WPS7797 Policy Research Working Paper 7797 Combining Mentoring Programs with Cash Transfers for Adolescent Girls in Liberia Baseline Report Kelly Hallman Elizabeth Kelvin Berk Özler Juliette Seban Erica Kuhlik Cooper Alton Joseph Kamara Sarah Goodman Development Research Group Poverty and Inequality Team August 2016 Policy Research Working Paper 7797 Abstract This report presents findings from the baseline assessment females reported having previously had sex. Within this of International Rescue Committee’s (IRC) Girl Empower group, 29 percent indicated that their first sexual act was non- (GE) program in Nimba County, Liberia. GE seeks to consensual. Among all GE girls, 37 percent reported having help 13 to 14 year-old girls make healthy life choices and experienced sexual violence of some type, such as being phys- decrease their risk of sexual abuse. The program centers on ically forced to have sex, non-physically pressured (coerced/ weekly meetings between girls and trained local mentors, persuaded) to have sex, someone unsuccessfully attempting during which the girls learn about life skills and financial to have sex with them, and being touched in a sexual way. literacy. GE also holds monthly discussion groups for par-   The levels of nonconsensual first sex and any experi- ticipants’ caregivers, and trains local health and psychosocial ence of nonconsensual sex are at the high end of the range care providers on how to improve and expand services for reported by the UNICEF Violence against Children Sur- survivors of gender-based violence. This baseline report is veys (VACS) in Swaziland, Tanzania, Kenya and Zimbabwe. part of a cluster-randomized controlled trial, which aims As the VACS reporting is for (a variety of ) age ranges, each to assess the program’s impact 24 months after baseline. of which is higher than that in this study, the levels of   21 percent of the baseline sample of 13–14 year-old sexual violence reported here are very high in comparison. This paper is a product of the Poverty and Inequality Team, Development Research Group. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The authors may be contacted at bozler@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Combining Mentoring Programs with Cash Transfers for Adolescent Girls in Liberia: Baseline Report Kelly Hallman*, Elizabeth Kelvin**, Berk Özler***, Juliette Seban†, Erica Kuhlik‡, Cooper Alton‡, Joseph Kamara‡, Sarah Goodman* JEL codes: J13, J16 Keywords: Adolescent girls, mentoring programs, non-consensual sex, sexual violence * Population Council, ** City University of New York School of Public Health, *** The World Bank, † International Rescue Committee, and ‡ Innovations for Poverty Action. Corresponding author: khallman@popcouncil.org. Acronym Table Acronym Meaning IRC International Rescue Committee GE Girl Empower GE+ Girl Empower plus, a study arm offering caregivers cash transfer incentives IPA Innovations for Poverty Action CRIES Children’s Revised Impact of Events Scale RSES Rosenberg Self Esteem Scale SMFQ Short Moods and Feelings Questionnaire MCEF Monitoring, Conflict, Emotional Support and Financial Support Scale 1 Contents  Summary........................................................................................................................................ 4  Introduction ............................................................................................................................... 5  Study Design .............................................................................................................................. 6  Survey Procedures..................................................................................................................... 7  Randomization........................................................................................................................... 9  Balance Table .............................................................................................................................. 11  Introduction to Results ............................................................................................................... 13  Basic Household Characteristics ............................................................................................... 13  Basic Girl Characteristics .......................................................................................................... 15  Basic Caregiver/Head of Household Characteristics............................................................... 17  Girl Survey .................................................................................................................................. 19  Ebola Impact............................................................................................................................ 19  School Participation ................................................................................................................ 20  Sexual History.......................................................................................................................... 22  Consent for First Sexual Act .................................................................................................. 24  Condom use.............................................................................................................................. 25  Pregnancy................................................................................................................................. 25  Marriage and Relationships ................................................................................................... 26  Physical Violence ..................................................................................................................... 