns are found for the male learners and the Grade-10 learners. Among learners in both provinces who had ever had an HIV test, Grade-10 girls were the most likely to have received and shared the results of the test with someone else; 91 percent of Grade-10 girls in MP had received the results, and 72 percent had shared the results with someone else. See Appendix Tables A10a–A10c for details on HIV testing and counseling by grade, province, and sex. Figure 13. Percentage of learners who have ever been tested for HIV, by grade, sex, and province 53 42 56 50 35 50 Grade-8 boys Grade-8 girls Grade-10 girls KZN MP 44 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa DISCUSSION This baseline report has demonstrated some important characteristics of the learners in the South African schools from KZN and MP who participated in this study. First, it is worth noting that few, if any, differences were observed within a province between the intervention or control group. This indicates that randomization as part of the evaluation study was successful. This is important for the overall evaluation and means that an intention to treat analysis is appropriate for the study data. The baseline data also indicate that many of the learners are orphaned in the study sample, either as single orphans (about one-third) or double orphans (one-tenth). The mother is most often the primary caregiver, followed in frequency by a grandparent. This is an important consideration for programs that are seeking to reach young people with messages about HIV. In the South African context, with high HIV prevalence, most young people have had experience with HIV and AIDS, either within their household or nearby. Programs need to avoid stigmatizing HIV in their messages and approaches. Most learners get information on HIV and AIDS and STIs from television and radio; and at school, LO educators are the most common source of information on sex, sexuality, and HIV. Many of the young people in the sample value and appreciate what they are learning in their LO curriculum. With the new SLPs, it will be possible to see if these attitudes improve over time, or if knowledge of HIV risks and myths about HIV change with improved teaching methods, particularly in the intervention schools. For the Grade-8 girls and boys, the program needs to consider issues of gender norms, self-efficacy, and aspirations about the future. These baseline data indicate areas where young people are already excelling, but they also indicate gaps in young people’s knowledge, experience, and self-efficacy that can be addressed through the SLPs of the LO program and supplemental activities. Finally, there exist opportunities as well as the need to intervene before young people become sexually experienced. Even though the majority of the young people surveyed had never had sex, there are HIV prevention needs that the LO program should address. Among the Grade-8 boys, one-quarter in KZN and one-fifth in MP were sexually active. Although three-quarters of these boys reported having used a condom the last time they had sex—and two-thirds to three-quarters reported using a condom every time—there are still gaps in condom use that put these young people at risk of acquiring HIV. Furthermore, by Grade 10, one-quarter of the female learners were sexually active, and about 59 percent of them had used a condom at last sex or consistently. The Grade-8 female learners were the least likely to be sexually experienced (10%), but many will initiate sex in the next couple of years, so information about safer sex practices (i.e., protection against HIV and unintended pregnancy) is important for these young girls. Program activities must motivate girls and boys to avoid pregnancy and HIV and AIDS by first helping them recognize their risk, and then helping them develop a prevention plan. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 45 CONCLUSION The EDC/DBE SLPs for the LO program, with supplemental activities, is seeking to improve knowledge, attitudes, self-efficacy, and behaviors among young people in South Africa. This evaluation will illuminate gaps in the program and ways to strengthen the program to meet the needs of all learners. The EDC team can use this baseline report to determine if they are currently meeting the needs of young people. Over time, with the midline and end line data, they will be able to see whether knowledge, attitudes, self-efficacy, and sexual and health-seeking behaviors improve with exposure to the LO program. The evidence produced by this large-scale evaluation of the LO program will be of great value prior to national-level scale-up. Evaluation results may also be used to inform the design and implementation of life skills curricula and related HIV prevention activities for young people in other developing countries. The next step for the evaluation will be to collect the midline quantitative data among Grade-9 female and male learners as well as qualitative data from learners, educators, principals, and administrators. This information will provide a context of initial changes and satisfaction with the program that can be used for program strengthening. Midline data collection will happen in the third quarter of the 2017 school year and will provide information on program exposure; changes in knowledge, attitudes, and behaviors; and improvements in the school environments with the new LO program. 46 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa REFERENCES Abdool Karim, Q., Kharsany, A. B., Leask, K., Ntombela, F., Humphries, H., Frohlich, J., . . . Abdool Karim, S. S. (2014). Prevalence of HIV, HSV-2 and pregnancy among high school learners in rural KwaZulu-Natal, South Africa: A bio-behavioural cross-sectional survey. Journal of Sexually Transmitted Infections, 90 (8): 620–626. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/24873967 Department of Basic Education. (2010). Integrated strategy on HIV and AIDS 2012–2016: Full report. Pretoria, South Africa: Department of Basic Education. Retrieved from http://www.gov.za/documents/draft-integrated-strategy-hiv-and-aids-2012-2016 Duflo, E., Dupas, P., Kremer, M., & Sinei, S. (2006). Education and HIV/AIDS Prevention: Evidence from a Randomized Evaluation in Western Kenya. Retrieved from https://openknowledge.worldbank.org/handle/10986/9007 Dupas, P. (2006). Relative risks and the market for sex: Teenagers, sugar-daddies and HIV in Kenya. (MPRA Paper No. 248) Munich, Germany: Munich University Library. Retrieved from https://mpra.ub.uni-muenchen.de/248/ Jewkes, R., Nduna, M., Levin, J., Jama, N., Dunkle, K., Puren, A., & Duvvury, N. (2008). Impact of Stepping Stones on incidence of HIV and HSV-2 and sexual behaviour in rural South Africa: Cluster randomized controlled trial. The British Medical Journal, (37): 1–11. Retrieved from http://www.bmj.com/content/337/bmj.a506 Kirby, D., Laris, B. A., & Rolleri, L. (2006). Sex and HIV education programs for youth: Their impact and important characteristics. Scotts Valley, CA: ETR Associates. Retrieved from http://hivhealthclearinghouse.unesco.org/library/documents/sex-and-hiv-education-programs-youth￾their-impact-and-important-characteristics Kish, Leslie. (1965). Survey sampling. Biometrical Journal, 10(1), 1–95. Retrieved from http://onlinelibrary.wiley.com/doi/10.1002/bimj.19680100122/abstract McCoy, S. I., Kangwende, R. A., & Padian, N. S. (2010). Behavior change interventions to prevent HIV infection among women living in low and middle income countries: A systematic review. AIDS and Behavior, 14(3) 469–482. Retrieved https://www.ncbi.nlm.nih.gov/pubmed/19949847 McQueston, K., Silverman, R., & Glassman, A. (2013). The efficacy of interventions to reduce adolescent childbearing in low- and middle-income countries: A systematic review. Studies in Family Planning, 44 (4): 369-388. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/24323658 Minnis, A. M., & Padian, N. S. (2001). Reliability of adolescents’ self-reported sexual behavior: A comparison of two diary methodologies. Journal of Adolescent Health, 28(15), 394–403. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/11336869 Republic of South Africa Department of Health. (2015). National HIV Counselling and Testing Policy Guidelines. Pretoria, South Africa: South Africa Department of Health. Retrieved from https://www.health-e.org.za/wp-content/uploads/2015/07/HCT-Guidelines-2015.pdf Ross, D., Changalucha, J., Obasi, A. I., Todd, J., Plummer, M. L., Cleophas-Mazige, B. Anemona, A., . . . Hayes, R. J. (2007). Biological and behavioral impact of an adolescent sexual health intervention in Tanzania: A community-randomized trial. AIDS, 21, 1943–1955. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/17721102 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 47 Shisana, O., Rehle, T., Simbayi, L. C., Zuma, K., Jooste, S., Zungu, N., . . . Onoya, D. (2014). South African national HIV prevalence, incidence and behaviour survey, 2012. Cape Town, South Africa: Human Sciences Research Council (HSRC) Press. Retrieved from http://www.hsrc.ac.za/en/research-data/view/6871 Statistics South Africa. (2014). General household survey 2013. Pretoria, South Africa: Statistics South Africa. Retrieved from https://www.statssa.gov.za/publications/P0318/P03182013.pdf Statistics South Africa. (2015). General household survey 2014. Pretoria, South Africa: Statistics South Africa. Retrieved from http://www.statssa.gov.za/publications/P0318/P03182014.pdf 48 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa APPENDIX 1. BASELINE FINDINGS Table Set A1. Demographics characteristics Table A1a. Demographic characteristics by province and intervention group, Grade-8 girls* KwaZulu-Natal Mpumalanga Intervention (n=1,064)** Control (n=924)** Total (n=1,988)* * Intervention (n=833)** Control (n=785)** Total (n=1,618)* * Mean age (range) 13.6 (11-25) 13.6 (11-25) 13.6 (11-25) 13.6 (11-25) 13.5 (11-25) 13.6 (11-25) Primary caregiver Mother Father Older sibling Aunt and/or uncle Grandparent Cousin Other adult Other child Respondent 74.2 5.1 3.6 3.3 12.3 0.7 0.5 0.1 0.3 73.4 5.7 3.4 3.8 11.7 1.5 0.4 0.1 0.1 73.9 5.3 3.5 3.5 12.0 1.0 0.5 0.1 0.1 74.0 7.4 3.4 2.6 11.0 0.9 0.4 0.1 0.3 74.5 5.4 4.0 2.3 12.1 0.4 0.7 0 0.5 74.3 6.4 3.7 2.4 11.5 0.7 0.5 0.1 0.4 Orphanhood*** Not an orphan Single orphan Double orphan 54.8 31.7 13.6 57.4 30.6 12.0 55.9 31.2 12.9 59.9 30.0 10.1 63.2 27.5 9.4 61.5 28.8 9.7 Number of household items (range=0-6)**** 1.2 (0-6) 1.4 (0-6) 1.3 (0-6) 1.1 (0-6) 1.0 (0-6) 1.0 (0-6) *Missing data ≤3.9% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. ***Child defined as orphan if s/he reported parent was not alive or did not know whether parent was alive. ****Calculated as a sum score of number of different items that the respondent reported having in his or her household. Items were electricity, radio, tap water, television, refrigerator, and car. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 49 Table A1b. Demographic characteristics, by province and intervention group, Grade-8 boys* KwaZulu-Natal Mpumalanga Intervention (n=736)** Control (n=699)** Total (n=1,435)** Intervention (n=654)** Control (n=699)** Total (n=1,353)* * Mean age (range) 14.2 (11-25) 14.4 (11-24) 14.3 (11-25) 14.5 (11-25) 14.2 (11-25) 14.3 (11-25) Primary caregiver Mother Father Older sibling Aunt and/or uncle Grandparent Cousin Other adult Other child Respondent 66.4 13.1 3.6 3.6 11.2 0.4 1.2 0.2 0.4 72.4 11.8 3.6 1.2 10.6 0.1 0.3 0.0 0.1 69.0 12.5 3.6 2.6 10.9 0.3 0.8 0.1 0.3 71.8 10.5 5.2 2.8 7.9 0.2 1.0 0.4 0.2 64.1 15.0 2.6 3.1 12.0 0.9 0.9 0.9 0.5 67.9 12.7 3.9 2.9 10.0 0.5 1.0 0.6 0.4 Orphanhood*** Not an orphan Single orphan Double orphan 53.4 35.1 11.5 56.2 31.8 12.0 54.6 33.7 11.7 58.6 30.0 11.4 60.7 29.4 9.9 59.6 29.7 10.7 Number of household items (range=0-6)**** 1.2 (0-6) 1.3 (0-6) 1.2 (0-6) 1.0 (0-6) 0.9 (0-6) 1.0 (0-6) *Missing data ≤3.1% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. ***Child defined as orphan if s/he reported parent was not alive or did not know whether parent was alive. ****Calculated as a sum score of number of different items respondent reported having in his or her household. Items include electricity, radio, tap water, television, refrigerator, and car. 50 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table A1c. Demographic characteristics, by province and intervention group, Grade-10 girls* KwaZulu-Natal Mpumalanga Intervention (n=1,033)** Control (n=975** Total (n=2,008)** Intervention (n=866)** Control (n=845)** Total (n=1,731)* * Mean age (range) 16.0 (11-25) 16.1 (11-25) 16.1 (11-25) 16.0 (11-25) 15.8 (11-25) 15.9 (11-25) Primary caregiver Mother Father Older sibling Aunt and/or uncle Grandparent Cousin Other adult Other child Respondent 71.3 7.6 4.8 3.5 10.6 0.8 1.0 0.2 0.2 70.7 6.9 2.8 4.8 12.1 1.3 1.2 0.2 0.1 71.0 7.3 3.9 4.1 11.3 1.0 1.1 0.2 0.2 76.0 5.1 4.0 2.7 10.4 0.6 0.9 0.1 0.2 70.3 6.1 3.6 3.9 14.2 0.7 0.6 0.3 0.3 73.2 5.6 3.8 3.3 12.3 0.6 0.8 0.2 0.2 Orphanhood*** Single orphan Double orphan 50.7 34.7 14.5 49.1 36.1 14.8 50.0 35.4 14.7 57.9 30.8 11.3 61.2 28.8 10.0 59.6 29.8 10.7 Number of household items (range=0-6)**** 1.4 (0-6) 1.5 (0-6) 1.4 (0-6) 1.0 (0-6) 0.9 (0-6) 0.9 (0-6) *Missing data ≤3.1% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. ***Child defined as orphan if s/he reported parent was not alive or did not know whether parent was alive. ****Calculated as a sum score of number of different items respondent reported having in his or her household. Items include electricity, radio, tap water, television, refrigerator, and car. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 51 Table Set A2. School attendance, performance, and relationships Table A2a. School attendance, performance, and relationships, by province and intervention group, Grade-8 girls* KwaZulu-Natal (n=1,988)** Mpumalanga (n=1,618)** Intervention Control Total Intervention Control Total Number of days of missed school None 1–2 days 3–4 days 5–6 days 7 or more days 63.6 20.0 7.1 3.6 5.7 58.8 23.0 7.8 4.5 6.0 61.6 21.2 7.4 4.0 5.8 64.4 18.3 6.3 4.2 6.7 65.5 19.1 5.6 3.8 5.9 65.0 18.7 5.9 4.0 6.3 Number of days late to school None 1–2 days 3–4 days 5–6 days 7 or more days 60.7 23.7 7.5 2.9 5.1 60.8 25.1 6.7 3.0 4.4 60.8 24.3 7.2 3.0 4.8 61.7 27.5 6.5 1.4 3.0 66.6 23.6 5.3 1.4 3.1 64.1 25.6 5.9 1.4 3.0 Overall mark for last term ≤20% 20–29% 30–39% 40–49% 50–59% 60–69% 70–79% ≥80% 5.8 5.9 11.4 17.7 21.4 16.2 10.3 11.4 4.5 5.2 10.3 18.0 22.9 18.2 12.9 8.1 5.3 5.6 10.9 17.8 22.0 17.0 11.3 10.0 5.1 5.5 13.1 23.9 23.2 14.2 11.1 8.4 5.6 5.1 14.3 17.1 20.1 18.0 12.0 7.8 5.3 5.3 13.7 18.3 21.7 16.0 11.6 8.1 Learner feels like s/he belongs in school 92.0 91.6 91.8 90.9 89.3 90.1 Learner feels s/he has some chance or high chance of being promoted to the next grade 90.4 92.3 91.2 83.7 82.3 83.0 *Missing data ≤2.2% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. 52 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table A2b. School attendance, performance, and relationships, by province and intervention group, Grade-8 boys* *Missing data ≤1.7% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. KwaZulu-Natal (n=1,435)** Mpumalanga (n=1,353)** Intervention Comparison Total Intervention Comparison Total Number of days of missed school None 1–2 days 3–4 days 5–6 days 7 or more days 58.92 23.26 8.76 3.39 6.3 58.2 22.09 7.65 4.4 7.65 58.25 22.76 8.28 3.83 6.89 64.36 18.63 6.05 4.99 5.98 62.16 17.39 7.72 4.61 8.13 63.26 18.01 6.88 4.8 7.06 Number of days late to school None 1–2 days 3–4 days 5–6 days 7 or more days 53.37 26.86 10 5.13 4.64 53.07 27.08 10.14 3.66 6.06 53.24 26.94 10.06 4.49 5.25 59.8 26.16 6.68 4.47 2.89 56 27.75 8.72 3.78 3.74 57.91 26.95 7.7 4.13 3.31 Overall mark for last term ≤20% 20–29% 30–39% 40–49% 50–59% 60–69% 70–79% ≥80% 8.32 7.25 12.98 23 21.29 13.62 7.24 6.3 7.77 7.26 12.29 21.21 22.07 13.57 10.71 5.12 8.08 7.25 12.68 22.22 21.62 13.6 8.74 5.79 7.35 8.57 16.4 21.2 22.4 11.58 5.73 6.72 8.06 8.85 15.07 20.82 20.74 13.06 7.14 6.26 7.7 8.71 15.75 21.01 21.57 12.32 6.43 6.49 Learner feels like s/he belongs in school 92.45 92.55 92.49 88.05 87.14 87.6 Learner feels s/he has some chance or high chance of being promoted to the next grade 87.33 92.52 90.1 82.09 81.07 81.58 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 53 Table A2c. School attendance, performance, and relationships, by province and intervention group, Grade-10 girls* KwaZulu-Natal (n=2,008)** Mpumalanga (n=1,731)** Intervention Comparison Total Intervention Comparison Total Number of days of missed school None 1–2 days 3–4 days 5–6 days 7 or more days 59.0 26.5 6.4 3.8 4.3 54.4 26.7 8.8 3.7 6.3 56.9 26.6 7.5 3.8 5.2 65.5 20.0 4.0 3.7 6.8 71.0 17.4 5.0 2.2 4.5 68.2 18.7 4.5 3.0 5.6 Number of days late to school None 1–2 days 3–4 days 5–6 days 7 or more days 56.0 28.1 8.5 3.3 4.2 61.2 25.4 8.0 1.7 3.6 58.4 26.9 8.2 2.5 3.9 64.3 26.3 4.5 2.2 2.8 63.9 23.9 6.6 2.4 3.3 64.1 25.1 5.5 2.3 3.1 Overall mark for last term ≤20% 20–29% 30–39% 40–49% 50–59% 60–69% 70–79% ≥80% 3.4 5.8 18.5 29.4 20.1 12.9 7.2 2.9 3.9 6.8 19.2 27.6 20.6 11.3 6.9 3.7 3.6 6.2 18.8 28.5 20.3 12.2 7.0 3.3 3.7 6.3 21.1 30.1 20.6 10.3 4.7 3.2 1.5 5.5 18.9 31.9 22.79 10.19 6.1 2.74 2.7 5.9 20.0 331.0 21.66 10.7 5.5 2.3 Learner feels like s/he belongs in school 91.2 91.6 91.8 90.9 89.3 90.1 Learner feels s/he has some chance or high chance of being promoted to the next grade 96.17 93.8 95.1 87.3 86.9 87.0 *Missing data is ≤0.7% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. 54 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table Set A3. Participation in and perceptions of Life Orientation Curriculum Table A3a. Participation in and perceptions of Life Orientation Curriculum, by province and intervention group, Grade-8 girls* ¥ KwaZulu-Natal (n=1,988)** Mpumalanga (n=1,618)** Intervention Control Total Intervention Control Total Respondent indicated the following is mostly true or very true (reference category is not true or a little true) The things we learn about gender roles, sexuality, and HIV in the Life Orientation class are similar to what I experienced in my life. 57.7 55.2 56.7 64.0 62.1 63.1 I have learned a lot about sexuality- and HIV- related topics in my Life Orientation class. 76.2 73.0 75.0 83.1 77.6 80.4 I am able to apply some of the things I have learned about gender roles, sexuality, and HIV in the Life Orientation class to my personal life. 66.7 65.4 66.1 70.6 69.8 70.2 I talk to my parent/caregiver about the sexual- and HIV￾related topics I learn in the Life Orientation class. 51.7 50.0 51.0 63.7 59.5 61.6 My parents/caregiver think it is a good thing I am learning about HIV/AIDS in school. 77.7 76.6 77.2 87.3 86.3 86.8 *Missing data ≤2.4% ¥Cronbach’s alpha for participation and perception of LO scale for all Grade-8 girls = 0.72. **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 55 Table A3b. Participation and perception of life orientation curriculum, by province and intervention group, Grade-8 boys* ¥ KwaZulu-Natal (n=1,435)** Mpumalanga (n=1,353)** Intervention Control Total Intervention Control Total Respondent indicated the following is mostly true or very true (reference category is not true or a little true) The things we learn about gender roles, sexuality, and HIV in the Life Orientation class are similar to what I experienced in my life. 54.1 58.0 55.8 64.1 61.8 63.0 I have learned a lot about sexuality- and HIV-related topics in my Life Orientation class. 