26  Sexual Abuse and Exploitation .............................................................................................. 27  Post-traumatic Stress .............................................................................................................. 30  Social Networks and Safety .................................................................................................... 30  Self-Esteem............................................................................................................................... 32  Aspirational Attitudes ............................................................................................................. 33  Attitudes Regarding Gender Roles and Violence ................................................................. 34  Relationship and Communication with Adults .................................................................... 35  Mental Health .......................................................................................................................... 38  Financial Literacy ................................................................................................................... 39  Health Literacy ........................................................................................................................ 41  Knowledge about Family Planning Methods ........................................................................ 42  Knowledge about Condoms .................................................................................................... 43  Knowledge about HIV/AIDS and Sexually Transmitted Infections................................... 44  Caregiver Survey ........................................................................................................................ 44  2 Attitudes about Girls: Gender Role Norms .......................................................................... 44  Attitudes about Girls: Child Rearing .................................................................................... 45  Discussion..................................................................................................................................... 47  SEXUAL DEBUT .................................................................................................................... 47  FIRST SEX UNWILLING ..................................................................................................... 47  EVER EXPERIENCED ANY SEXUAL ABUSE ................................................................ 48  NATURE OF SEXUAL ABUSE EVER EXPERIENCED .................................................. 49  EXPERIENCE OF ANY SEXUAL ABUSE IN PAST 12 MONTHS ................................ 49  NATURE OF SEXUAL ABUSE EXPERIENCED IN PAST 12 MONTHS ..................... 50  PERSPECTIVES ON OTHER KEY OUTCOMES ............................................................ 50  School enrollment ................................................................................................................. 50  Parental co-residence and survival........................................................................................ 51  Not enrolled in school and not living with either parent ...................................................... 51  Pregnancy .............................................................................................................................. 51  Marriage ................................................................................................................................ 51  Age-disparate relationships ................................................................................................... 52  HIV knowledge ..................................................................................................................... 52  Condom use .......................................................................................................................... 52  Gender norms ........................................................................................................................ 53  Rosenberg Self-Esteem Scale (RSES) .................................................................................. 53  Short Mood and Feelings Questionnaire (SMFQ): Assessment for Depression .................. 53  Children’s Revised Impact of Event Scale (CRIES-8): Assessment for PTSD.................... 54  CAREGIVER SURVEY ......................................................................................................... 55  Gender Norm Attitudes ......................................................................................................... 55  Attitudes about Child Rearing ............................................................................................... 55  References .................................................................................................................................... 57  Appendix ...................................................................................................................................... 60  Supplementary Survey Tables ............................................................................................... 60  Respondent Comprehension and Survey Environment ...................................................... 88  Caregiver Survey .................................................................................................................. 88  Girl Survey ............................................................................................................................ 