71.1 70.0 70.6 78.0 74.4 76.2 I am able to apply some of the things I have learned about gender roles, sexuality, and HIV in the Life Orientation class to my personal life. 61.1 61.0 61.0 62.6 66.3 64.5 I talk to my parent/caregiver about the sexual- and HIV￾related topics I learn in the Life Orientation class. 46.5 49.7 47.9 48.0 50.8 49.4 My parents/caregiver think it is a good thing I am learning about HIV/AIDS in school. 74.5 74.8 74.6 79.3 79.3 79.3 *Missing data ≤2.1% ¥Chronbach’s alpha for participation and perception of LO scale for all Grade-8 boys= 0.75. **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. 56 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table A3c. Participation and perception of life orientation curriculum, by province and intervention group, Grade-10 girls* ¥ KwaZulu-Natal (n=2,008)** Mpumalanga (n=1,731)** Intervention Control Total Intervention Control Total Respondent indicated the following is mostly true or very true (reference category is not true or a little true) The things we learn about gender roles, sexuality, and HIV in the Life Orientation class are similar to what I experienced in my life. 60.4 61.2 60.7 69.3 65.4 67.4 I have learned a lot about sexuality- and HIV-related topics in my Life Orientation class. 83.0 81.8 82.4 85.1 82.9 84.0 I am able to apply some of the things I have learned about gender roles, sexuality, and HIV in the Life Orientation class to my personal life. 70.7 69.2 70.0 74.6 70.8 72.7 I talk to my parent/caregiver about the sexual- and HIV￾related topics I learn in the Life Orientation class. 42.2 43.4 42.7 54.6 52.2 53.4 My parents/caregiver think it is a good thing I am learning about HIV/AIDS in school. 83.9 81.1 82.6 89.0 89.2 89.1 *Missing data ≤0.7% ¥Chronbach’s alpha for participation and perception of LO scale for all Grade-10 girls = 0.69. **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 57 Table Set A4. Knowledge of STIs/HIV Table A4a. Knowledge of STIs/HIV, by province and intervention group, Grade-8 girls* KwaZulu-Natal (n=1,988)** Mpumalanga (n=1,618)** Intervention Control Total Intervention Control Total Respondent indicated correctly whether the following statements are true or false (CORRECT ANSWER IN PARENTHESES) You can usually tell if someone has HIV/AIDS by the way they look. (FALSE) 36.3 35.23 36.1 44.0 41.3 42.8 If you have an STI you will definitely know because you will see/feel symptoms. (FALSE) 17.4 17.9 17.7 13.0 12.8 12.9 Not all sexually transmitted infections are curable. (TRUE) 44.4 13.8 13.4 48.2 41.4 44.8 Oral sex poses no risk for STIs. (FALSE) 29.1 25.6 27.7 42.2 43.5 42.8 If a mosquito bites you it can infect you with HIV. (FALSE) 34.5 32.6 33.7 43.6 44.8 44.2 'You can get HIV from kissing a person who is HIV-positive. (FALSE) 52.6 53.6 53.0 63.1 65.9 64.5 A woman who is pregnant can do nothing to prevent her baby from being born with HIV. (FALSE) 41.1 41.9 41.4 47.6 44.9 46.3 58 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Respondents indicated they heard or saw information on HIV/AIDS from… Radio commercials or public services announcements 36.4 30.9 34.1 30.0 27.6 28.9 Radio shows/programs 26.5 26.1 26.3 25.7 26.7 26.2 T.V. commercials/public service announcements 24.8 26.5 25.5 26.9 22.1 24.5 T.V. shows 35.9 34.5 35.3 53.0 54.9 53.9 Cinema/movies 13.2 14.2 13.6 17.4 15.0 16.2 Street performances 6.3 6.6 6.4 10.0 12.0 11.0 Friends 15.2 15.6 15.3 27.5 29.9 28.6 Family members 22.5 23.6 23.0 32.7 31.9 32.3 Library 18.0 16.7 17.5 21.7 20.5 21.1 At schools, respondents learned most about sex, sexuality and HIV from… The LO educator 86.2 85.1 85.7 84.1 77.2 80.7 Other educators at the school 4.7 5.6 5.1 3.8 4.2 4.0 External people who visit the school (e.g., NGO representatives and nurses) 5.0 5.1 5.1 7.2 13.3 10.2 School library books or computers 2.2 2.4 2.2 1.7 1.7 1.7 Friends or other learners at school 1.1 0.5 0.9 2.0 2.3 2.2 Other sources 0.8 1.2 1.0 1.2 1.4 1.3 *Missing data ≤5.0% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 59 Table A4b. Knowledge of STIs/HIV, by province and intervention group, Grade-8 boys* KwaZulu-Natal (n=1,435)** Mpumalanga (n=1,353)** Intervention Control Total Intervention Control Total Respondent indicated correctly whether the following statements are true or false (CORRECT ANSWER IN PARENTHESES) You can usually tell if someone has HIV/AIDS by the way they look. (FALSE) 36.8 35.0 36.0 41.9 41.7 41.8 If you have an STI you will definitely know because you will see/feel symptoms. (FALSE) 19.7 19.1 19.4 19.9 19.6 19.8 Not all sexually transmitted infections are curable. (TRUE) 48.5 49.7 49.0 51.9 50.5 51.2 Oral sex poses no risk for STIs. (FALSE) 35.7 28.6 32.6 46.7 45.3 46.0 If a mosquito bites you it can infect you with HIV. (FALSE) 32.4 34.0 33.1 38.5 40.5 39.4 'You can get HIV from kissing a person who is HIV-positive. (FALSE) 50.7 54.1 52.1 60.4 61.4 60.9 A woman who is pregnant can do nothing to prevent her baby from being born with HIV. (FALSE) 42.2 39.1 40.8 44.7 44.3 44.5 60 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Respondents indicated they heard or saw information on HIV/AIDS from… Radio commercials or public services announcements 39.6 38.0 38.9 30.7 30.5 30.6 Radio shows/programs 30.0 28.4 29.3 26.1 27.3 26.7 T.V. commercials/public service announcements 30.0 28.9 29.5 21.2 25.3 23.2 T.V. shows 36.5 38.7 37.4 53.7 55.0 54.4 Cinema/movies 16.5 15.9 16.3 15.5 16.4 15.9 Street performances 8.4 6.8 7.7 7.2 9.0 8.1 Friends 13.8 16.7 15.1 24.3 28.1 26.2 Family members 16.6 16.7 16.6 22.5 27.5 25.0 Library 13.7 14.0 13.8 14.5 16.7 15.6 At schools, respondents learned most about sex, sexuality and HIV from… The LO educator 83.1 84.9 83.9 83.2 82.0 82.6 Other educators at the school 5.4 4.9 5.2 4.5 5.0 4.7 External people who visit the school (e.g., NGO representatives and nurses) 5.5 5.6 5.6 6.7 8.1 7.4 School library books or computers 3.0 2.1 2.6 1.9 2.0 1.9 Friends or other learners at school 2.1 1.0 1.6 2.5 2.1 2.3 Other sources 0.8 1.5 1.1 1.3 0.9 1.1 *Missing data ≤4.0% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 61 Table A4c. Knowledge of STIs/HIV, by province and intervention group, Grade-10 girls* KwaZulu-Natal (n=2,008)** Mpumalanga (n=1,731)** Intervention Control Total Intervention Control Total Respondent indicated correctly whether the following statements are true or false (CORRECT ANSWER IN PARENTHESES) You can usually tell if someone has HIV/AIDS by the way they look. (FALSE) 47.2 48.5 47.8 52.2 54.2 53.2 If you have an STI you will definitely know because you will see/feel symptoms. (FALSE) 15.2 15.0 15.1 15.6 13.0 14.3 Not all sexually transmitted infections are curable. (TRUE) 54.7 53.7 54.3 58.0 56.6 57.3 Oral sex poses no risk for STIs. (FALSE) 38.1 36.1 37.2 51.1 50.2 50.6 If a mosquito bites you it can infect you with HIV. (FALSE) 40.4 39.3 39.9 45.2 42.6 43.9 'You can get HIV from kissing a person who is HIV-positive. (FALSE) 74.7 69.8 72.4 78.0 77.6 77.8 A woman who is pregnant can do nothing to prevent her baby from being born with HIV. (FALSE) 62.0 62.1 62.0 67.0 65.4 66.2 62 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Respondents indicated they heard or saw information on HIV/AIDS from… Radio commercials or public services announcements 64.9 67.8 66.3 26.4 27.7 27.1 Radio shows/programs 24.6 28.4 26.4 28.0 28.5 28.2 T.V. commercials/public service announcements 24.2 24.9 24.5 21.7 23.4 22.6 T.V. shows 39.3 38.3 38.9 53.7 55.1 54.4 Cinema/movies 11.2 14.6 12.8 11.5 12.7 12.1 Street performances 7.3 6.2 6.8 9.3 8.9 9.1 Friends 19.9 18.5 19.2 27.4 28.4 27.9 Family members 27.3 24.5 26.0 30.8 34.5 32.7 Library 20.8 17.3 19.2 15.5 17.4 16.5 At schools, respondents learned most about sex, sexuality and HIV from… The LO educator 87.5 85.7 86.7 75.8 77.7 76.7 Other educators at the school 2.8 3.4 3.1 4.9 4.1 4.5 External people who visit the school (e.g., NGO representatives and nurses) 6.6 7.2 6.8 14.1 12.6 13.4 School library books or computers 1.1 1.4 1.3 2.0 1.0 1.5 Friends or other learners at school 1.4 1.5 1.5 1.8 3.3 2.5 Other sources 0.6 0.8 0.7 1.3 1.4 1.3 *Missing data ≤2.0% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 63 Table Set A5. Confidence and expectations about the future Table A5a. Confidence¥ and expectations about the future, by province and intervention group, Grade-8 girls* KwaZulu-Natal (n=1,988)** Mpumalanga (n=1,618)** Intervention Control Total Intervention Control Total Respondent indicated that the following is true most or all of the time (reference category is none or some of the time) When I have a problem, I can come up with lots of ways to solve it. 42.5 43.5 42.9 51.6 51.9 51.7 The things I have done in the past will help me in the future. 57.7 59.7 58.5 52.6 53.2 52.9 I believe I can find ways to solve a problem even when others want to quit. 51.1 49.5 50.5 57.8 58.9 58.3 Thinks life will be better, about the same, or worse 1 year from now Better Same Worse 78.0 17.6 4.4 78.7 17.5 3.9 78.3 17.5 4.2 84.9 11.2 4.0 83.1 12.7 4.2 84.0 11.9 4.1 Thinks life will be better, the same, or worse 5 years from now Better Same Worse 77.7 16.6 5.7 76.1 17.3 6.6 77.0 16.9 6.1 79.6 14.9 5.6 80.0 13.4 6.6 79.8 14.2 6.1 *Missing data ≤2.1% ¥Chronbach’s alpha for confidence scale for all Grade-8 girls= 0.57. **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. 64 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa .Table A5b. Confidence¥ and expectations about the future, by province and intervention group, Grade 8 boys* KwaZulu-Natal (n=1,435)** Mpumalanga (n=1,353)** Intervention Control Total Intervention Control Total Respondent indicated that the following is true most or all of the time (reference category is none or some of the time) When I have a problem, I can come up with lots of ways to solve it. 43.3 40.2 42.0 46.8 45.1 46.0 The things I have done in the past will help me in the future. 54.0 54.0 54.0 52.8 56.8 54.8 I believe I can find ways to solve a problem even when others want to quit. 51.8 49.6 50.9 53.1 55.2 54.1 Thinks life will be better, about the same, or worse 1 year from now Better Same Worse 74.1 20.4 5.4 75.4 20.7 3.9 74.7 20.5 4.8 77.2 18.3 4.5 80.5 15.4 4.1 78.8 16.9 4.3 Thinks life will be better, the same, or worse 5 years from now Better Same Worse 74.7 19.3 6.0 75.1 20.0 4.9 74.9 19.6 5.5 72.0 20.8 7.2 75.1 18.7 6.2 73.6 19.8 6.7 *Missing data i≤1.7% ¥Chronbach’s alpha for confidence scale for all Grade-8 boys=0.40. **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 65 Table A5c. Confidence¥ and expectations about the future, by province and intervention group, Grade-10 girls* KwaZulu-Natal (n=2,008)** Mpumalanga (n=1,731)** Intervention Control Total Intervention Control Total Respondent indicated that the following is true most or all of the time (reference category is none or some of the time) When I have a problem, I can come up with lots of ways to solve it. 44.3 43.7 44.0 50.3 49.0 49.6 The things I have done in the past will help me in the future. 56.5 53.1 54.9 52.5 54.5 53.5 I believe I can find ways to solve a problem even when others want to quit. 53.9 52.7 53.4 58.0 57.1 57.5 Thinks life will be better, about the same, or worse 1 year from now Better Same Worse 80.4 15.7 4.9 80.3 15.2 4.6 80.3 15.4 4.3 87.6 9.4 3.0 87.4 9.3 3.3 87.5 9.4 3.2 Thinks life will be better, the same, or worse 5 years from now Better Same Worse 84.7 11.9 3.5 83.8 12.0 4.2 84.3 11.9 3.8 83.8 11.0 5.3 86.8 8.5 4.7 85.3 9.7 5.0 *Missing data ≤0.8% ¥Cronbach’s alpha for confidence scale for all Grade-10 girls=0.30. **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. 66 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table Set A6. Gender norm attitudes Table A6a. Gender norm attitudes, by province and intervention group, Grade 8 girls* ¥ KwaZulu-Natal (n=1,988)** Mpumalanga (n=1,618)** Intervention Control Total Intervention Control Total Respondents agreed a lot, somewhat, or not at all with the following statements It is the man who decides when to have sex. Agree a lot Somewhat agree Not at all 15.7 16.8 67.5 14.7 15.6 69.8 15.3 16.3 68.5 28.5 19.7 51.9 24.2 21.4 54.4 26.4 20.5 53.1 Men are always ready to have sex. Agree a lot Somewhat agree Not at all 32.4 19.1 48.5 33.6 19.3 47.1 32.9 19.2 47.9 49.6 25.7 24.8 46.9 24.2 28.9 48.2 25.0 26.8 Women are always ready to have sex. Agree a lot Somewhat agree Not at all 7.0 25.7 67.3 6.5 22.0 71.5 6.8 24.1 69.1 9.5 31.0 59.5 8.6 30.0 61.4 9.1 30.5 60.4 Men need sex more than women do. Agree a lot Somewhat agree Not at all 34.3 18.7 47.0 36.1 19.7 44.2 35.1 19.1 45.8 54.8 20.9 24.3 53.9 21.3 24.9 54.3 21.1 24.6 A man needs other women even if things with his wife/partner are fine. Agree a lot Somewhat agree Not at all 28.9 22.4 48.7 29.1 19.0 51.9 29.0 21.0 50.0 43.1 23.4 33.5 45.7 22.2 32.1 44.4 22.8 32.8 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 67 You don’t talk about sex, you just do it. Agree a lot Somewhat agree Not at all 9.1 13.0 77.9 7.9 11.9 80.2 8.6 12.6 78.8 12.3 12.4 75.3 11.9 14.7 73.5 12.1 13.5 74.4 A woman should not initiate sex. Agree a lot Somewhat agree Not at all 18.4 24.2 57.5 19.2 23.1 57.7 18.7 23.7 57.6 25.0 33.3 41.7 24.7 33.4 42.0 24.8 33.3 41.9 A woman who has sex before she is married does not deserve respect. Agree a lot Somewhat agree Not at all 23.6 20.4 56.0 25.4 17.0 57.6 24.4 19.0 56.7 33.4 20.3 46.4 31.5 20.6 47.9 32.5 20.5 47.1 A man who has sex before he is married does not deserve respect. Agree a lot Somewhat agree Not at all 21.8 20.9 57.3 23.9 20.1 56.0 22.7 20.6 56.8 31.0 21.0 48.0 29.4 22.9 47.7 30.2 21.9 47.9 Women who carry condoms on them are loose. Agree a lot Somewhat agree Not at all 23.0 20.3 56.6 25.7 20.7 53.6 24.2 20.5 55.4 30.3 21.3 48.3 28.5 21.2 50.4 29.4 21.3 49.3 In my opinion, women can suggest using condoms just like a man. Agree a lot Somewhat agree Not at all 41.8 20.8 37.4 40.5 20.3 39.3 41.3 20.6 38.2 51.7 25.0 23.3 49.0 21.8 29.2 50.4 23.4 26.2 68 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa A couple should decide together if they want to have children. Agree a lot Somewhat agree Not at all 63.0 14.8 22.2 64.2 13.7 22.1 63.5 14.3 22.1 80.3 10.2 9.5 77.8 10.8 11.4 79.1 10.5 10.5 It is only the woman’s responsibility to avoid getting pregnant. Agree a lot Somewhat agree Not at all 41.1 18.0 40.9 38.6 22.2 39.2 40.0 19.8 40.2 56.7 17.6 25.8 54.3 18.8 26.9 55.5 18.2 26.3 If a man gets a woman pregnant, a child is the responsibility of both. Agree a lot Somewhat agree Not at all 61.6 14.4 24.1 64.0 14.0 22.0 62.6 14.2 23.2 68.9 13.2 18.0 65.9 16.0 18.2 67.4 14.6 18.1 It is important that a father is present in the lives of his children, even if he is no longer with the mother. Agree a lot Somewhat agree Not at all 64.8 15.6 19.5 69.2 13.4 17.4 66.7 14.7 18.6 74.7 13.2 12.1 71.0 14.9 14.1 72.8 14.1 13.1 *Missing data ≤4.6% data. ¥Cronbach’s alpha for gender norms attitudes scale for all Grade-8 girls=0.81. **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 69 Table A6b. Gender norm attitudes, by province and intervention group, Grade-8 boys* ¥ KwaZulu-Natal (n=1,435)** Mpumalanga (n=1,353)** Intervention Control Total Intervention Control Total Respondents agreed a lot, somewhat, or not at all with the following statements It is the man who decides when to have sex. Agree a lot Somewhat agree Not at all 28.7 27.1 44.1 29.9 29.7 40.4 29.2 28.2 42.6 34.0 29.2 36.8 34.0 30.6 35.5 34.0 29.9 36.1 Men are always ready to have sex. Agree a lot Somewhat agree Not at all 37.7 29.2 33.1 37.2 31.6 31.3 37.4 30.3 32.3 46.1 30.6 23.3 46.5 27.4 26.2 46.3 29.0 24.7 Women are always ready to have sex. Agree a lot Somewhat agree Not at all 19.5 35.2 45.3 19.9 36.4 43.6 19.7 35.7 44.6 24.9 40.2 34.9 24.7 41.5 33.8 24.8 40.9 34.4 Men need sex more than women do. Agree a lot Somewhat agree Not at all 40.1 26.1 33.8 42.9 23.8 33.3 41.3 25.1 33.6 45.1 28.3 26.6 47.3 25.8 26.9 46.2 27.1 26.7 A man needs other women even if things with his wife/partner are fine. Agree a lot Somewhat agree Not at all 29.6 31.2 39.1 30.1 31.7 38.2 29.8 31.4 38.8 41.0 26.9 32.2 38.5 27.9 33.6 39.8 27.4 32.9 70 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa You don’t talk about sex, you just do it. Agree a lot Somewhat agree Not at all 18.7 25.5 55.8 17.5 24.7 57.8 18.2 25.2 56.7 26.8 22.3 51.0 24.1 23.8 52.1 25.4 23.1 51.5 A woman should not initiate sex. Agree a lot Somewhat agree Not at all 19.6 35.2 45.2 22.2 35.8 42.0 20.7 35.5 43.8 23.7 42.8 33.5 24.9 41.5 33.6 24.3 42.2 33.5 A woman who has sex before she is married does not deserve respect. Agree a lot Somewhat agree Not at all 30.6 24.2 45.3 28.9 25.9 45.2 29.8 24.9 45.2 33.1 24.8 42.1 34.0 21.4 44.6 33.5 23.1 43.3 A man who has sex before he is married does not deserve respect. Agree a lot Somewhat agree Not at all 26.0 26.2 47.8 25.2 25.2 49.7 25.6 25.8 48.6 28.4 27.2 44.4 29.0 25.6 45.4 28.7 26.4 44.9 Women who carry condoms on them are loose. Agree a lot Somewhat agree Not at all 33.1 22.0 44.9 31.8 23.0 45.1 32.6 22.5 45.0 32.8 28.4 38.8 34.4 23.6 42.1 33.6 26.0 40.4 In my opinion, women can suggest using condoms just like a man. Agree a lot Somewhat agree Not at all 51.3 23.8 25.0 48.2 26.2 25.6 49.9 24.8 25.2 51.2 27.8 21.0 55.8 24.4 19.8 53.5 26.1 20.4 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 71 A couple should decide together if they want to have children. Agree a lot Somewhat agree Not at all 64.5 17.9 17.7 63.7 19.6 16.8 64.1 18.6 17.3 72.2 15.2 12.5 68.8 17.6 13.7 70.5 16.4 13.1 It is only the woman’s responsibility to avoid getting pregnant. Agree a lot Somewhat agree Not at all 39.9 24.9 35.2 37.3 24.8 37.9 38.8 24.9 36.3 48.8 25.1 26.1 50.1 25.6 24.3 49.5 25.3 25.2 If a man gets a woman pregnant, a child is the responsibility of both. Agree a lot Somewhat agree Not at all 60.6 20.0 19.4 63.0 18.2 18.8 61.6 19.3 19.1 59.4 20.0 20.6 55.8 22.2 22.1 57.6 21.1 21.3 It is important that a father is present in the lives of his children, even if he is no longer with the mother. Agree a lot Somewhat agree Not at all 63.3 21.4 15.4 62.4 23.4 14.2 62.9 22.2 14.9 62.1 22.9 15.0 62.6 22.1 15.3 62.4 22.5 15.2 *Missing data ≤4.0% ¥Cronbach’s alpha for gender norms attitude scale for all Grade-8 boys=0.79. **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. 72 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table A6c. Gender norm attitudes, by province and intervention group, Grade-10 girls* ¥ KwaZulu-Natal (n=2,008)** Mpumalanga (n=1,731)** Intervention Control Total Intervention Control Total Respondents agreed a lot, somewhat, or not at all with the following statements It is the man who decides when to have sex. Agree a lot Somewhat agree Not at all 17.7 22.5 59.8 16.5 21.6 62.0 17.1 22.1 60.8 26.3 24.6 49.2 24.7 23.7 51.6 25.5 24.2 50.4 Men are always ready to have sex. Agree a lot Somewhat agree Not at all 42.9 25.6 31.5 39.2 26.2 34.5 41.2 25.9 32.9 47.5 24.3 28.2 49.2 24.9 25.9 48.3 24.6 27.0 Women are always ready to have sex. Agree a lot Somewhat agree Not at all 5.8 32.1 62.1 6.7 29.9 63.4 6.2 31.1 62.7 8.3 34.5 57.2 6.7 34.9 58.4 7.5 34.7 57.8 Men need sex more than women do. Agree a lot Somewhat agree Not at all 49.2 22.7 28.1 44.4 20.9 34.7 47.0 21.9 31.1 54.0 21.2 24.8 56.7 19.4 23.9 55.3 20.3 24.4 A man needs other women even if things with his wife/partner are fine. Agree a lot Somewhat agree Not at all 40.2 24.4 35.4 37.7 24.5 37.8 39.1 24.4 36.5 50.4 20.2 29.4 48.7 17.9 33.4 49.6 19.0 31.4 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 73 You don’t talk about sex, you just do it. Agree a lot Somewhat agree Not at all 9.7 13.8 76.5 9.3 15.5 75.2 9.5 14.6 75.9 12.5 20.0 67.5 9.5 18.7 71.8 11.0 19.3 69.6 A woman should not initiate sex. Agree a lot Somewhat agree Not at all 18.5 33.1 48.4 21.0 29.5 49.6 19.6 31.4 48.9 22.4 36.5 41.1 23.2 33.2 43.5 22.8 34.9 42.3 A woman who has sex before she is married does not deserve respect. Agree a lot Somewhat agree Not at all 17.9 20.4 61.7 21.3 15.7 63.0 19.5 18.3 62.3 23.4 17.2 59.4 18.4 16.4 65.2 20.9 16.8 62.3 A man who has sex before he is married does not deserve respect. Agree a lot Somewhat agree Not at all 14.5 20.4 65.1 19.5 17.2 63.3 16.8 18.9 64.3 23.8 17.9 58.4 17.6 16.2 66.1 20.7 17.0 62.2 Women who carry condoms on them are loose. Agree a lot Somewhat agree Not at all 20.7 20.7 58.6 20.8 17.0 62.2 20.8 19.0 60.2 23.7 18.9 57.4 24.0 18.0 58.0 23.9 18.4 57.7 In my opinion, women can suggest using condoms just like a man. Agree a lot Somewhat agree Not at all 57.9 20.0 22.0 53.6 20.7 25.7 55.9 20.4 23.7 59.5 20.2 20.3 59.9 20.8 19.4 59.7 20.5 19.8 74 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa A couple should decide together if they want to have children. Agree a lot Somewhat agree Not at all 80.1 9.6 10.3 78.4 10.3 11.3 79.3 9.9 10.8 83.9 8.5 7.6 86.6 6.1 7.4 85.2 7.3 7.5 It is only the woman’s responsibility to avoid getting pregnant. Agree a lot Somewhat agree Not at all 37.5 22.9 39.6 41.1 19.1 39.8 39.1 21.2 39.7 52.9 17.0 30.1 47.5 16.2 36.3 50.2 16.6 33.2 If a man gets a woman pregnant, a child is the responsibility of both. Agree a lot Somewhat agree Not at all 78.4 10.8 10.9 78.8 11.0 10.3 78.6 10.9 10.6 74.5 12.1 13.4 78.5 9.5 12.0 76.5 10.8 12.7 It is important that a father is present in the lives of his children, even if he is no longer with the mother. Agree a lot Somewhat agree Not at all 79.6 11.1 9.3 79.3 11.5 9.2 79.5 11.3 9.3 77.4 13.4 9.2 82.3 9.0 8.6 79.9 11.2 8.9 *Missing data is ≤2.3% ¥Cronbach’s alpha for gender norms attitude scale=0.73 **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 75 Table Set A7. Risk perception Table A7a. Risk perception, by province and intervention group, Grade-8 girls* KwaZulu-Natal (n=1,988)** Mpumalanga (n=1,618)** Intervention Control Total Intervention Control Total Respondent's perception of the chance that each of the following will happen to him or her The chance of getting HIV in his or her lifetime No chance Some chance High chance 63.7 21.4 15.0 66.8 21.1 12.1 65.0 21.3 13.7 59.9 26.0 14.1 56.4 31.0 12.7 58.2 28.5 13.4 The chance that he or she already has HIV No chance Some chance High chance 73.3 15.5 11.2 74.3 14.5 11.2 73.7 15.1 11.2 67.0 21.9 11.1 72.0 19.1 8.9 69.5 20.5 10.1 * Missing data ≤5.2% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. 76 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table A7b. Risk perception, by province and intervention group, Grade-8 boys* KwaZulu-Natal (n=1,435)** Mpumalanga (n=1,353)** Intervention Control Total Intervention Control Total Respondent's perception of the chance that each of the following will happen to him or her The chance of getting HIV in his or her lifetime No chance Some chance High chance 61.2 26.7 12.1 59.9 29.2 11.0 60.6 27.8 11.6 54.9 31.0 14.1 55.1 29.8 15.1 55.0 30.4 14.6 The chance that he or she already has HIV No chance Some chance High chance 65.2 24.3 10.5 68.1 22.4 9.5 66.4 23.5 10.1 64.4 23.5 12.2 64.7 22.9 12.4 64.5 23.2 12.3 * Missing data ≤4.6% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 77 Table A7c. Risk perception, by province and intervention group, Grade-10 girls* KwaZulu-Natal (n=2,008)** Mpumalanga (n=1,731)** Intervention Control Total Intervention Control Total Respondent's perception of the chance that each of the following will happen to him or her The chance of getting HIV in his or her lifetime No chance Some chance High chance 59.6 31.8 8.6 60.6 30.3 9.1 60.1 31.1 8.8 56.5 29.7 13.8 58.1 28.0 13.9 57.3 28.9 13.9 The chance that he or she already has HIV No chance Some chance High chance 77.0 15.9 7.2 78.3 14.7 7.0 77.6 15.3 7.1 68.4 19.9 11.7 75.4 16.8 7.9 71.9 18.3 9.8 *Missing data ≤1.6% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. 78 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table Set A8. Self-efficacy Table A8a. Self-efficacy, by province and intervention group, Grade-8 girls* ¥ KwaZulu-Natal (n=1,988)** Mpumalanga (n=1,618)** Intervention Control Total Intervention Control Total Respondent's reported confidence to make the following choices Could refuse to start having sex if he or she did not want to, even if friends were having sex with their boyfriends or girlfriends 57.2 57.6 57.3 50.7 54.0 52.4 Could refuse to have sex with boyfriend or girlfriend, even if b/f or g/f was pressuring him or her 63.9 63.3 63.6 58.3 60.4 59.4 Would be able to talk about condoms with b/f or g/f if b/f or g/f was asking him or her to have sex 62.4 60.4 61.6 66.2 68.0 67.1 Felt confidence he or she could refuse to have sex even if someone offered him or her a meal, gifts, money, or favor 49.3 49.6 49.4 46.2 49.2 47.7 Knows where to get tested for HIV 78.2 78.0 78.1 84.4 84.4 84.4 *Missing data ≤6.1% ¥Cronbach’s alpha of self-efficacy scale for all Grade-8 girls=0.74. **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 79 Table A8b. Self-efficacy, by province and intervention group, Grade-8 boys* ¥ KwaZulu-Natal (n=1,435)** Mpumalanga (n=1,353)** Intervention Control Total Intervention Control Total Respondent's reported confidence to make the following choices Could refuse to start having sex if he or she did not want to, even if friends were having sex with their boyfriends or girlfriends 56.8 57.1 57.0 48.2 45.8 47.0 Could refuse to have sex with boyfriend or girlfriend, even if b/f or g/f was pressuring him or her 58.0 56.8 57.5 50.7 50.1 50.4 Would be able to talk about condoms with b/f or g/f if b/f or g/f was asking him or her to have sex 68.1 71.5 69.6 70.0 70.3 70.2 Felt confidence he or she could refuse to have sex even if someone offered him or her a meal, gifts, money, or favor 50.2 54.0 51.9 46.9 44.9 45.9 Knows where to get tested for HIV 78.7 75.9 77.4 80.6 82.7 81.7 *Missing data is ≤5.1% ¥Cronbach’s alpha of self-efficacy scale for all Grade-8 boys=0.68. **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. 80 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Table A8c. Self-efficacy, by province and intervention group, Grade-10 girls* ¥ KwaZulu-Natal (n=2,008)** Mpumalanga (n=1,731)** Intervention Control Total Intervention Control Total Respondent's reported confidence to make the following choices Could refuse to start having sex if he or she did not want to, even if friends were having sex with their boyfriends or girlfriends 62.6 66.0 64.1 54.6 61.6 58.0 Could refuse to have sex with boyfriend or girlfriend, even if b/f or g/f was pressuring him or her 70.6 71.7 71.1 63.3 70.4 66.8 Would be able to talk about condoms with b/f or g/f if b/f or g/f was asking him or her to have sex 77.6 75.8 76.8 80.2 82.0 81.1 Felt confidence he or she could refuse to have sex even if someone offered him or her a meal, gifts, money, or favor 59.0 61.4 60.1 54.2 58.1 56.1 Knows where to get tested for HIV 87.9 88.0 88.0 90.8 92.2 91.5 *Missing data ≤1.8% ¥Cronbach’s alpha of self-efficacy scale for all Grade-10 girls=0.70. **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 81 Table Set A9. Sexual behavior Table A9a. Sexual behavior, by province and intervention group, Grade-8 girls* KwaZulu-Natal (n=1,988)** Mpumalanga (n=1,618)** Intervention Control Total Intervention Control Total Currently have a boyfriend or girlfriend 29.8 25.6 28.1 37.3 36.7 37.0 Age of boyfriend/girlfriend (mean) 15.9 15.8 15.9 15.6 15.5 15.6 Ever had sexual intercourse 11.4 8.5 10.2 10.3 8.9 9.6 Among those who ever had sexual intercourse KwaZulu Natal (unweighted n=183; weighted n=233.4) Mpumalanga (unweighted n=163; weighted n=65.65) Intervention Control Total Intervention Control Total Age at first sex (mean) 12.0 11.7 11.9 12.3 11.7 12.1 Number of sex partners in past 12 months (mean) 3.1 4.4 3.6 5.3 3.5 4.5 Number of different sex partners in past 3 months (mean) 4.5 2.3 3.8 3.6 3.3 3.5 Used a condom at last sex 59.4 67.0 62.0 84.5 74.3 80.3 Used a condom every time in past 3 months 49.4 65.2 54.4 79.7 62.3 73.3 Most recent or current partner attends the same school 39.9 47.0 42.5 48.7 51.3 49.8 82 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa At last sex, received money, gifts, or favors in exchange for sex 39.0 17.4 31.0 37.0 41.7 39.1 In lifetime, gave money, gifts, or favors in exchange for sex 29.4 24.8 27.7 18.4 19.4 18.9 In lifetime, received money, gifts, or favors in exchange for sex 28.0 21.2 25.4 22.9 24.4 23.5 Ever had sex with a person 5 years older 22.7 24.2 23.2 18.1 25.0 21.2 Ever did something sexual that she did not want to do Yes I’m not sure 24.8 37.4 31.6 31.5 27.2 35.4 32.6 22.9 30.8 28.1 31.8 25.2 *Missing data ≤4.7% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. At last sex, gave money, gifts, or favors in exchange for sex 34.0 25.2 30.1 35.3 29.9 32.9 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 83 Table A9b. Sexual behavior, by province and intervention group, Grade-8 boys* KwaZulu-Natal (n=1,435)** Mpumalanga (n=1,353)** Intervention Control Total Intervention Control Total Currently have a boyfriend or girlfriend 41.2 44.1 42.5 47.2 50.6 48.9 Age of boyfriend/girlfriend (mean) 14.8 15.0 14.9 14.7 14.8 14.7 Ever had sexual intercourse 23.5 23.7 23.6 19.5 19.8 19.7 Among those who ever had sexual intercourse KwaZulu-Natal (unweighted n=335; weighted n=606.6) Mpumalanga (unweighted n=286; weighted n=138.7) Intervention Control Total Intervention Control Total Age at first sex (mean) 12.0 11.6 11.8 11.7 11.3 11.5 Number of sex partners in past 12 months (mean) 4.7 3.9 4.4 5.8 5.9 5.8 Number of different sex partners in past 3 months (mean) 5.3 3.8 4.6 6.2 6.3 6.2 Used a condom at last sex 78.3 73.5 76.2 73.1 79.1 76.2 Used a condom every time in past 3 months 67.7 66.5 67.1 67.8 79.7 73.7 Most recent or current partner attends the same school 50.7 50.2 50.5 55.1 59.9 57.5 At last sex, gave money, gifts, or favors in exchange for sex 33.2 29.7 31.6 53.4 38.4 45.8 84 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa At last sex, received money, gifts, or favors in exchange for sex 31.9 34.9 33.2 42.7 42.8 42.8 In lifetime, gave money, gifts, or favors in exchange for sex 24.2 26.0 25.0 34.1 23.8 28.4 In lifetime, received money, gifts, or favors in exchange for sex 26.6 24.7 25.8 34.4 23.8 29.2 Ever had sex with a person 5 years older 34.1 31.6 33.0 26.5 31.4 28.9 Ever did something sexual that he did not want to do Yes I’m not sure 28.6 24.0 32.5 21.5 30.2 22.9 40.3 23.7 44.0 22.2 42.2 23.0 *Missing data ≤9.4% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 85 Table A9c. Sexual behavior, by province and intervention group, Grade-10 girls* KwaZulu-Natal (n=2,008)** Mpumalanga (n=1,731)** Intervention Control Total Intervention Control Total Currently have a boyfriend or girlfriend 57.8 53.6 55.9 67.9 68.1 68.0 Age of boyfriend/girlfriend (mean) 18.9 19.0 18.9 18.4 18.1 18.3 Ever had sexual intercourse 25.4 26.8 26.0 24.6 29.4 32.0 Among those who ever had sexual intercourse KwaZulu-Natal (unweighted n=521; weighted n=696.7) Mpumalanga (unweighted n=578; weighted n=313.5) Intervention Control Total Intervention Control Total Age at first sex (mean) 15.3 15.3 15.3 15.4 15.1 15.3 Number of sex partners in past 12 months (mean) 2.1 1.9 2.0 1.7 2.0 1.8 Number of different sex partners in past 3 months (mean) 2.8 2.9 2.9 3.0 2.6 2.8 Used a condom at last sex 59.0 64.8 61.8 73.5 72.2 72.8 Used a condom every time in past 3 months 58.2 54.0 56.2 62.8 72.3 67.3 Most recent or current partner attends the same school 35.4 30.1 32.9 38.3 44.1 41.1 At last sex, gave money, gifts, or favors in exchange for sex 15.5 15.4 15.4 17.5 17.3 17.4 86 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa At last sex, received money, gifts, or favors in exchange for sex 16.2 15.6 15.9 23.21 22.1 22.7 In lifetime, gave money, gifts, or favors in exchange for sex 14.2 12.5 13.4 11.9 15.7 13.7 In lifetime, received money, gifts, or favors in exchange for sex 15.2 13.4 14.3 11.6 17.5 14.3 Ever had sex with a person 5 years older 27.4 25.9 26.7 18.1 24.2 20.9 Ever did something sexual that she did not want to do Yes I’m not sure 33.5 14.7 34.0 13.5 33.8 14.2 23.7 16.9 29.6 9.8 26.4 13.6 *Missing data ≤9.0% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 87 Table Set A10. HIV testing and counseling and other services Table A10a. HIV testing and counseling and other services, by province and intervention group, Grade-8 girls* KwaZulu-Natal (n=1,988)** Mpumalanga (n=1,618)** Intervention Control Total Intervention Control Total Visited a clinic or hospital in past 12 months to get information about sex￾related issues or get condoms or other contraceptives 35.6 32.6 34.3 28.7 30.3 29.5 Was referred for counseling or testing for STIs or HIV 33.2 33.8 33.4 23.4 27.3 25.3 Has ever been tested for HIV 40.4 44.3 42.1 33.4 37.1 35.2 Among those who have ever been tested for HIV KwaZulu-Natal (unweighted n=786; weighted n=927) Mpumalanga (unweighted n=584; weighted n=238) Intervention Control Total Intervention Control Total Received results of the HIV test 81.5 84.8 82.9 85.1 81.7 83.4 Shared the results of the HIV test with someone 65.4 64.7 65.1 61.6 70.4 66.1 Among those who were referred for counseling or testing for STIs or HIV KwaZulu-Natal (unweighted n=611; weighted n=727) Mpumalanga (unweighted n=396; weighted n=169) Intervention Control Total Intervention Control Total Went to the facility he or she was referred to 81.5 84.8 82.9 85.1 81.7 83.4 88 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Among those who were referred but did not go to referral KwaZulu-Natal (unweighted n=200; weighted n=467.1) Mpumalanga (unweighted n=134; weighted n=109.9) Intervention Control Total Intervention Control Total Reasons he or she did not Referral was too far go to referral 49.8 56.5 52.5 34.0 44.5 39.7 Referral was too costly 31.2 27.2 29.6 40.7 32.0 36.0 Worries about confidentiality 28.8 34.5 31.1 38.4 41.2 39.9 *Missing data ≤9.5% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 89 Table A10b. HIV testing and counseling and other services, by province and intervention group, Grade-8 boys* KwaZulu-Natal (n=1,435)** Mpumalanga (n=1,353)** Intervention Control Total Intervention Control Total Visited a clinic or hospital in past 12 months to get information about sex￾related issues or get condoms or other contraceptives 48.2 48.3 48.2 41.1 43.0 42.1 Was referred for counseling or testing for STIs or HIV 40.1 43.8 41.7 36.3 33.1 34.7 Has ever been tested for HIV 50.9 56.2 53.2 55.1 44.6 49.9 Among those who have ever been tested for HIV KwaZulu-Natal (unweighted n=716; weighted n=1307) Mpumalanga (unweighted n=652; weighted n=346.3) Intervention Control Total Intervention Control Total Received results of the HIV test 87.3 85.9 86.7 87.8 83.2 85.7 Shared the results of the HIV test with someone 62.7 64.9 63.7 60.0 67.6 63.3 Among those who were referred for counseling or testing for STIs or HIV KwaZulu-Natal (unweighted n=556; weighted n=1010) Mpumalanga (unweighted n=463; weighted n=238.6) Intervention Control Total Intervention Control Total Went to the facility he or she was referred to 65.2 62.7 64.1 67.4 71.9 69.6 90 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Among those who were referred but did not go to referral KwaZulu-Natal (unweighted n=200; weighted n=636.4) Mpumalanga (unweighted n=139; weighted n=162.4) Intervention Control Total Intervention Control Total Reasons he or she did not go to referral Referral was too far 48.6 53.4 50.8 50.1 52.9 51.3 Referral was too costly 30.6 37.3 33.7 33.3 44.2 38.1 Worries about confidentiality 27.1 21.5 24.5 34.1 29.8 32.2 *Missing data is ≤10.2% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 91 Table A10c. HIV testing and counseling and other services, by province and intervention group, Grade-10 girls* KwaZulu-Natal (n=2,008)** Mpumalanga (n=1,731)** Intervention Control Total Intervention Control Total Visited a clinic or hospital in past 12 months to get information about sex￾related issues or get condoms or other contraceptives 40.7 39.3 40.1 44.4 36.7 40.5 Was referred for counseling or testing for STIs or HIV 35.3 39.0 37.0 30.2 26.5 28.4 Has ever been tested for HIV 55.6 56.7 56.1 54.3 46.0 50.2 Among those who have ever been tested for HIV KwaZulu-Natal (unweighted n=1,111; weighted n=1481) Mpumalanga (unweighted n=845; weighted n=487) Intervention Control Total Intervention Control Total Received results of the HIV test 95.0 91.4 93.3 93.2 88.8 91.2 Shared the results of the HIV test with someone 73.1 72.3 72.7 71.4 72.8 72.0 Among those who were referred for counseling or testing for STIs or HIV KwaZulu-Natal (unweighted n=725; weighted n=968.7) Mpumalanga (unweighted n=476; weighted n=272.4) Intervention Control Total Intervention Control Total Went to the facility he or she was referred to 68.9 67.3 68.1 73.8 72.0 72.9 92 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Among those who were referred but did not go to referral KwaZulu-Natal (unweighted n=232; weighted n=655.8) Mpumalanga (unweighted n=139; weighted n=195.4) Intervention Control Total Intervention Control Total Reasons he or she did not go to referral Referral was too far 59.6 44.0 52.1 38.4 67.3 52.0 Referral was too costly 24.2 36.7 30.2 26.8 25.7 26.3 Worries about confidentiality 22.4 29.0 25.6 45.4 17.0 32.0 *Missing data is ≤5.3% **Unweighted Ns shown here. Calculations based on weighted Ns. See Table 3 for weighted Ns. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 93 APPENDIX 2. ADDITIONAL FINDINGS The analyses and figures presented below were generated after in-country dissemination meetings to explore specific areas of interest in greater depth. Each figure reports information about a subsample of youth who were at or within two years of the typical age for their grade. This eliminated from our study learners who were two or more years older than the typical age for their grades: 72 Grade-8 boys (2.6%), 59 Grade-8 girls (1.7%), and 142 Grade-10 girls (3.8%). We dropped these older learners because they would have been exposed to the outcomes of interest (e.g., sexual experience, pregnancy, and HIV testing) for a longer period than their peers, thus biasing the results. To examine characteristics of young people who reside in food-insecure households, we used the variable, how many of the past three days the household went without food. We considered any days without food to be “food-insecure” and no days without food to be “not food-insecure.” We coded orphanhood based on whether the learner’s biological mother or father was alive. Thus, we coded young people who reported either that their mother or father was not alive or that they did not know if their mother or father was alive as orphans, because they lacked access to a parent. All values are weighted and significance testing was undertaken using an F-test; p-values are shown for significant differences. All figures are for both provinces combined. Figure A.1. Percentage of learners who went any of the past three days without food in their home (food-insecure), by orphanhood status **p≤0.01; ***p≤0.001 Summary: Orphans (single and double) were significantly more likely to reside in food-insecure households than nonorphans. Double orphans came from the most food-insecure households. The same results were seen for all grade groups. In total, 33.8 percent of Grade-8 boys, 29.5 percent of Grade-8 girls, and 31.1 percent of Grade-10 girls reported that their homes had gone without food for any of the past three days. 29.3 37.4 45.6 24.7 33.7 41.1 27.9 32.7 39.5 Not orphan Single orphan Double orphan Not orphan Single orphan Double orphan Not orphan Single orphan Double orphan Grade-8 boys*** Grade-8 girls*** Grade-10 girls** 94 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Figure A.2. Percentage of learners who missed school, by grade and orphan status Summary: Among Grade-8 boy and girl learners, double orphans were more likely than other learners to have missed three or more days of school (p≤0.001). Among Grade-10 girls, single and double orphans were more likely than nonorphans to have missed three or more days of school (p<0.05). Figure A.3. Percentage of learners who had ever had sex, by grade and orphan status +p≤0.10 Summary: Double orphans were more likely than other learners to have ever had sex, although this was only borderline significant among Grade-10 girls (p≤0.10). 22.1 21.3 24.2 9.3 9.6 10.6 24.5 26.7 30.5 Not orphan Single orphan Double orphan Not orphan Single orphan Double orphan Not orphan Single orphan Double orphan Grade-8 boys Grade-8 girls Grade-10 girls+ 17.9 15 13.2 17.9 17.1 16.6 27.6 25.2 15.8 Grade-8 boys*** Grade-8 girls*** Grade-10 girls* Not orphans Single orphans Double orphans Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 95 Figure A.4. Percentage of learners who had ever had sex, by grade and by household food insecurity (i.e., went any days without food in the past 3 days vs. had food all those days) ***p≤0.001 Summary: Learners who resided in food-insecure households were significantly more likely to have ever had sex than the learners who did not reside in food-insecure households. Figure A.5. Percentage of learners who had ever been pregnant (or gotten a partner pregnant) by grade and by household food insecurity for a) all learners; and b) learners who had ever had sex* *These comparisons are all significant at p≤0.01. 18.3 29.0 7.5 14.2 23.3 32.1 No days without food Any days without food No days without food Any days without food No days without food Any days without food Grade-8 boys*** Grade-8 girls*** Grade-10 girls*** 2.0 6.9 1.0 3.9 4.2 8.5 12.4 24.9 13.9 29.2 18.2 27.5 No days without food Any days without food No days without food Any days without food No days without food Any days without food Grade-8 boys Grade-8 girls Grade-10 girls All learners Learners who ever had sex 96 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Summary: Learners who resided in food-insecure households were more likely to have ever been pregnant (or to have gotten a partner pregnant) than learners who did not reside in food-insecure households. Among learners who had ever had sex, a large proportion of those who resided in food￾insecure household had ever been pregnant or gotten a partner pregnant (about one quarter). Figure A.6. Percentage of learners who had ever been tested for HIV, by sexual experience and grade* *These comparisons are significant at p≤0.001. Summary: A high percentage of learners had ever been tested, and the percentage was higher among learners who had had sex than among those who had never had sex. Figure A.7. Learners’ perceived chance (no chance, some chance, high chance) of acquiring HIV, by sexual experience and grade (%)* *p≤0.001 for all comparisons Summary: Learners who had ever had sex were significantly more likely to perceive some chance or a high chance of acquiring HIV in their lifetime than learners who had never had sex. 60.8 56.3 64.8 51.5 62.9 51.9 27.8 29.0 21.9 33.8 28.0 36.2 11.3 14.7 13.3 14.7 9.2 11.9 Never sex Ever sex Never sex Ever sex Never sex Ever sex Grade-8 boys Grade-8 girls Grade-10 girls No chance Some chance High chance 49.8 59.3 38.8 52.6 49.2 65.1 Never sex Ever sex Never sex Ever sex Never sex Ever sex Grade-8 boys Grade-8 girls Grade-10 girls Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 97 Figure A.8. Learners’ perceived chance (no chance, some chance, high chance) of getting HIV, by household food insecurity and grade (%)* * p≤0.05 for all comparisons. Summary: Learners who had gone any days without food perceived a higher risk of acquiring HIV in their lifetime than those who had had food in the past three days. Figure A.9. Learners’ perceived chance (no chance, some chance, high chance) of getting pregnant or getting a partner pregnant while in high school, by sexual experience and grade (%)* *p≤0.05 for boys and p≤0.001 for all girl comparisons. Summary: Learners who had ever had sex (particularly females) perceived a greater chance of getting pregnant or getting a partner pregnant while still in high school than learners who had never had sex. 62.8 54.0 65.1 60.6 62.0 56.0 26.2 31.5 22.6 23.7 29.2 31.9 11.0 14.5 12.4 15.7 8.8 12.1 No days without food Any days without food No days without food Any days without food No days without food Any days without food Grade-8 boys Grade-8 girls Grade-10 girls No chance Some chance High chance 53.7 45.8 62.8 41.5 64.5 52.1 23.9 32.1 14.1 30.0 13.0 24.5 22.4 22.1 23.1 28.5 22.5 23.4 Never sex Ever sex Never sex Ever sex Never sex Ever sex Grade-8 boys Grade-8 girls Grade-10 girls No chance Some chance High chance 98 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Figure A.10. Learners’ perceived chance of getting pregnant or getting a partner pregnant while in high school, by household food insecurity and grade (%)* *p≤0.01 for boys and p≤0.001 for all girl comparisons. Summary: Learners who had gone without food any of the past three days perceived a higher risk of getting pregnant or getting a partner pregnant while in high school than learners who had not gone without food. 55.4 45.5 64.7 52.2 63.2 58.4 22.8 31.1 13.6 18.7 15.6 15.9 21.8 23.4 21.7 29.1 21.3 25.7 No days without food Any days without food No days without food Any days without food No days without food Any days without food Grade-8 boys Grade-8 girls Grade-10 girls No chance Some chance High chance Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 99 APPENDIX 3. BASELINE QUESTIONNAIRE Baseline Questionnaire for Grades 8 and 10 Learners Province District School Student Number Name of Life Orientation Teacher Class Code (facilitator will tell you the class code to select) Date (dd/mm/yy) Time A. Demographics and HH Composition Tell us a bit about yourself and your household Questions Response Options Codes A1 How old were you on your last birthday? [Respondent enters age] A2 In what school grade are you currently enrolled? Grade 8 Grade 10 1 2 A3 Are you female or male? Female Male 2 1 A4 What race group are you? Black Indian White Coloured Other 1 2 3 4 5 A5 What is your religion? Christian Jewish Muslim Hindu Ba’hai Traditional Other 1 2 3 4 5 6 7 A6 Aside from wedding and funerals, how often do you attend religious services/church? Usually once a week Usually once or twice a month Usually a few times a year Seldom Never 1 2 3 4 5 A7 Who is your primary care giver? This could be your mother, father, granny or someone else. We call this person your primary caregiver. Mother Father Older brother or sister Aunts and/or uncles 1 2 3 4 100 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Grandparents Cousins Other adults who are not blood relatives Another child I am the head of the house 5 6 7 8 9 A8 What is the highest level of education your primary caregiver completed? Did not go to school Attended some primary, but did not complete primary school Completed primary school Attended some secondary, but did not complete secondary school (grade 12) Completed secondary school (grade 12) Completed a diploma Completed a university or technikon degree Don’t know 1 2 3 4 5 6 7 8 A9 Is your biological mother still alive? Yes No Don’t know 1 2 8 A10 Is your biological father still alive? Yes No Don’t know 1 2 8 A11 Do you live with your biological mother? ASKED ONLY IF “Yes” to question A9 Yes No 1 2 A12 Do you live with your biological father? ASKED ONLY IF “Yes” to question A10 Yes No 1 2 A13. Does your household have any of these things? A13a Electricity Yes No 1 2 A13b Radio Yes No 1 2 A13c Tap water Yes No 1 2 A13d Television Yes No 1 2 A13e Fridge Yes No 1 2 A13f Motor car Yes No 1 2 A14 What kind of toilet does your house have? Flushing toilet Pit toilet/long drop No toilet facility 1 2 3 A15 What kind of floor does your house have? CHOOSE ONLY 1 ANSWER THAT APPLIES TO THE MAIN AREA OF YOUR HOUSE Earth or Sand or Dung Bare wood planks Parquet or polished wood Vinyl or asphalt strips Ceramic tiles Cement Carpet 1 2 3 4 5 6 7 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 101 Don’t know 8 A16 What is the main material used for the outside walls of your house? Bricks Cement block/concrete Corrugated iron/zinc Wood Plastic Cardboard Mud or cement mix Wattle and daub Tile Mud Thatching/grass Asbestos Don’t know 1 2 3 4 5 6 7 8 9 10 11 12 98 A17 In the past 3 days, how many days have you been without food in your home? 0 days 1 day 2 days 3 days 1 2 3 4 A18 How long have you lived in your house? Less than 12 months 12 months or more → SKIP TO A21 1 2 A19 The last time you moved, where did you move from? From a house in the same village or town From a different village or town From a different Province From outside South Africa 1 2 3 4 A20 How many times have you moved to a different village or town in the past 12 months? None Once 2-3 times 4-5 times 6 or more times 0 1 2 3 4 A21 Is someone in your household HIV positive? Yes No Don’t know 1 2 8 B. Relationship to Caregivers Now please tell us about your relationship to your caregivers, whether that is your parents, other family members, or someone else Questions Response Options Codes B1 If I am going to be home late I am expected to tell my parent/caregiver to let them know where I am Never Rarely Sometimes Often Very often 1 2 3 4 5 B2 My parent/caregiver knows who my friends are Never Rarely Sometimes Often Very often 1 2 3 4 5 B3 My parent/caregiver comes to events at my school like parent meetings, prize giving, sports days, etc. Never Rarely Sometimes Often 1 2 3 4 102 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Very often 5 B4 My parent/caregiver asks me what I learned in school Never Rarely Sometimes Often Very often 1 2 3 4 5 B5 If I have a problem I can talk to my parent/caregiver Never Rarely Sometimes Often Very often 1 2 3 4 5 C. School Attendance, Performance, and Relationships Please tell us about your schooling Questions Response Options Codes C1 Of the last 10 school days, how many did you miss? None 1-2 days 3-4 days 5-6 days 7 or more days 1 2 3 4 5 C2 Last month how many days were you late to school? None 1-2 days 3-4 days 5-6 days 7 or more days 1 2 3 4 5 C3 Which month of this school year did you miss the most days of school when school was open? I have not missed any days of school → SKIP TO C5 January February March April May June July August September October November 0 1 2 3 4 5 6 7 8 9 10 11 C4 How many days of school did you miss in that month? None 1-2 days 3-4 days 5-6 days 7 or more days 1 2 3 4 5 C5 What was your overall mark across all the subjects for the last term? ≤20% 20-29% 30-39% 40-49% 50-59% 60-69% 70-79% ≥80% 1 2 3 4 5 6 7 8 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 103 C6 Do you feel emotionally close to other classmates at your school? By emotionally close, we mean that you consider classmates to be your friends and like to spend time with them. Yes No 1 2 C7 Do you feel like you belong in your school? Yes No 1 2 C8 During the last full school year, were you happy to be at your school? Yes No 1 2 C9 During the last full school year, did you feel that the teachers at your school treated learners fairly? Yes No 1 2 C10 In your opinion, what is the chance that you will be promoted to the next grade? No chance Some chance A high chance 1 2 3 C11 In your opinion, what is the chance that you will still be in school next year? No chance Some chance A high chance 1 2 3 D. Aspirations & Expectations about the Future How often are the following statements true? Questions Response Options Codes D1 When I have a problem, I can come up with ways to solve it Not at all Sometimes Mostly Always 1 2 3 4 D2 The things I have done in the past will help me in the future Not at all Sometimes Mostly Always 1 2 3 4 D3 I believe I can find ways to solve a problem even when others want to quit Not at all Sometimes Mostly Always 1 2 3 4 Tell us what you think about your future. Questions Response Options Codes D4 Do you think your life will be better, about the same, or worse 1 year from now? Better Same Worse 1 2 3 D5 Do you think your life will be better, about the same, or worse 5 years from now? Better Same Worse 1 2 3 E. Participation and Perceptions of Life Orientation Curriculum Please tell us about the Life Orientation (LO) Curriculum. How true are the following statements? Questions Response Options Codes E1 The things we learn about gender roles, sexuality, and HIV in the Life Orientation Not true A little true Mostly true 1 2 3 104 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa class are similar to what I experience in my life Very true 4 E2 I have learned a lot about sexuality and HIV related topics in my Life Orientation class Not true A little true Mostly true Very true 1 2 3 4 E3 I am able to apply some of the things I have learned about gender roles, sexuality, and HIV in the Life Orientation class to my personal life Not true A little true Mostly true Very true 1 2 3 4 E4 I talk to my parent/caregiver about the sexual and HIV related topics I learn in the Life Orientation class Not true A little true Mostly true Very true 1 2 3 4 E5 My parents/caregiver think it is a good thing I am learning about HIV and AIDS in school Not true A little true Mostly true Very true 1 2 3 4 E6 How would you rate your level of participation in class discussions in Life Orientation lessons? I don’t really participate I participate a little I participate a lot We don’t have class discussions 1 2 3 4 E7 How often do you ask the teacher questions during your LO lessons? I don’t really ask questions I ask questions sometimes I ask questions often 1 2 3 E8 To what extent do you listen to what the teacher is teaching you in LO lessons? I don’t really listen to my teacher I listen a little to what my teacher teaches in LO I listen most of the time to what my teacher teaches in LO 1 2 3 E9 How motivated are you to learn about LO? I am not motivated to learn LO I am a little motivated to learn about LO I am very motivated to learn about LO 1 2 3 E10 How much time and effort do you put into doing assignments or studying for LO tests? I don’t put in much time or effort I put in some time and effort I put in a lot of time and effort 1 2 3 How true are the following statements? E11 My LO teacher encourages the students to learn the material. Not true A little true Mostly true Very true 1 2 3 4 E12 My LO teacher understands the material that he/she presents Not true A little true Mostly true Very true 1 2 3 4 E13 My LO teacher wants all students to feel respected Not true A little true Mostly true Very true 1 2 3 4 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 105 F. Knowledge of STIs/HIV Now please tell us what you know about HIV/AIDS and other sexually transmitted infections (STIs) Questions Response Options Codes F1 You can usually tell if someone has HIV and AIDS by the way they look True False Don’t know 1 2 8 F2 If you have a sexually transmitted infection (STI) you will definitely know because you will see/feel symptoms True False Don’t know 1 2 8 F3 Not all sexually transmitted infections are curable True False Don’t know 1 2 8 F4 Oral sex has no risk for STIs True False Don’t know 1 2 8 F5 When used correctly and consistently, condoms protect you from all STIs True False Don’t know 1 2 8 F6 If a mosquito bites you it can infect you with HIV True False Don’t know 1 2 8 F7 You can get HIV from kissing a person who is HIV positive True False Don’t know 1 2 8 F8 A woman who is pregnant can do nothing to prevent her baby from being born with HIV True False Don’t know 1 2 8 F9 In the past 6 months, where did you hear or see information on HIV and AIDS? Tick all that apply Radio commercials Radio shows/programs T.V. commercials T.V. shows Cinema/Movies Cell phone messages Social media like Facebook, WhatsApp or Twitter School clubs Religious gatherings Street performances Friends Family members Library Sporting events Don’t remember I did not hear or see information on HIV and AIDS in the last 6 months A b c d e f g h i j k l m n y z Tell us from where you most learn about sex and HIV F10 At your school, where do you learn the most about sex, sexuality, and HIV? The Life Orientation educator Other educators at the school External people that visit the school (like NGOs, nurses, doctors, or social workers) 1 2 3 106 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa The school library books or computers My friends and other learners at school Another source (specify) 4 5 6 G. Sexual Behavior We would like to ask you some questions about your sexual activity. Try to answer as honestly as you can. None of your answers will be shared with anyone. These questions are not meant to make you feel uncomfortable. If there is a question that you really do not feel comfortable asking, you can skip that question. Question Response Options Code G1 Do you currently have a boyfriend or girlfriend? A boyfriend or girlfriend is someone you have a romantic relationship with. Yes No → SKIP TO G3 1 2 G2 How old is your boyfriend or girlfriend? [Respondent enters age] G3 Have you ever had sexual intercourse (sex) before? By ‘sex’ we mean ‘going all the way’ or vaginal or anal penetration. Yes No → STOP THIS SECTION. SKIP TO H1 1 2 G4 How old were you when you had sex for the first time? [Respondent enters age] G5 In the past 12 months how many different people have you had sex with? Respondent enters number IF RESPONDENT ENTERS 0, THEN SKIP TO G13 G6 Did you or your most recent (or current) sex partner use a condom the last time you had sex? A sex partner is someone you have had sex with. Yes No 1 2 G7 Did (Does) your most recent (or current) sex partner attend the same school as you? Yes No 1 2 G8 The last time you had sex, did you give money, gifts, or favours (for example, cell phone, cash, or clothes?) in exchange for sex? Yes No 1 2 G9 The last time you had