89  3 Summary  This report presents the results from the baseline assessment of International Rescue Committee’s (IRC) 1 “Girl Empower” (GE) program in Nimba County, Liberia. GE is a program in rural communities which seeks to help 13 to 14 year-old girls make healthy life choices and decrease their risk of sexual abuse. The program centers on weekly meetings between girls and trained local mentors, during which the girls learn about life skills and financial literacy. GE also holds monthly discussion groups for participants’ caregivers, as well as trains local health and psychosocial care providers on how to improve and expand services for survivors of gender- based violence. Girls are also equipped with savings accounts, and small deposits are made on their behalf. The GE baseline assessment is part of a randomized evaluation, which will assess the program’s impact. Primary investigators from Population Council, the World Bank, and the IRC lead the evaluation’s research team, and Innovations for Poverty Action (IPA) is responsible for the survey data collection. Villages were identified for inclusion in the study based both on research and programmatic considerations. The research team made an initial selection of Nimba County villages based on reported village population, distance from the IRC office and likelihood of accessibility. Staff from IRC Liberia then visited all of the identified villages to verify population and accessibility. If a village was significantly outside the target population range or was inaccessible, the research team replaced it with an alternate. Between July 12 and September 6, 2015, four teams of IPA enumerators completed data collection in 100 study villages across Nimba County. Among those villages, however, 16 had fewer than five girls who were eligible for the program (girls had to be between the ages of 13 and 14 to qualify). Given that GE mentorship groups require a minimum of five girls, these 16 villages were dropped from the study. Among the final 84 study villages, the enumeration team completed a total of 10,930 household listing surveys, 1,216 girl surveys, and 1,055 caregiver surveys. After data collection was complete, the research team randomly assigned the 84 study villages to the control group (28 villages), the GE treatment group (28 villages) and the GE+ treatment group (28 villages). Both the GE and GE+ treatment groups will receive the standard girl mentorship, savings account and caretaker discussion group components of the program, while the GE+ group will also receive a small cash incentive to caregivers to encourage regular attendance by girls at mentorship meetings. At the beginning of the program implementation 817 girls and 130 mentors were participating in the GE and GE+ arms; girl group size ranged from 6 to 20, with a mean of 14.6. The GE impact evaluation is designed to assess a variety of outcomes, primarily the program’s effectiveness in reducing instances of and vulnerability to sexual abuse and exploitation. Also evaluated is the program’s effectiveness when participants’ caregivers receive cash transfers based on girls’ attendance at the weekly sessions. Further investigated is the program’s ability to help reduce teen pregnancies and early marriages, increase social connectedness, school 1 See Acronym Table on page 1 for full list of abbreviations. 4 participation, and self-esteem, and improve girls’ life and financial literacy skills. Other relevant outcomes include whether or not the girl has ever had sex, whether the girl’s first and subsequent sexual acts were consensual and whether or not the girl has ever been married or pregnant. In order to get pre-randomization measures of these (and all other) outcome indicators, all of the outcome measures were included in the baseline assessment. Among girls interviewed in the baseline assessment, only 0.74% reported that they were married or were living with a man as if they were married. One-fifth (20.72%) reported having previously had sex. Among girls who had had sex, 7.94% reported having ever been pregnant (n=20; 1.64% of the total sample). For those who have sexually debuted, 71.43% indicated that their first sexual act was consensual, while 13.89% reported that they were physically forced and 14.68% reported that they were tricked. Among all GE girls, 37.3% report having ever experienced sexual violence of some type. The nature of these experiences were reported as: 7.8% had been physically forced to have sex, 8.4% had been non-physically pressured (coerced/persuaded) to have sex, 24.7% had had someone unsuccessfully attempt to have sex with them, and 28.9% had been touched in a sexual way. The percentage of GE girls who had experienced sexual debut as of the baseline survey was not noticeably higher for their age than available comparable sources from the region. The levels of nonconsensual first sex and any experience of nonconsensual sex are within, but at the high end of, the range of those reported by the UNICEF Violence against Children Surveys (VACS) - the program of surveys most comparable to GE with regard to age group and sexual violence survey question style and content - in Swaziland, Tanzania, Kenya and Zimbabwe. In particular, GE respondents reported much higher levels of unsuccessful attempted sex and sexual touching than in the other four countries. It should also be noted that the VACS reporting was for (a variety of) age ranges that were each older than the GE respondents. The levels of all types of sexual violence reported in the GE survey are therefore very high in comparison. Nearly 98 percent of all survey participants indicated having ever attended school; of those, the largest proportion (nearly 21 percent) indicated second grade as their highest level of schooling reached. Typical for the overall setting, 85.03 percent of GE participants reported having been enrolled during the academic year prior to the survey (in this case, the 2014-2015 school year). The mean and median self-esteem scores reported by GE survey respondents were within the normal range. Introduction  The following presents results from the baseline assessment conducted as part of an evaluation of International Rescue Committee’s (IRC) Girl Empower (GE) program in Nimba County, Liberia. GE is an innovative program designed by the IRC to equip girls with the skills and experiences necessary to make healthy, strategic life choices and to stay safe from sexual and abuse and exploitation. The program aims to accomplish these goals through a combination of mentorship, asset building, girl and caregiver discussion groups and local capacity-building. As part of the 5 program, the IRC invites adolescent girls in rural communities throughout Nimba County to join mentorship groups led by trained local mentors. During the weekly mentorship meetings these groups hold, the girls learn about life skills and financial literacy and open savings accounts with seed money from the program. In addition, the program holds monthly discussion groups for caregivers of program participants and trains local health and psychosocial care providers on how to improve services for survivors of gender-based violence. The baseline assessment described in this report is part of a clustered randomized controlled trial being conducted to evaluate the impact of the GE program. The evaluation has three study arms: comparison, GE and GE+. The GE treatment arm includes the program components described above. The GE+ treatment arm includes the same program components, but also provides a small cash incentive to families of girls in the program to encourage regular attendance at mentorship meetings. Primary investigators Kelly Hallman (Population Council), Elizabeth Kelvin (City University of New York), Berk Özler (World Bank) and Juliette Seban (IRC) led the research team for the GE baseline assessment and Innovations for Poverty Action (IPA) carried out the data collection with support from the PIs. The survey field team was led by Erica Kuhlik, Joseph Kamara and Cooper Allton. The baseline assessment was originally scheduled to being in July, 2014, but had to be postponed as a result of the Ebola outbreak in Liberia in 2014. Preparations for the assessment resumed in June, 2015. IPA began data collection on July 12th and subsequently completed it on September 6th. The ethics procedures used in the research were approved by Internal Review Boards of both the Population Council (protocol number 645) and the University of Liberia – Pacific Institute for Research and Evaluation (UL-PIRE). Study Design  The primary research questions to be answered through the evaluation include:  What impact does the GE program, which includes mentoring, asset building (this includes life skills, financial literacy, and savings) and caregiver discussion groups, have on reducing adolescent girls’ experiences of sexual abuse and exploitation?  Does the effectiveness of the GE program improve when a cash transfer to caregivers component is added? In addition, the evaluation will specifically seek to answer the following questions: does program participation:  Decrease experiences of sexual abuse and sexual exploitation in the target population?  Reduce teen pregnancies and early marriages?  Decrease their social isolation and deepen their social networks (e.g. number and diversity of friends)?  Increase their school participation?  Increase their self-esteem, self-confidence, and self-efficacy?  Increase girls’ capacity for crucial life skills (decision-making, communication, negotiation, self-protection, understanding and awareness of violence, financial literacy)? 6  Increase the protective factor of family/home life through increased attention to their wellbeing by their caregivers? The primary outcome indicators identified in evaluation plan are (1) whether the girl has ever had sex (2) whether she has ever been pregnant (3) whether she has ever been married (4) whether her first sexual act was non-consensual and (5) whether she experienced violence in the past year. In addition, school enrollment is also identified as a primary outcome. Other outcome measures are scale-based, including the Rosenberg Self Esteem Scale (girls) (Rosenberg, 1965), the Short Mood and Feelings Questionnaire (SMFQ) Scale to assess depression (girls) (Angold et al., 1995), the Gender Equitable Attitudes Scale (girls) (Stephenson et al., 2012), the Children’s Revised Impact of Event Scale (CRIES-8) (girls) (Perrin, et al, 2005), a Monitoring, Conflict Emotional Support and Physical Support scale (girls) (Bingenheimer, et al., 2014), and the Gender Norms Attitudes Scale (caregivers) (Waszak et al., 2000). The study frame for the GE baseline assessment consisted of 100 villages in Nimba County, Liberia. The research team selected these potential program villages on the basis both of size and of accessibility. Within each village randomly assigned to program status, all 13-14 year old girls are eligible to participate in the GE program. The IRC planned to have one mentorship group of approximately 20 girls in each program village. To help ensure this, the research team used village population data from the Liberia Institute of Statistics and Geo-Information Services (LISGIS) to develop a list of villages likely to have the appropriate number of eligible girls. All villages in the initial selection had a listed population of between 700 and 1500. The IRC team in Nimba County then checked the