sex, did you receive money, gifts, or favours (for example, cell phone, cash, or clothes?) in exchange for sex? Yes No 1 2 G10 In the past 3 months, how many different people have you had sex with? [Respondent enters number] IF RESPONDENT ENTERS 0, THEN SKIP TO G13 G11 How many times have you had sex in the last 3 months? [Respondent enters number] Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 107 G12 Did you use a condom every time you had sex in the last 3 months? Yes No Don’t know 1 2 8 G13 In your life, have you ever given money, gifts, or favours in exchange for sex? ASKED ONLY OF RESPONDENTS WHO ANSWER NO TO G8 Yes No 1 2 G14 In your life, have you ever received money, gifts, or favours in exchange for sex? ASKED ONLY OF RESPONDENTS WHO ANSWER NO TO G9 Yes No 1 2 G15 In your life, have you ever had sex with a person who was more than 5 years older than you? Yes No Don’t know 1 2 8 G16 Have you ever done something sexual with someone else that you wish you had not done? Yes No I’m not sure 1 2 3 H. Gender norm attitudes (from GEM scale) We would like to ask you about your opinion about men’s and women’s roles. For each statement tell us if you agree a lot, somewhat agree, or do not agree at all. Question Response Options Code H1 It is the man who decides when to have sex Agree a lot Somewhat agree Do not agree at all 1 2 3 H2 Men are always ready to have sex Agree a lot Somewhat agree Do not agree at all 1 2 3 H3 Women are always ready to have sex Agree a lot Somewhat agree Do not agree at all 1 2 3 H4 Men need sex more than women do Agree a lot Somewhat agree Do not agree at all 1 2 3 H5 A man needs other partners even if things with his wife/partner are fine Agree a lot Somewhat agree Do not agree at all 1 2 3 108 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa H6 A woman needs other partners even if things with her husband/partner are fine Agree a lot Somewhat agree Do not agree at all 1 2 3 H7 You don’t talk about sex, you just do it Agree a lot Somewhat agree Do not agree at all 1 2 3 H8 A woman should not initiate sex Agree a lot Somewhat agree Do not agree at all 1 2 3 H9 A woman who has sex before she is married does not deserve respect Agree a lot Somewhat agree Do not agree at all 1 2 3 H10 A man who has sex before he is married does not deserve respect Agree a lot Somewhat agree Do not agree at all 1 2 3 H11 Women who carry condoms on them are loose Agree a lot Somewhat agree Do not agree at all 1 2 3 H12 Men who carry condoms on them are loose Agree a lot Somewhat agree Do not agree at all 1 2 3 H13 In my opinion, women can suggest using condoms just like a man can Agree a lot Somewhat agree Do not agree at all 1 2 3 H14 A couple should decide together if they want to have children Agree a lot Somewhat agree Do not agree at all 1 2 3 H15 It is only the woman’s responsibility to avoid getting pregnant Agree a lot Somewhat agree Do not agree at all 1 2 3 H16 If a man gets a woman pregnant the child is the responsibility of both Agree a lot Somewhat agree Do not agree at all 1 2 3 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 109 H17 It is important that a father is present in the lives of his children, even if he is no longer with the mother Agree a lot Somewhat agree Do not agree at all 1 2 3 I. Risk Perception People have different perceptions about their risk of getting HIV and AIDS. What do you think the chances are that each of the following will happen to you? Question Response Options Code I1 The chances that you will get HIV in your lifetime? No chance Some chance A high chance 1 2 3 I2 The chances that you already have HIV No chance Some chance A high chance 1 2 3 J. Attitudes Towards People Living with HIV and AIDS Please tell us if you have fear or do not have fear about getting HIV, in response to the following statements: Question Response Options Code J1 Being exposed to the spit of a person with HIV or AIDS Have fear of this Do not have fear of this Do not know 1 2 3 J2 Being exposed to the sweat of a person with HIV or AIDS Have fear of this Do not have fear of this Do not know 1 2 3 J3 Sharing eating utensils with a person who has HIV or AIDS Have fear of this Do not have fear of this Do not know 1 2 3 J4 Physically caring for a person living with HIV or AIDS Have fear of this Do not have fear of this Do not know 1 2 3 Please tell us if you agree or disagree with the following statements: Question Response Options Code J5 HIV is a punishment from God Agree Disagree 1 2 J6 HIV and AIDS are punishment for bad behavior Agree Disagree 1 2 110 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa J7 It is women prostitutes who spread HIV in the community Agree Disagree 1 2 J8 It is men prostitutes who spread HIV in the community Agree Disagree 1 2 J9 People with HIV are promiscuous Agree Disagree 1 2 J10 I would be ashamed if I were infected with HIV Agree Disagree 1 2 J11 I would be ashamed if someone in my family had HIV and AIDS Agree Disagree 1 2 K. Self-Efficacy Now please tell us about your confidence to make choices. Question Response Options Code K1 If your friends were having sex with their boyfriends or girlfriends, could you refuse to start having sex if you did not want to? Yes A little bit I’m not sure No 1 2 3 4 K2 If your boyfriend or girlfriend was pressuring you to have sex and you did not want to, could you refuse to have sex with him or her? Yes A little bit I’m not sure No 1 2 3 4 K3 If your boyfriend or girlfriend was asking you to have sex, would you be able to talk about condoms with him or her? Yes A little bit I’m not sure No 1 2 3 4 K4 Do you feel confident you could refuse to have sex even if someone offered you a meal, gifts, money or favour in exchange for sex? Yes A little bit I’m not sure No 1 2 3 4 How confident are you that you would be able to go to the clinic or hospital if you needed to for the following reasons? K5 To get information about sex-related issues (e.g. STIs, HIV, pregnancy, contraception) Not at all confident Somewhat confident Confident Very Confident 1 2 3 4 K6 To get condoms Not at all confident Somewhat confident Confident Very Confident 1 2 3 4 K7 To have an HIV test Not at all confident Somewhat confident Confident 1 2 3 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 111 Very Confident 4 K8 To get medical treatment for an STI or HIV if needed Not at all confident Somewhat confident Confident Very Confident 1 2 3 4 Please tell us more about your ability to get health services K9 If you wanted to get tested for HIV, would you know where to go? Yes No 1 2 Below are some statements about sex and condoms. For each statement please tell us if you agree a lot, somewhat agree, or do not agree at all with the statements Code K10 I can ask a new partner to use condoms Agree a lot Somewhat agree Do not agree at all 1 2 3 K11 I can ask a partner to use a condom even if we have had sex before without a condom Agree a lot Somewhat agree Do not agree at all 1 2 3 K12 I can refuse sex when I don’t have a condom available Agree a lot Somewhat agree Do not agree at all 1 2 3 K13 I can find condoms if I want to use them Agree a lot Somewhat agree Do not agree at all 1 2 3 L. HIV Testing and Counseling and Other Health Services Question Response Options Code L1 In the last 12 months, have you been to the clinic or hospital to get information about sex related issues (such as pregnancy or HIV) or to get condoms or other contraceptives? Yes No 1 2 L2 We do not want to know the result, but have you ever tested for HIV? Yes No→ SKIP TO L10 1 2 L3 When was the last time you were tested? [Respondent enter month and year] mm/yy L4 The last time you were tested, where was the test done? Government hospital Government health center Stand-alone VCT center Family planning clinic Government mobile clinic Community health worker/Ward-based outreach worker Other public medical site Private hospital Private clinic Private stand-alone health center Pharmacy NGO mobile clinic Other private medical site Home Other (specify) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 112 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa L5 The last time you were tested, did anyone accompany you to get tested? Yes No  SKIP TO L7 1 2 L6 Who accompanied you to get tested? Tick all that apply Your boyfriend or girlfriend Other sexual partner(s) Mother Father Brother Sister Other relative Friend Classmate Educator Religious leader Community leader Your doctor, nurse, or other health care worker Teacher Peer educator Other (specify) a b c d e f g h i j k l m n o x L7 The last time you were tested, did you get the results of the test? Yes No→SKIP TO L10 1 2 L8 The last time you got the results of the HIV test did you share the results with anyone? Yes No→SKIP TO L10 1 2 L9 Who did you share your results with? Tick all that apply Your boyfriend or girlfriend Other sexual partner(s) Mother Father Brother Sister Other relative Friend Classmate Educator Religious leader Community leader Your doctor, nurse, or other health care worker Teacher Peer educator Other (specify) A b c d e f g h i j k l m n o x L10 Have you ever been referred for counseling or testing for STIs or HIV? Yes No→SKIP TO L13 1 2 L11 Did you go to the facility/place you were referred to? Yes→SKIP TO L13 No 1 2 L12 Why did you not go for the referral? Tick all that apply Referral was too far Referral was too expensive I was worried about confidentiality A b c ASKED ONLY OF RESPONDENT WHO ANSWER YES TO L1, L2 OR L11. We would like to ask you some additional questions about your experience with getting health service. L13 The last time you went to a clinic, hospital or health facility for sexual health-related information or services did you feel respected by the healthcare staff? Yes No 1 2 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 113 L14 The last time you went to a clinic, hospital or health facility for sexual health-related information or services did you feel comfortable asking for the information or services you needed? Yes No 1 2 M. Pregnancy (Girls Only) Now we would like to ask you about pregnancy. Your answers will not be shared with anyone. These questions are not meant to make you feel uncomfortable. If there is a question that you really do not feel comfortable answering, you can skip that question. Question Response Options Code M1 When a woman reaches a certain age, she starts her monthly period (menstruation). Have you started your monthly period? Yes No→SKIP TO M3 1 2 M2 When was the first day of your last period? [Respondent enters date] dd/mm/yy M3 How important is it for you to avoid pregnancy now? Not important Somewhat important Very important 1 2 3 M4 What do you think are the chances that you will fall pregnant before you finish school? No chance Some chance A high chance 1 2 3 QUESTIONS M5 - M17 ASKED ONLY OF FEMALE RESPONDENTS WHO ANSWER YES TO G3 M5 Are you currently pregnant? Yes No→SKIP TO M8 Don’t know→SKIP TO M8 1 2 8 M6 How many weeks pregnant are you? [Respondent enters number] M7 Have you had any antenatal visits? Yes No 1 2 M8 Have you ever been pregnant? Yes No→SKIP TO M15 1 2 M9 What was your age at first pregnancy? [Respondent enters age] M10 Was your first pregnancy wanted at the time, wanted later, or not wanted at all? Wanted at the time Wanted later Not wanted at all 1 2 3 M11 What was the outcome of your first pregnancy? Live birth Miscarriage Abortion 1 2 3 M12 How many times have you been pregnant? In addition to babies carried to term, this may include miscarriage, babies that were stillborn, or pregnancies that were terminated [Respondent enters number] M13 How many living children do you have, that you have given birth to? [Respondent enters number] M14 How many children that you have given birth to live with you? [Respondent enters number] M15 Did you use birth control the last time you had sex? Yes No →SKIP TO M17 1 2 M16 What birth control method(s) did you use the last time you had sex? Tick all that apply Birth control pills Injectable: Depo-Provera (3 monthly injection) a b 114 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Injectable: NET-EN (known as Nuristerate, 2 monthly injections) Male condoms Female condoms Mirena/IUD Implant/Implanon Patch Cycle beads/Standard days method Emergency contraception Rhythm method/observation of menstrual cycle Traditional methods Withdrawal Other (specify) c d e f g h i j k l m x M17 Why did you not use birth control the last time you had sex? Tick all that apply I am not in a sexual relationship It is too expensive I do not know where to get it I am too embarrassed to ask for it I do not know about or understand birth control I did not think about it I am worried about the side effects Healthcare worker refused to give me birth control I am opposed to birth control My partner is opposed to birth control Other reason(s) (specify) a b c d e f g h i j x N. Pregnancy (Boys Only) Now we would like to ask you about pregnancy. Your answers will not be shared with anyone. These questions are not meant to make you feel uncomfortable. If there is a question that you really do not feel comfortable answering, you can skip that question. Question Response Options Code N1 How important is it for you to avoid getting a partner pregnant now? Not important Somewhat important Very important 1 2 3 N2 What do you think are the chances that you will get a partner pregnant before you finish school? No chance Some chance A high chance 1 2 3 QUESTIONS N3 – N9 ASKED ONLY OF MALE RESPONDENTS WHO ANSWER YES TO G3 N3 Have you ever gotten a partner pregnant? Yes No→SKIP TO N7 Don’t know→SKIP TO N7 1 2 98 N4 How many times have you gotten a partner pregnant? In addition to babies carried to term, this may include miscarriage, babies that were stillborn, or pregnancies that were terminated [Respondent enters number] N5 How many living children do you have, for whom you are the biological father? [Respondent enters number] Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 115 N6 How many children live with you, for whom you are the biological father? [Respondent enters number] N7 Did you or your partner use birth control the last time you had sex? Yes No →SKIP TO N9 1 2 N8 What birth control method(s) did you or your partner use the last time you had sex? Tick all that apply Birth control pills Injectable: Depo-Provera (3 monthly injection) Injectable: NET-EN (known as Nuristerate, 2 monthly injections) Male condoms Female condoms Mirena/IUD Implant/Implanon Patch Cycle beads/Standard days method Emergency contraception Rhythm method/observation of menstrual cycle Traditional methods Withdrawal Other (specify) a b c d e f g h i j k l m x N9 Why did you or your partner not use birth control the last time you had sex? Tick all that apply I am not in a sexual relationship It is too expensive I do not know where to get it My partner did not know where to get it I am too embarrassed to ask for it My partner is too embarrassed to ask for it I do not know about or understand birth control My partner does not know about or understand birth control I did not think about it I am worried about the side effects My partner is worried about the side effects Healthcare worker refused to give me birth control Healthcare worker refused to give my partner birth control I am opposed to birth control My partner is opposed to birth control Other reason(s) (specify) a b c d e f g h i j k l m n o x END OF SURVEY 116 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa APPENDIX 4. PROTOCOL PROTOCOL USAID South Africa Title: Impact Evaluation of a School-Based Sexuality and HIV Prevention Education Activity in South Africa Name of program/intervention being evaluated: South Africa School-Based Sexuality and HIV Prevention Education Activity Principal investigators: Ilene Speizer UNC School of Public Health, Maternal and Child Health Email: speizer@email.unc.edu Phone: 919-480-9324 Andy Beke SHSPH Email: andy.beke@up.ac.za, beke@intekom.co.za Phone: (c)0824154176 office (0123542482) Office: HW Snyman Building North 5-12 Implementing partner(s): Measure Evaluation / University of North Carolina Enterprises / University of Pretoria Implementing mechanism: MEASURE Evaluation Phase IV Leader with Associates Cooperative Agreement Version date: 31 July 2016 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 117 Brief Summary Purpose: To assess the impact of a school-based sexuality andHIV prevention education activity on learners over time Participants: Cohort of 5000 girls in grade 8 (followed through grade 10), cross-sectional sample of 5000 girls in grade 10 at baseline, and cross-sectional samples of 5000 boys in grades 8, 9 and 10 at baseline, midline and endline, from a total of 115 schools in Mpumalanga and KwaZulu-Natal provinces in South Africa. Procedures (methods): The evaluation will employ a two-arm, cluster randomized trial, where a secondary school is a cluster and a learner is the unit of observation/analysis. Impact will be measured by comparing changes over time in the incidence of HSV-2 or pregnancy; and knowledge, attitudes, and self-reported risk behaviors among the cohort of female learners and a cross-section of male learners enrolled in the selected schools between the intervention and control arms. Qualitative data will be collected, in addition to existing M&E data, to understand if the scripted lesson plans and related activities are implemented with fidelity to the design, as well as program acceptability and sustainability. 