list to assess the accessibility and distance of proposed study villages. When villages were found to be inaccessible or too distant, the research team replaced them with villages of similar size. Once this updated list was complete, the IRC Liberia team traveled to each of the potential study villages to verify population and accessibility. If IRC Liberia found that any of the proposed villages were inaccessible or were too large or small for the program, the research team replaced the village with an alternate. This resulted in a final list of 100 villages that were both accessible and within the targeted population range. After data collection was complete, the research team determined that 16 of the study villages had fewer than 5 five girls eligible for the GE program (i.e. between the ages of 13 and 14 years old). Given that GE mentorship groups had to have a minimum of 5 girls each, these villages were dropped. As a result, 84 communities were included in the final study frame for the GE baseline assessment. Survey Procedures  Four IPA teams, each consisting of four enumerators and one team leader, completed all data collection for the baseline assessment. Given the sensitive nature of the subject matter, all enumerators were female. Enumerators were responsible for interviewing respondents, while team leaders were responsible for enumerator supervision and spot-checking interviews. Upon entering a new study village, the team leader informed the Town Chief about the assessment and requested a town meeting to explain it. Once convened, the team leader explained the purpose of the assessment, emphasizing the fact that participation was voluntary and that respondents would not receive any material benefits. Afterwards, the team leader answered questions from 7 members of the community. Once the community entry process was complete, the team leader assigned enumerators to different zones in the village and they started data collection. Enumerators began data collection in each home by using chalk to mark a door or window with a unique house code. The enumerators then asked to speak with the head of household. If the head of household was not home, the enumerator asked to speak with any other adult (at least 18 years old). If the head of household or another adult was present, the enumerator introduced herself, read the household consent statement, answered any questions the respondent had and then asked for the respondent’s consent to complete the household listing survey. If the respondent consented, the enumerator completed the interview. The household listing survey asked general questions about the household, including how many families lived there and whether there were any adolescent girls. If the respondent reported that there were adolescent girls in the household, the enumerator asked for further details including their ages. If any of the girls listed were between 13 and 14 years old, the enumerator asked to speak with their caregiver. Once the enumerator located the girl’s caregiver, she explained the purpose of the study and read and explained the girl survey’s caregiver consent statement. If the caregiver consented for the girl to be interviewed, the enumerator met the girl, explained the purpose of the assessment and read and explained to her the girl survey assent statement. If the girl assented, the enumerator completed the girl survey and tracing survey with her. After completing the interview, the enumerator asked to speak with the person the girl identified as her caregiver in the survey (often the same person who completed the caregiver consent for the girl). Once the enumerator located that person, she introduced herself, explained the purpose of the assessment and read and explained the caregiver survey consent statement to the caregiver. If the caregiver consented, the enumerator completed the caregiver survey with him/her. If a respondent could not be located or was not available when the enumerator visited a particular home, she noted the house code in a log sheet and moved on to the next house. She later returned to the home and again attempted to find the respondent. In cases where, after multiple attempts, the enumerator was still unable to locate the respondent, she entered the house code and other pertinent information in a “mop-up” sheet. A separate “mop-up” team then revisited the village after the team left and attempted to find the respondent. In addition to the four enumeration teams, a separate team of two back-checkers provided an additional layer of data quality control. The back-checking team visited each study village after the enumeration team finished data collection and re-interviewed randomly-chosen respondents using a selection of the original survey questions. If back-check survey responses were significantly different from the original survey responses for a particular respondent, the IPA Field Manager followed up to investigate the cause of the discrepancies. Enumerators completed all data collection with encrypted PDAs (smartphones) running SurveyCTO (a survey application based on Open Data Kit). Each survey an enumerator finished was saved in her PDA for later upload. Team leaders periodically collected enumerator PDAs and uploaded their data to an encrypted cloud server using a 3G data connection and/or a secure mobile Wifi hotspot. Uploaded data was immediately available to the IPA management team, who ran daily high-frequency checks to verify data quality. 