118 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Introduction Background Young people in South Africa, especially females, are at high risk of human immunodeficiency virus (HIV) acquisition. In the 2012 national survey, HIV rates rose from 3.1% among girls under age 14, to 5.2% among adolescent girls age 15-19. In 2013, 5.4% of girls 14–19 years old were pregnant (General Household Survey). Among female learners in 14 high schools in KwaZulu-Natal (KZN) province, the prevalence of HIV and herpes simplex virus-2 (HSV-2) are 6.4% and 10.7%, respectively, and the rate of teenage pregnancy is 3.6% (Karim et al, 2014). The authors of this study conclude: “The high prevalence of HIV, HSV-2 & pregnancy underscore the need for school￾based sexual & reproductive health services.” South Africa’s Department of Basic Education (DBE) has created a national HIV/AIDS strategy with the goals of reducing the incidence of HIV, other STIs, and pregnancy among learners grades 7-9 in public schools in all nine provinces (DBE 2010). The need for intervening with young women to reduce HIV incidence is further underscored by the objectives of the U.S. President’s Emergency Plan for AIDS Relief’s (PEPFAR) new Determined, Resilient, Empowered, AIDS-free, Mentored, and Safe (DREAMS) initiative. Gaining the knowledge and skills necessary to make healthy choices about sexual behavior as learners’ transition to young adulthood is key to growing an AIDS-free generation and attenuating the potentially devastating effects of the epidemic. Additionally, it is important to identify both perinatally and behaviorally HIV-infected young people in schools and link these youths to HIV testing and counseling with onward linkages to care, treatment and positive prevention. Through its life skills curriculum, in place since 2000, the DBE has attempted to address some of these issues. Previous studies found exposure to the curriculum to be associated with better knowledge, attitudes, and some practices, but there is uneven implementation among teachers and schools and challenges with the curriculum (DBE 2010). Due to these problems, DBE, with support from the United States Agency for International Development (USAID), has invested in developing scripted lesson plans (SLP) and supporting activities to increase the rigor of the life skills program, based on recommendations from the DBE report and the Southern and Eastern Africa Consortium for Monitoring Educational Quality III (SACMEQ III) project in South Africa (2010). The main component of the new program is the sexuality and HIV education program targeted to secondary school learners and delivered by life skills teachers trained in SLPs developed from an extensive review of existing life skills curricula and enhanced with interactive pedagogies. The SLPs aim to delay sexual debut, reduce unprotected sex, increase male and female condom use, reduce the number of sexual partners and reduce violence and risk. Supportive activities include mobilization andengagement of parents, school managementteams (SMTs), andschool governing bodies (SGBs); focusing on sexual and reproductive health and gender-based violence; and strengthening of referrals to health and social services. Additional supportive activities include linkages and increased access to youth friendly sexual and the integrated school health program. The new SLPs will be made mandatory, timetabled, and assessed in order to institutionalize sexuality education in schools. To ensure implementation and the roll out of the new scripted lesson plans in select areas, USAID is currently providing technical support through a new award managed by the Education Development Center (EDC) to support DBE with in-service teacher training activities, in targeted provinces and districts to strengthen DBE’s ability to implement and monitor the new sexuality and HIV prevention education program. In addition, DBE is revising current policies to promote access to HIV prevention services in schools using biomedical, behavioral, social and structural interventions to reduce transmission and vulnerability to HIV. Expected Relationship between Program Exposure and Primary Outcome Measure: Figure 1 presents the logic model that describes the expected relationship between the program exposure and primary outcome measures. The program is expected to reduce the incidence of HSV-2 or Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 119 pregnancy among female learners through one or more channels: improved knowledge, attitudinal change, increased school retention, risk mitigation, increased STI and HIV testing, and increased completed referrals for health services. The logic model guides the proposed evaluation strategy and forms the evaluation questions. Figure 1. Logic Model of DBE sexuality programme Most studies of similar programs have relied on self-reported risk behaviors, levels of HIV knowledge, and other behavioral data for assessing effectiveness. Of 83 studies of school-based sexuality programs, 65% found a significant impact on at least one sexual behavior but did not include biomarkers (Kirby et al., 2006). A randomized trial assessing the impact of a school-based program in 328 primary schools in Kenya found that the program had a significant effect on risky HIV-related behaviors and a 65% decrease in pregnancies with adult fathers (Duflo et al. 2006; Dupas 2006). In another large randomized trial in Tanzania, a teacher-led and peer-assisted sexual health education program with a community component was found to have had a significant impact on knowledge, attitudes, reported sexually transmitted infection symptoms and sexual behavioral outcomes (Ross et al. 2007). McCoy et al (2010) reviewed the results from studies of nine behavioral interventions conducted in Africa, India, Thailand and Mexico and concluded that, overall, the evidence for behavioral interventions for women and girls in low- and middle-income countries successfully changing risk behavior and prevent HIV infection is limited. Locally, the Stepping Stones program, an intensive 50-hour sexual health education activity aimed at changing gender norms and HIV risk behaviors, was evaluated in a large (n = 2,776) RTC of South African adolescents and young adults ages 15–26 by Jewkes et al. (2008). The evaluation found the program had prompted improvement in HIV testing rates, HIV-related communication, HIV-related sexual risk behaviors including condom use, and fewer men participating in the program reported IPV perpetration at 24 months than those in the control group (6% vs. 10%, p = 0.054). There was also a 33% decrease in incidence of HSV-2 in the intervention group when compared to the control group (Jewkes et al., 2008). Given that the incidence and prevalence of HSV-2 is higher for young people than HIV, this is often the biomarker of choice for these type of evaluation studies. Further, by including a biomarker for HSV-2, it is possible to have a better indication of sexual behaviors that are often poorly reported through self-report. Before rolling out the new program nationwide, the USAID Mission in South Africa and DBE want to evaluate the program to see if it is having the expected impact. By contributing to the country’s evidence base of effective HIV programming for youth, the proposed study will help to ensure that young people in South Africa receive high-quality sexuality and HIV prevention education while in school. It will help to institutionalize prevention education in the South African school system, therefore helping to ensure that programs aimed at preparing the country’s youth to address sexual health and HIV challenges are sustainable. Aims of the Impact Evaluation The goal of the proposed impact evaluation is to assess the impact of a school-based sexuality and HIV prevention education activity on learners over a period of two years. The evaluation aims to Inputs •Resources • Schools •DBE curriculum with scripted lessons and supporting activities Program activities • Engage schools • Train teachers • Implement lessons • Implement supportive activities • Site visits to schools Outputs • Schools implementing the program • Teachers trained • Lessons administered • Site visits undertaken • Supportive activites delivered Outcomes • Improvement in: • Knowledge & attitudes • School retention • Risk behavior • HIV testing • Referrals for & uptake of health services Impact • Incidence reduction among females in: • HIV (not measured) • HSV-2, or • Pregnancy 120 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa provide the South African Department of Basic Education (DBE) and PEPFAR South Africa with evidence for the effectiveness of the sexuality and HIV education program by estimating the impact of the scripted lesson plans and supporting activities on the incidence of HSV-2 or prevalence pregnancy among a cohort of girls in secondary school in two provinces of South Africa. The primary evaluation question is: What is the effect of the scripted lesson plans and supporting activities on the incidence of HSV-2 or pregnancy at the end of 10th grade among a cohort of girls enrolled in grade 8 at the beginning of 2016 in intervention schools compared to a cohort of girls in grade 8 at control schools providing the current life skills program (i.e., the standard of practice)? The secondary evaluation questions include: - What is the effect of the scripted lesson plans and supporting activities on knowledge, attitudes, school retention, and self-reported risk behavior, STI and HIV testing, and completed referrals for health services at the end of 8th, 9th, and 10th grade among a cohort of girls first interviewed in grade 8 and among a cross-section boys interviewed in grade 8, grade 9 and grade 10? - If there is a reduction in the primary outcome, does the intervention work by increasing school retention, or is the effect independent of school retention? - Do effects differ by sex or by rural vs. urban schools? The primary hypothesis is: The scripted lesson plans and supporting activities will result in a reduction of the incidence of HSV-2 and/or pregnancy over a period of three years among a cohort of girls enrolled in grade 8 at the beginning of 2016 compared to comparable girls in schools with the current life skills program. The primary outcomes of interest to test the hypothesis are HSV-2 incidence, pregnancy and the composite variable of HSV-2 incidence and pregnancy. The secondary hypotheses include: The scripted lesson plans and supporting activities will enhance school retention, knowledge, attitudes, self-reported risk and health behavior (i.e., condom use at last sex and delayed sexual debut), uptake of STI and HIV testing and counseling, and completed referrals for reproductive health/family planning services among girls and boys in intervention schools with no change or less change among their counterparts at control schools. The primary outcome measure on which this impact evaluation is powered is: the incidence of HSV-2, measured through dried blood spots, or self-reported pregnancy. There are secondary outcome measures including school retention, knowledge, attitudes, self-reported risk behavior (i.e., condom use at last sex, delayed sexual debut), uptake of STI and HIV testing and counseling, and completed referrals for reproductive health/family planning services: see Table 1 for details. The study is not powered on the secondary measures. Notably, because the changes in secondary outcomes are expected to be greater since these outcomes include key messages of the program, it is thought that powering the study for the primary outcome will result in a sufficient sample size to examine the secondary outcomes as well. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 121 Table 1. Primary and secondary outcome measures, data sources, and data collection Primary outcome Incidence of HSV-2 or pregnancy – coded 1 if young person is positive (yes) on any of these two measures; zero otherwise. Female learners HSV-2: biomarker (dried blood spots) testing at Time 1 and 3 (cohort only) Pregnancy: self-report from surveys at Time 1, 2 and 3 Secondary outcomes School retention Female learners School records of dropouts Knowledge Female and male learners Self-report from surveys at Time 1, 2 and 3 Attitudes Female and male learners Self-report from surveys at Time 1, 2 and 3 Risk behaviors Female and male learners Self-report from surveys at Time 1, 2 and 3 STI and HIV testing Female and male learners Self-report from surveys at Time 1, 2 and 3 Completed referrals for health services Female and male learners Self-report from surveys at Time 1, 2, and 3 We expect that the knowledge generated from this study will be generalizable to other provinces in South Africa and to other African nations that are seeking to meet similar goals (e.g., reduce HSV-2 and pregnancy; enhance school retention; reduce self-reported risk and health behavior; increase uptake of STI and HIV testing and counseling; and increase completed referrals for reproductive health/family planning services among school-age girls and boys) through similar life skills curricula and scripted lesson plans. The findings of the study can be used by the South African government, and governments of other African nations, to inform HIV prevention and sexuality programming for in-school youth. 122 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Methods Evaluation design The evaluation will employ a two-arm, cluster randomized trial, where a secondary school is a cluster and a learner is the unit of observation/analysis. In the intervention arm school teachers were trained to provide sexuality and HIV prevention education based on the new scripted lesson plans to learners enrolled in grade 8 at the beginning of 2016, while in the control arm the existing life orientation curriculum will be continued. Week-long teacher trainings were led by Education Development Corporation (EDC) as part of the main intervention. Notably, the training of the teachers did not happen in all intervention schools prior to the beginning of the 2016 school year (January 2016); therefore, some of the intervention students will not have full coverage of the program in the first year, however, it is expected that they will be exposed to the full program starting in grade 9. The study applies a stratified multi-stage cluster sampling approach whereby a sample of schools in each arm will be randomly selected for a longitudinal observation of learners. The impact of the new program will be measured by comparing the incidence of HSV-2 or pregnancy among the cohort of female learners enrolled in the selected schools between the intervention and control arms. The cohort of female learners in the intervention and control arms will take part in an annual panel survey from grades 8 through 10. Biomarker collection for HSV-2 will be collected from the cohort of girls in grade 8 and in grade 10. All biomarker samples collected during grade 10 will be tested for HSV-2. Biomarkers collected at grade 8 will be stored at -80 degrees Celsius and only be tested, for baseline comparison, if the corresponding grade 10 biomarker result is positive for HSV-2; this permits assessing if the HSV-2 is an incident infection since baseline. Pregnancy self-report will be obtained at each survey time. Learners will be followed even if they drop out of school to ensure a complete picture of intervention effects. Additionally, each year a cross-sectional sample of male learners in the same schools will participate in the annual surveys. A cross-sectional sample of grade 10 girls in the same schools will also be surveyed during the first year of data collection (when the cohort is in grade 8). Grade 10 girls will be used as a comparison group for the grade 10 longitudinal sample at endline. Target population and program assignment The initial coverage of the program for the evaluation will include two provinces from which five high-priority education districts with a high prevalence of STIs and pregnancy are selected for program implementation, as identified by the USAID mission in South Africa (USAID/SA) and DBE. Public secondary schools in the five districts will be randomly assigned to either the intervention or control arm for the purpose of the evaluation. DBE will provide MEASURE Evaluation with information on each secondary school in the five priority education districts, including the location, number of enrolled learners, a measurement of socioeconomic status (SES) of the catchment learners, catchment area, and drop-out and retention rates from previous years. Retention means school completion at end of the year during the years of the study. Target schools will be in the three lowest SES quintiles, schools that include grades 8-10, and schools with at least 25 learners that will be randomly assigned to either the intervention or control arm. The population covered will be a cohort of female learners enrolled in grade 8 at the beginning of 2016 and a cross-section of female learners in grade 10 and male learners in grade 8 at baseline. Teachers in the same schools, parents, School Management Teams, and School Governing Bodies will also be involved in the intervention activities. If feasible, qualitative data will be collected at midterm (2017) and endline (2018) with a sample of these groups to provide greater depth on program acceptability and sustainability. The DBE will play a key role in engaging control schools in the study and ensuring that the program will be rolled out to include the control schools upon completion of the program evaluation. Information will be collected from control schools throughout the project to determine what types of sexuality education activities are being undertaken (i.e., if any new program activities are introduced over the follow-up period). Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 123 The geographic focus where the sexuality education activity is implemented is aligned with the 27 identified priority districts for PEPFAR support in South Africa. Based on USAID guidance, in collaboration with the Department of Basic Education, the EDC intervention activity will focus in 7 sub-districts located in 7 target districts; these are City of Johannesburg – Johannesburg West (Gauteng), Thabo Mofutsanyane (Free State), Ehlanzeni – Bohlabela and Gert Sibande (Mpumalanga), City of Cape – Metro North (Western Cape), uThungulu and eThekwini – Pinetown and Umlazi (KwaZulu Natal). The evaluation will only take place in the districts in Mpumalanga and KwaZulu Natal. In the same districts where the program is being implemented, USAID also has other grantees that are undertaking HIV testing and counseling services. USAID intentionally placed the EDC program in these districts so that if referrals for further counseling and testing was needed, this would be available. Some of the sites will include mobile clinics that may visit the schools over the course of the study. As part of the data collection, the evaluation team will seek to monitor exposure at the school level (intervention and control schools) to these mobile clinic services. The School-Based Sexuality and HIV Prevention Activity is also aligned with the DREAMS Initiative guidance due to targeted focus on adolescent girls using evidence based interventions as per the UNESCO, International Technical Guidance on Sexuality Education: An Evidence informed Approach for Schools, Teachers and Health Educators, 2009. This impact evaluation will overlap with the DREAMS impact evaluation in Kwazulu-Natal, eThekwini schools located in Umlazi and Pinetown DBE districts. Exposure measurement All female and male learners enrolled in grade 8 at the beginning of 2016 in an intervention school will be considered exposed to the new program, while those enrolled in a control school will be considered exposed to the standard curriculum (i.e. the control program). The cohort of learners will receive their respective program in each school year until the end of grade 10, with an expected exposure of three school years. Surveys will ask learners questions about their exposure to the various lesson plans. This will provide information on the number of teachers trained and the level of exposure of the new program to the learners. Additionally, teachers will be interviewed about the scripted lesson plans and supporting activities they delivered during a school year for possible analysis of types of interventions (e.g. scripted lesson plans and supporting activities vs. supporting activities only) actually implemented. However, the sample size will not be powered on analyses with these different exposure variables as this is powered based on an intention to treat analysis. Data collection The data will be collected at multiple observation points over three study years from grade 8 until the end of grade 10 (i.e. between 2016 and 2018): baseline data collection in July-September 2016 (Time 1), midterm data collection in July-September 2017(Time 2), and endline in July-September 2018 (Time 3). Biomarker data collection – Dried blood spots for HSV-2 testing will be collected at baseline and stored for future testing at endline. Biomarkers will be collected again at endline and tested for HSV-2; biomarkers from baseline will be stored at -80 degrees Celsius and the corresponding baseline samples will only be tested for comparison, for those samples that are HSV-2 positive at endline. The later testing of the baseline biomarkers will permit a determination of whether an individual with HSV-2 at endline had an incident infection between baseline and endline or if she was already HSV-2 positive at baseline. All positive biomarker tests will be subject to confirmatory tests to minimize measurement error. Grade 10 learners will receive the results of the HSV-2 testing. Referral procedures will be determined prior to IRB submission prior to endline data collection. The selection of HSV-2 testing was determined because the prevalence and incidence of HSV-2 is expected to be higher than other STI and this is a better indication of sexual experience (unprotected sexual activity) than self-reported behaviors. 