8 Randomization  After data collection was complete, the research team randomly assigned the 84 remaining study villages to the control group or one of the two treatment groups (GE and GE+). Before carrying out the randomization, the team first blocked (stratified) the villages based on the number of girls surveyed and the average level of school enrollment. The randomization process resulted in 28 of the villages being assigned to the GE treatment arm, 28 to the GE+ treatment arm and 28 to the control arm. During the randomization, two of the principal investigators traveled to Liberia to oversee and explain the randomization process to IRC staff and to ensure that all parties were aware of which villages had been assigned to the control and two treatment groups. The results are shown in the figure below (Figure 1). Once this was complete, the IRC began preparing for the implementation of the GE program in all of the villages that were assigned to one of the two treatment arms. 9 10 Balance Table  The table below shows the level of balance in average values for select variables between individuals assigned to the treatment and control groups. The blocking variables used in randomization and the primary outcome variables of the impact evaluation were included in the table. The first column presents the mean (standard deviation) of each variable in the control group. Columns 2-4 show the difference in the average value of the selected variable between control and treatment groups (both for combined treatment and the individual GE and GE+ treatment arms), as well as the standard error of that difference. Column 5 presents the p-value for the test of equality of means between the two arms of the GE treatment group (GE and GE+). The bottom row presents p-values from a test of joint orthogonality to indicate whether all the indicators included in this table are jointly different in one study arm than another. Examining the individual differences between randomly assigned treatment arms, we see no statistically significant differences between the control group and any of the treatment groups (or the combined treatment group). This is reflected in the p-values from the joint orthogonality tests reported in the bottom row, the lowest of which is 0.39 (columns 2-4). However, we do detect a slight imbalance between the two treatment arms (p-value=0.094 in the bottom row of column 5). This imbalance seems to be driven by a large difference in the Children’s Revised Impact of Event Survey (CRIES) scale scores in the GE and GE+ groups: the GE group is less likely to have a CRIES scale score (meaning that more of them were below the screening cutoff for PTSD) and, among those screened in, the scores are significantly lower in the GE group than in GE+. As this is a cluster-RCT and randomization was not blocked on this particular outcome indicator, chance imbalances such as this one can occur. Overall, the balance in the trial seems acceptable, with no notable differences between the control group and any treatment arm and only a difference that is significant at the 10% level between the two treatment arms, which is primarily due to an imbalance in the CRIES outcome indicator. 11 Table 1. Balancing of Variables across Groups Test for Mean (and Difference in Means (with Equality of Number of standard respect to control group) Parameters Variable Obser- deviation) for (p-values) vations control group All GE GE+ GE = GE+ treated 0.209 -0.002 -0.010 0.006 Ever had sex 0.596 1,215 (0.407) (0.029) (0.030) (0.034) 0.013 0.006 0.005 0.007 Ever pregnant 0.827 1,215 (0.112) (0.006) (0.008) (0.007) 0.015 0.008 0.015 0.002 Married 0.515 1,215 (0.212) (0.014) (0.021) (0.014) 0.063 -0.005 -0.013 0.002 Firs sex non-consensual 0.320 1,215 (0.243) (0.013) (0.014) (0.015) 30.843 0.235 0.020 0.444 Rosenberg scale 0.300 1,212 (4.249) (0.446) (0.492) (0.484) 9.010 -0.753 -0.087 -1.396* SMFQ scale 0.097* 1,214 (6.739) (0.718) (0.876) (0.713) 7.633 -0.164 -0.151 -0.176 Gender Relations Scale 0.947 1,197 (3.041) (0.301) (0.371) (0.337) 24.947 -0.297 -0.306 -0.288 MCEF Composite Scale 0.959 1,212 (3.602) (0.247) (0.295) (0.316) 17.490 1.13 -0.893 2.76* CRIES Scale 0.001*** 465 (8.157) (1.42) (1.42) (1.48) 0.002 0.038 -0.033 No CRIES Score 0.617 0.103 1,216 (0.035) (0.041) (0.040) (0.487) 0.844 0.011 0.008 0.014 Enrolled 2014-15 0.864 1,216 (0.095) (0.024) (0.033) (0.026) 19.286 1.930 2.471 1.406 Number girls 0.819 1,216 (9.546) (3.318) (4.450) (3.598) Chi-squared test of joint 0.8432 0.9170 0.3903 0.0938* 1,184 orthogonality (p-value) Notes: stars represent statistical significance according to: .01 ***; .05 **; .1 *. Cross-section OLS regressions at girl level with SE (between parentheses) clustered at the village level. In the column 'Mean (and standard deviation) for control group', numbers in parentheses are standard deviations, not standard errors. 12 Introduction to Results   The following sections present results from the GE baseline assessment. The first three sections describe the basic characteristics of the households and respondents interviewed. Results from these sections are taken from all three of the baseline survey instruments (household, girl and caregiver surveys). Unless otherwise noted, the tables are for single-response questions. For questions in which the respondent could choose more than one response, the corresponding table has a note at the bottom indicating “multiple responses allowed”. Tables with two or more joined sections are for contingent questions (questions for which the preceding question determines whether or not the next question will be asked). For questions with continuous responses (such as number of days missed from school), tables list basic descriptive statistics including the mean, standard deviation, minimum and maximum for the variable. Basic Household Characteristics  As part of the caregiver survey, respondents were asked about household water sources and sanitation facilities. The most common water source in caregiver homes was a dug well with a hand pump in the yard/plot (49.05%), followed by a public tube well, borehole, or dug well with a pump (29.46%) (see appendix Table A1). Furthermore, a significant percentage (47.57%) reported that people in their homes did not use toilet facilities, and instead practiced open defection. The next most common practice reported was the use of a pit latrine with slab (cement floor) (22.96%) and pit latrine without slab (open pit) (15.08%) (see appendix Table A2). 