124 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Prior to obtaining the dried blood spots, individual parental consent and learner assent will be obtained for the biomarker part of the study. Learners will be asked to take consent forms home for parent/guardians to sign/consent. Follow up reminders will be given to the learners to return consent forms. The nurse will provide simple explanations of the testing procedures to a group scheduled for data collection on a specific date. The trained nurses will also provide a health screening for the cohort of grade 8 female learners across all three time points. The health screening involves going through a standardized check list of the algorithm for syndromic diagnosis of STI. Any student who presents a health issue, including symptoms of HSV-2, will be referred to the School health nurse (from the District Health PHC team of that catchment area) for access and treatment at the local public health clinic for further management. Furthermore, identification or recognition of any condition or situation with a legal mandatory reporting requirement by the health care worker will be managed in the best interest of the learner using the established reporting and referral network for the school. Due precautions would be taken to adhere to universal precautions of blood borne infections, needle stick injury prevention and medical waste management. The counseling and referral process based on the health screening form is as follows: The nurse will ask the girl the questions on the screening form. General counselling will be given to all girls, which informs that the discussed symptoms might be related to an STI and could be indicative of health problems that girls should get care for. We also mention that some STI spread when a person does not have symptoms, and that they should to see a doctor if they start to experience any of the symptoms. Girls who respond yes to any of the health screening questions will also be advised and given a referral letter to see a doctor for further evaluation of the symptom(s). They will be given a referral form from the nurse and be reminded that they can decide on whether they want to share the referral letter with their parent(s)/guardian. The referral letter is pre-signed by the school so that the school does not know which girls are being referred. At midterm, we will not collect the biomarkers, however, the nurses will also undertake the health screening to identify any health issues of the cohort of girls and to refer the girls. Between baseline and endline, all dried blood spots will be stored in a -80C freezer. The specimens will all remain in South Africa at the HIV Immunopathology Laboratory of the Department of Immunology of the University of Pretoria. No names will be stored with the dried blood spots; only study identification numbers will be stored for linking the endline survey and specimens over time within individuals. ODK surveys – Audio-assisted self-administered interviews will be completed by learners at each of the three survey times to measure self-reported outcomes including pregnancy, knowledge, attitudes, etc., among the cohort of female learners and a cross-section of male learners. Parental consent and learner assent will be obtained before the learner completes the self￾interview. Trained data collectors will oversee the data collection process and be available to assist the learners with the ODK survey. The surveys will be translated to local languages by professional translators. The surveys will be available in the local languages of choice for the learners. The languages available for this survey include English, Sepedi, Swati, Tsonga and IsiZulu. Adequate care and precautions would be taken to protect the safety of all field workers and the electronic instruments used for the ODK survey. The welfare of the learners will remain a priority. With due consultation with the school authorities and the school governing bodies, arrangement would also be made and funded to ensure the safe return home of the learners. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 125 Built Environment Tool – During data collection, interviewers will administer an observational checklist with a small number of questions for school administrators. The checklist will obtain information on the school environment including access to toilets, tap water, electricity, learner engagement in risk-taking behaviors (e.g., smoking on the premises). Information from the school administrator will include questions about other sexuality education activities in the last quarter, visits from a school nurse, attendance rates, and drop-out rates. Information on the built environment will be collected at baseline, midterm, and endline and will be collected from both intervention and control schools. The tools will not capture personally sensitive data. Consent will be obtained from the administrators before the built environment tool is administered. Qualitative tools – After reviewing the baseline quantitative data and in collaboration with EDC, the MEASURE Evaluation team will determine what gaps in information can better be assessed with qualitative data collection. Some thoughts on qualitative data collection include undertaking focus groups or other qualitative methods with teachers, parents and administrators to understand their engagement in the program. Details of what will be included in the qualitative data collection will be determined at a later date and all data collection guides will be submitted to the IRB as an amendment prior to data collection. We also propose to collect costs related to program implementation to inform cost-effectiveness analyses and to inform plans for scale up. Other sources – To monitor the process and quality and fidelity to program implementation, we will collect teachers’ training registers and reports, and use program reports of lessons delivered. This information will come from EDC and its partners. Data quality assurance – The University of Pretoria will hire data collection teams that will include trained interviewers, nurses and trained supervisors. All study team members will be trained on consent procedures and how to collect the data through ODK surveys, dried blood spots (nurses only), and qualitative methods. No data (hard or soft copy) will have personal identifiers. All data collection forms will be checked by the team supervisor before senior supervisors will hand￾deliver completed forms to the headquarters office. Electronic data from ODK surveys are uploaded and automatically transferred to UNC servers through a secure platform. Using ODK alleviates the need for data entry and makes data available in a more efficient timeframe. Country manager will oversee the fieldwork and submitted data, in addition to UNC staff. Sampling The study applies a stratified multi-stage cluster sampling approach in order to obtain a random sample of learners from the intervention and control domains, respectively. Sampling frame – The sampling frame will be constructed from a list of schools provided by the DBE, which has information on the location and number of learners. The frame will be stratified within each domain to enhance the representativeness of the sample (strata may include district, urbanicity, poverty level, and school size depending on the number of schools). We will finalize and make any necessary revisions to the proposed sampling design when all necessary information is provided to ensure that the study objectives are met. Because the program needed to push forward with implementation, school selection has already occurred. In total, the study includes 58 intervention schools (23 in Mpumalanga and 35 in KZN) and 57 control schools. The school is classified as the Primary Sampling Unit (PSUs). The current plan is that all female and male learners in grade 8 in January 2016 in the selected schools will be approached and asked to take a parental consent form home; i.e., all are eligible for inclusion in the study producing a self-weighting sample within provinces such that larger schools contribute more learners than smaller schools. Given that the number of learners in 2016 was not 126 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa available by the time of sampling, we included supplementary schools (the intention was to have 50 intervention and 50 control schools). Statistical power/sample size The goal of the sample size calculation is to power the statistical analysis on the primary outcome: the incidence of HSV-2, or pregnancy among female learners. We performed sensitivity analysis of the sample size calculations based on various combinations of parameter assumptions to inform the sample size of learners and schools presented here. The required sample size based on simple random sampling with an assumed attrition rate at 5% per year (or 14.3% over three school years) is 675 female learners per arm to detect the difference in the HSV-2 incidence rate of 0.04 in the intervention arm vs. 0.08 in the control arm over three school years at α of 0.05 (two-sided) and power (1-β) of 0.8. Likewise, the required sample size based on simple random sampling with the same level of attrition rate at 5% per year is 614 female learners per arm to detect the difference in the incidence rate of HSV-2 or pregnancy of 0.06in the intervention arm vs. 0.11 in the control arm over three years at α of 0.05 (two-sided) and power (1-β) of 0.8. We adjusted the sample size for: 1) design effect to account for elevated standard errors in a cluster sample design, and 2) baseline prevalence of the primary outcome to account for loss of units available to estimate the incidence rate. First, the design effect from clustering is approximated as 1 + ICC × (𝑀̅ − 1) where ICC is intra-cluster correlation and 𝑀̅ is the average cluster size (Kish, 1965). We assumed a maximum ICC of 0.03 and plan to take a random sample of an average of 50 female learners per school, implying a design effect of 2.47. Next, we assumed the prevalence at 1% of HSV-2 and 0% of pregnancy, implying the prevalence at 1% in the incidence of HSV-2 or the composite indicator of HSV-2 and pregnancy among the target population. With a total sample size in both arms together of 5,000 female learners (50 female learners per school and 100 schools in total), we estimate a power of 92% for the analysis of the incidence of HSV-2, and 94% for the analysis of the incidence of HSV-2 or pregnancy. The study is not specifically powered on the secondary outcomes; however, given that the level of change is expected to be larger for these outcomes we expect to have a sufficient power to perform the analysis. The expected sample size for boys is 5,000, or 50 boys per school per year over three years. Inclusion Criteria - The inclusion criteria include the following: - all girls in grade 8 in selected schools at time of enrollment (baseline); followed longitudinally those that enroll in the study - all boys in grade 8 in selected schools at time of enrollment (baseline) - all girls in grade 10 at the time of enrollment (baseline) - 58 intervention and 57 control schools were selected randomly to participate in the study. The inclusion criteria for the selection of schools include: - schools in PEPFAR priority provinces - secondary schools - schools in the Quintiles 1-3 range - schools with at least grades 8-10 present All schools that met these selection criteria were reviewed by the provincial DBE to determine if there are any reasons to eliminate any of the eligible schools prior to selection. Once the list was finalized, selection of the intervention and control schools happened. We need a minimum of 50 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 127 schools in each group, however, we selected a larger number in case of school refusal or small class sizes. Exclusion Criteria There will be no exclusion criteria of schools besides random selection of the schools. For the participating learners, the survey will be administered in English plus four local languages will be available. If a learner is unable to take the survey in one of these five languages, she may need to be excluded if the interviewer does not speak the necessary language to support completion of data collection. Selection Bias: The aforementioned randomization will minimize self-selection bias of schools and learners at baseline between the intervention arms as they have an equal chance of being assigned to either the intervention or control arm. However, selection bias may be of concern when learners of certain characteristics transfer or drop out of schools in a systematic way in different arms after their initial assignment. We will monitor learners’ transfer or drop out and their exposure to program throughout the evaluation to examine the extent of possible selection bias. The goal is to follow up with all learners in the cohort of grade eight girls at baseline, even if they transfer or drop out. However, we are not able to find learners that transfer or drop out, then the learner will be dropped out of the study. We will also apply an intention-to-treat analysis and analyze data of learners based on their initial assignment of program. Another source of selection bias is attrition of learners. If those learners are systematically different from those in school in the outcomes or relevant characteristics, it may bias the study results. We will monitor attrition of female learners and retrieve contact information from girls and schools to follow up with those at home who have left school at midterm and endline. An important component of the program is to follow ALL baseline participating learners. The survey at the end of the first year will obtain detailed contact information from all learners. This will happen for learners in both the intervention and control schools. Detailed tracking mechanisms including telephonic and home visits will be employed to ensure the highest follow-up rates possible, even among those who leave the school or community. A final source of bias is based on which learners receive parental consent and which ones do not. If there is a systematic reason for not receiving parental consent (e.g., parents are away or do not read/understand the form), this may bias the data. If possible, we can use school-level information to determine if there are differences in demographic characteristics of learners who receive consent to participate and those who do not. 128 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Methods of Recruitment DBE will provide MEASURE Evaluation with information on each secondary school in the five priority education districts, including the name, type, location, and quintiles. Public secondary schools in the five districts were randomly assigned to either the intervention or control arm. All schools will be informed by DBE and subsequently contacted by researchers prior to the project implementation and evaluation to obtain permission to be included in the study. After initial contact, trained study staff will explain the study and criteria for participation to both the learners and the parent/legal guardian. Consent/Assent Written informed consent will be obtained from the parents prior to the initiation of study related activities and prior to each round of data collection. For all learners who receive parental consent, on the day ofthe survey and biomarker collection, wewill requestthe learner’s assentto participate. These are standard procedures for data collection with young people. If the learner is 18 or older she will sign her own consent form. The consent and assent will detail all study procedures including potential risks and benefits and time spent participating in the research. Participants will read or have read to them the consents/assents before study procedures are begun. Participants will provide their signature on the assent and consent forms. Parents/legal guardians who are unable to read or sign their name can make a mark (a cross) in the presence of a witness. Signed copies of the consent and assent forms will be given to the participants. After endline data collection, any young person who tests positive for HSV-2 will be referred for further testing and treatment. Referral letters will be given to the learner to deliver to her parents. Potential study participants will be told that they are free at all times not to participate in the research and that refusal to participate will in no way affect their academic enrollment or performance at their school. Participants will also be encouraged to share with researchers any concern they might have regarding their researchparticipation. Measurement Challenges There may be measurement error in drop outs rates for girls and boys as reported by the schools due to inaccurate record keeping. Learner responses about their exposure to the sexuality education activities in intervention and control schools may be over-reported due to social desirability bias and recall error/bias. While measurement errors in these records will not affect the primary outcome measurement, it may affect the aforementioned dose-response analysis depending on the nature and magnitude of the measurement error through statistical power (when the error is random) or bias (when the error is systematic). Teacher delivery of the scripted lesson plans will be randomly monitored by the program team to ensure fidelity to the intervention. The proposed study will use the program reports from various site visits and report any deviations. Also teachers’ records of delivered lessons can be used as another program intervention exposure measure. Key confounding factors - Confounding is of concern when there is a systematic and unobserved difference in characteristics between intervention groups that also influences the outcome of interest. We attempt to prevent the problem through random assignment of schools to the intervention or control arms. Because the number of schools to be selected may be moderate and may not eliminate a potential imbalance by chance, we will stratify the schools by potentially relevant characteristics prior to randomization and ensure balance in these characteristics. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 129 Additionally, we will examine the baseline characteristics of schools and learners to assess the extent of attained balance between the arms and report the results; any observed differences will be controlled for in multivariable models. Social desirability bias and recall error – Self-reported outcomes, such as pregnancy, attitudes, and risk behaviors, may be subject to social desirability bias and recall error. One way to reduce this error is to use audio assisted self-administered survey software to implement the learner surveys. Open Data Kit (ODK) is an open source mobile data collection platform. With ODK, surveys are programmed into an electronic format, inclusive of data quality checks, skips and logics, to reduce data collection error. The survey is then uploaded onto a tablet via ODK, for field data collection. Audio files will be attached to each question, which allows the interviewee to both hear (through headphones) and see the question on the tablet. Biomarkers – Biomarker outcomes for HSV-2 are subject to false positives and negatives. All positive results will be sent for confirmatory biomarker testing. Spillover effects – There may be spillover effects when there are peer interactions between learners from different study arms (i.e., those not exposed to the new program gain information and knowledge provided through their peers in the new program). We will collect information on peer￾interactions through self-report surveys to examine the extent of spillovers at the data analysis phase. Contamination or inadequate program exposure - The aforementioned problem of peer interactions and learner’s transferring in and out of schools between the intervention and control arms may result in contamination of program exposure. By asking girls the school she attended in the previous year, we can attempt to know about exposure in the control schools. Additionally, delay or interruption in program implementation may result in inadequate program exposure. We will ensure that the lesson plans are delivered on schedule by reviewing monitoring sheets of implementation of the lesson plans. We will also monitor individual learner exposure to the lesson plans via teacher records and possibly via interviews with the learners. Analytical Framework The program impact on the primary and secondary outcomes will be measured through multivariable regression for repeated measures to analyze the longitudinal data. Specifically, we will apply Generalized Linear Mixed Models (GLMM) to allow for both fixed and random effects and response variables from different distributions (binary, continuous, Poisson, etc.). The impact will be estimated by the coefficient of the program indicator as a fixed effect. We will control for other covariates to enhance the estimate precision and to account for imbalance in observed characteristics (which should be minimal after randomization) as fixed effects with clusters (i.e. schools) treated as random effect. At the end of the first year and subsequently at the end of the second school year we can examine some of the secondary outcomes (e.g., school retention, knowledge, attitudes, and risk behaviors) to see if they are changing in the expected direction and determine if there are greater changes in the intervention as compared to the control group; similar analysis methods will be used for these analyses as for the endline analyses as described above. We will investigate impact heterogeneity by testing for coefficients of interactions between the program indicator and characteristics of interest (e.g. sex, school type, age, urban/rural, socioeconomic status, academic achievement, language group) in our analytic regression model, although the study is not powered for these tests. If the units have different observation periods and points (due to attrition despite our effort to follow up with female learners who have left school) we may apply survival analysis to address censoring. We will account for the sample design (variable sampling weights) to estimate the 130 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa parameters of interest. The analyses will be performed in Stata statistical software (StataCorp, College Station, TX). Confidentiality of Data Anonymous data collection strategies are put in place to maintain a high level of confidentiality. No questionnaires or specimen collection forms will contain names or other explicit identifiers. Confidentiality will be discussed in detail with all members of the research team, and will be addressed during a training session on research ethics. Site staff will receive Human Subjects Protection training. Training will emphasize the importance of protecting the confidentiality of study participants and the data collected about the participants. All data will be password protected and accessed only by authorized members of the research team. All data collection forms will be checked by the team supervisor before data are uploaded to UNC. All logs of data activities, data printouts, etc. will be kept in a locked and secure file cabinet, and digital data will be stored in a password-protected and encrypted computer database. Once data collection is complete and the data has been cleaned, any paper study materials will be locked up; these files will be kept for 15 years as required by the IRB. Complete data sets and the meta data and do files will be stored in a password-protected and encrypted computer database for 15 years at the Scholl of Health Systems and Public Health, HW Snyman building, University of Pretoria. The specimens of participants will be destroyed at the end of the study. Future researchers will not have access to the code that links the specimen to the participant. The UNC team will work with the in-country partners to identify a mechanism to destroy materials in a safe and secure manner. Ethical Review The final study protocol including study procedures, data collection tools, and consent forms will be approved by the Faculty of Health Sciences Research Ethics Committee at the University of Pretoria (UP). All schools will be contacted prior to the project implementation and evaluation to obtain permission to be included in the study. Prior to approaching the schools, we will work with the provincial and district-level DBE offices to obtain their approval. Prior to each round of data collection, a form will be sent home to the parents of the participating youth (e.g., grade 8 girls and boys at time 1; grade 9 girls and boys at time 2, and grade 10 girls at time 1 and time 3) requesting parental consent for their child to participate in the survey and biomarker data collection (grade 8 – baseline girls and grade 10 – endline girls only). For all youth who receive parental consent, on the day of the survey, we will request the young person’s assent to participate. These are standard procedures for data collection with young people. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 131 Timeline Table B-1 presents the timeline of the proposed evaluation activity. The baseline survey will take place in August 2016, followed up interim survey in August 2017 and the endline survey in August 2018. Table B-1. Timeline of the proposed evaluation (organized by calendar year) Activity 2015: Year 1 2016: Year 2 2017: Year 3 2018: Year 4 Q2- 3 Q 4 Q 1 Q 2 Q 3 Q 4 Q 1 Q 2 Q 3 Q 4 Q 1 Q 2 Q 3 Q 4 USG-DBE planning, approvals, and procurement X X Development of research X X X Submission of protocol to South African and U.S. ethics committees X Expected receipt of research ethics approval X Pre-survey sampling field visits X Pilot test of data collection tools X Data collector training X Data collection: annual survey X X X Data cleaning X X X Data analysis and disseminatio n meeting X Drafting of research report X Disseminatio n events X 132 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa Innovation The rigor of this study (randomization and biomarker outcomes) should make a significant contribution to the evidence base for behavioral interventions, as its design attempts to address shortcomings from previous studies on such interventions, including life skills programs. We are not aware of another life skills program that has been evaluated in such a rigorous manner. And the rigorous, standardized way in which the life skills curriculum is being delivered (through scripted lesson plans) is novel and maximizes the likelihood of such interventions having an impact, by reducing implementation variability. This study will therefore contribute to the evidence base of effective HIV programming for youth. The proposed randomized study design minimizes methodological limitations and problems often observed in quasi-experimental and observational studies, and provides a strong ground for examining the program impact. The study design will use a biomarker outcome (incidence of HSV-2) and self-reported pregnancy as a primary outcome for assessing the effectiveness of the sexuality and HIV prevention education program. This represents a novel and advantageous approach to estimating the impact of school￾based sexual health and HIV programs, as most similar programs have relied on self-reported risk behaviors and other behavioral data for assessing effectiveness. The advantage of utilizing biomarkers (for HSV-2) is that it gives a clearer, more direct and more accurate indication of learners’ sexual risk behaviors. The study design will also include components that examine teacher-learner modes of interaction that are mostly innovative in the South African school setting. Lessons are commonly delivered in a didactic manner. Teachers will be trained to utilize an interactive pedagogy with guides for addressing difficult issues using participatory methods. If feasible, the evaluation will use a mixed methodology and attempt to capture and measure the effects of this learning approach as well as the actual lesson content using quantitative and qualitative data sources. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 133 References Abdool Karim, Q. (2014) Prevalence of HIV, HSV-2 and pregnancy among high school learners in rural KwaZulu-Natal, South Africa: a bio-behavioural cross-sectional survey. Journal of Sexually Transmitted Infections doi:10.1136/sextrans-2014-051548 Department of Basic Education. 2010. Integrated strategy on HIV and AIDS 2012-2016, Full Report. Pretoria. Duflo, E, Dupas P, Kremer, M. and Sinei, S. (2006), “Education and HIV/AIDS Prevention: Evidence from a Randomized Evaluation in Western Kenya,” available at www.povertyactionlab.org. Dupas P. 2006. Relative risks and the market for sex: teenagers, sugar-daddies and HIV in Kenya. Available at https://www.povertyactionlab. Com/papers/Dupas_RelativeRisk-1007.pdf. Shisana O, Rehle T, Simbayi LC, Zuma K, Jooste S, Zungu N, Labadarios D, Onoya D et al. (2014) South African National HIV Prevalence, Incidence and Behaviour Survey, 2012. Cape Town: HSRC Press. Jewkes R, Nduna M, Levin J, Jama N, Dunkle K, Puren A, Duvvury N. (2008) Impact of stepping stones on incidence of HIV and HSV-2 and sexual behaviour in rural South Africa: Cluster randomized controlled trial. BMJ.37:1–11. doi: 10.1136/bmj.a506. Kirby D. et al. 2006. Sex and HIV education programs for youth: their impact and important characteristics. Scotts Valley. CA: ETR Associates. Kish, Leslie. "Survey sampling." (1965). McCoy SI, Kangwende RA, Padian NS. Behavior change interventions to prevent HIV infection among women living in low and middle income countries: a systematic review. AIDS Behav. 2010; 14(3):469-482. Minnis A. and N. Padian (2001). Reliability of adolescents’ self-reported sexual behavior: a comparison of two diary methodologies. Jrnl. of Adolescent Health 28(15):394-403. Ross D. et al. (2007). Biological and behavioral impact of an adolescent sexual health intervention in Tanzania: a community-randomized controlled trial. AIDS 21: 1943-1955. Stats SA (2014) General Household Survey 2013. Available at: www.statssa.gov.za. Stats SA (2015) General Household Survey 2014. Available at: www.statssa.gov.za. 134 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa APPENDIX 5. EVALUATION TEAM MEMBERS Ilene Speizer, PhD is the Principal Investigator and activity lead for the impact evaluation. Her responsibilities include overall development and implementation of the evaluation, collaboration with the local research partner, and coordination with USAID/South Africa and the implementing partner. She is a Research Professor at the University of North Carolina at Chapel Hill and a Research Associate on the MEASURE Evaluation project with more than 20 years of research and evaluation experience in public health. Her primary research interests focus on impact evaluations for adolescent sexual and reproductive health. This evaluation builds on previous work in program evaluation and studies on family planning, HIV and AIDS, and STI prevention. She is currently Co-Principal Investigator and the Technical Deputy Director on the Bill & Melinda Gates-funded Measurement, Learning & Evaluation (MLE) project that is evaluating family planning programs in four countries. Andy Beke, MMED is based in South Africa and serves as the local Principle Investigator for the evaluation. He is a senior lecturer and adjunct professor at the University of Pretoria and a principal specialist with the Mpumalanga Department of Health in South Africa. Dr. Beke has 35 years of experience providing direct medical care, managing health service delivery points, and providing leadership and support to occupational and environmental health programs. Dr. Beke has also been involved with monitoring and evaluating occupational and environmental health programs, and coordinates and teaches several public health courses. Mahua Mandal, PhD is a Monitoring and Evaluation Specialist in Youth and Gender and leads the qualitative component of the evaluation. Additionally, she works closely with the PI on the overall development and implementation of the evaluation and assists with data analysis and report writing. She has fifteen years of experience in public health, and has conducted mixed-methods research and evaluation in women’s, children’s, and adolescent health for eight years. She has conducted evaluations of HIV and AIDS, reproductive health, adolescent health, and gender-based violence programs in Africa and Asia. Prior to conducting research and evaluation, Dr. Mandal was the Youth Reproductive Health Technical Advisor at USAID/Washington. Khou Xiong, MPH is a Research Associate and the lead individual working closely with the PI on protocol development, tool development, and all phases of data collection. She works closely with the local research partners on the training of trainers and training of enumerators and supports all data-collection and field￾based activities. She liaises on a regular basis with the South Africa-based Research Associate on day-to-day activities. She assists with data analysis and report writing. Ms. Xiong’s work in public health incudes research in water sanitation and use, sexual reproductive health, HIV and AIDS, and mHealth applications to address health issues in under-resourced communities. Aiko Hattori, PhD is the Sampling Expert for the South Africa impact evaluation. She conducts sampling calculations and supports selection of the schools for the intervention and control sites. In addition, she conducts data analysis. Ms. Hattori is a Research Associate and Public Health Economist with MEASURE Evaluation and has been working in the field of public health for more than 15 years. She has a PhD in Maternal and Child Health and an MPH in Biostatistics from the University of North Carolina at Chapel Hill. Her research has centered on evaluations of health policies and programs both in developed and developing countries. Her research at MEASURE Evaluation since 2012 has involved impact evaluations of public health projects in multiple developing countries, including Bangladesh and Zambia. She has designed and conducted sample designs for population surveys and managed and analyzed the collected data. Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa 135 Milissa Markiewicz, MPH, PMP provides project management and logistic support to the evaluation. Ms. Markiewicz is a Research Associate at MEASURE Evaluation and currently serves as project manager for a regional impact evaluation on cross-border health in East Africa (Uganda, Rwanda, Kenya, and Tanzania); an integrated health services impact evaluation in Malawi; and an impact evaluation in Zambia focused on gender and groundnut farming. She previously served as Project Manager for the SE Region of the Network for Public Health Law at the University of North Carolina at Chapel Hill, and as a Program Director at the Terry Sanford Institute of Public Policy at Duke University. Ms. Markiewicz worked in Uganda for more than three years as an Academic Director and Special Projects Coordinator for the School for International Training. Rick O’Hara, BS is a Data Manager/Analyst at the Carolina Population Center, University of North Carolina at Chapel Hill, where he has provided data management services for more than two decades. He worked most recently with the Measurement, Learning, & Evaluation (MLE) project, where he managed data for four countries including India, Kenya, Nigeria, and Senegal. Prior to this, Mr. O’Hara worked as a data manager/analyst on several Carolina Population Center projects, including Add Health, South Africa Testing Innovations, National Children’s Study, Nang Rong Projects, Russia Longitudinal Monitoring Survey, and the Iowa Youth and Families Project. Ndinda Makina, MPH is the team’s South Africa-based Research Associate. She works with closely with local research partners on the training of trainers and training of enumerators and supports all data￾collection and field-based activities. She liaises closely with USAID/South Africa, DBE, and the U.S.-based evaluation team. She assists with data analysis and report writing. Ms. Ndinda has contributed to research design, fieldwork data collection, data analysis, and reporting in a number of studies, including a nationwide study assessing the barriers of education in South Africa and evaluating malaria interventions in Malawi. Stephanie Watson-Grant, DrPH serves as Country Portfolio Manager for the evaluation. She provides oversight and support to the MEASURE Evaluation team throughout the implementation of the evaluation. She monitors compliance with MEASURE Evaluation Phase IV Agreement conditions and reporting requirements and approves work-plan deliverables. Dr. Watson-Grant has more than 10 years of experience in the field of international health and development. Her areas of expertise are HIV planning and program implementation, assessment of HIV monitoring and evaluation systems, management of USAID-funded projects, survey implementation, capacity building training, and measurement of country ownership. 136 Baseline Report: Impact Evaluation of a Sexuality and HIV Prevention Activity in South Africa