13 Caregivers were also asked whether or not anyone living in the home had various items. More than half had a mattress (77.82%), a table (71.49%), a cell phone (59.88%), or a radio (58.84%); 7.63% had a generator, 13.52% had a motorcycle or motorbike and just 3.38% had access to electricity from the grid. Table 2. Caregiver survey item A17: Does your  household or someone living in your household have  any of the following?  Yes (%)  Number  A mattress (not made of straw or grass)?  77.82%  898  A table?  71.49%  825  A cell phone?  59.88%  691  A radio?  58.84%  679  Chairs?  48.53%  560  A watch?  15.16%  175  A motorcycle or motor bike?  13.52%  156  A cupboard?  9.27%  107  A generator?**  7.63%  88  A television/DVD player?  4.59%  53  An ice box?  3.55%  41  Electricity (current)?  3.38%  39  A sewing machine?  2.08%  24  A bicycle?  1.91%  22  A car or truck?  1.30%  15  A boat or a canoe?  0.87%  10  A computer?  0.78%  9  *multiple responses allowed  Respondents:  1,154  ** “Don’t know” recorded as a response in one case for “A generator".  Responses for all other items were either yes or no.  During the household listing survey, respondents were asked about the families and adolescent girls living in the house. The majority of households (70.47%) had just one family living in the house, whereas 13.16% had two families, 9.32% had three and just 7.05% had more than three (Table A3 in appendix). A minority (35.92%) of households reported having girls over the age of 12 living in the home. Among the households that had girls in that age range, 70.22% reported having one and 23.34% reported having two. Table 3. Household listing item A11. Are there any  adolescent girls, by which I mean girls over 12 years  old, living in this household?  By this I mean that they  sleep here most nights and eat food from here  throughout the year.  Percent  Number  No  64.08%  6957  Yes  35.92%  3899     Total  10856  14 Table 4. Household listing item A12. If adolescent girls  aged 12 or older (A11=yes), how many girls around this  age live here?  Percent  Number  1  70.22%  2738  2  23.34%  910  3  5.03%  196  4  1.13%  44  5  0.13%  5  6  0.05%  2  7  0.05%  2  8  0.03%  1  10  0.03%  1     Total  3899  Basic Girl Characteristics  Most (88.16%) of the girls interviewed during the GE baseline assessment reported having been born between 2000 and 2002 (the birth-year range that corresponded to the target age range of 13-14 years-olds).2 Another 10.61% did not know their year of birth. Girls were selected for interview based on the age the caregiver reported. As a result, a small number of girls whose self-reported birth year fell outside the target range were still interviewed (1.22% of cases) (see appendix Table A4). Almost all girls (97.94%) had attended school (see appendix Table A5). Figure 2. Girl survey item B2: Have you ever attended school? 2% No (25) Yes (1191) 98% The most common grade levels reached were grade 2 (20.99%), grade 3 (18.39%), grade 1 (17.13%) and grade 4 (15.20%) (see appendix Table A6). The average number of years of school 2 Depending on the month and day of, girls born in 2000 and 2002, however, could fall outside the target age range. All girls born in 2001 fall within the target age range. 15 (starting with ABC=1 year, K1=2 years, K2=3 years, and grades 1-9 equally 4-12 years) completed was 5.17, and the median was 5. Table 5. Girl survey item B3b: How many years of school have you completed?  Observations  Mean  Median  SD  Min  Max  1168  5.17  5  1.97  0  11  A larger percentage of girls lived with their biological mother than with their biological father (74.59% for mothers vs. 58.63% for fathers). Among girls who did not live with their mothers, the most common reasons reported were that the girl was sent away for school (42.39%), that the mother remarried (17.15%) or that the mother died (12.94%). Among girls who did not live with their fathers, the most common reasons reported were that the father died (26.84%), the father remarried (20.28%) or the girl was sent away to school (19.28%). Table 6. Girl survey item B12: Is your biological mother  living with you?  Percent  Number  No  25.41%  309  Yes  74.59%  907     Total  1216  Girl survey item B13: If not with mother, what is the  main reason you are not living in the same house?  Percent  Number   I left or was sent away for school   42.39%   131   Mother remarried   17.15%   53   Mother died   12.94%   40   Mother got divorced / separated   7.44%    23   Mother left for work   7.12%   22   Other reason(s)  6.80%   21   I left or was sent away for work   4.53%   14   I got married   1.62%  5     Total  309  16 Table 7. Girl survey item B14: Is your biological father  living with you?  Percent  Number  No  41.37%  503  Yes  58.63%  713     Total  1216  Girl survey item B15: If not with father, what is the  main reason you are not living in the same house?  Percent  Number  Father died  26.84%  135  Father remarried   20.28%   102  I left or was sent away for school  19.28%  97  Father left for work   12.33%   62  Father got divorced / separated   9.15%   46  Other reason(s)   6.36%   32  Don’t know   0.20%   1     Total  503    Basic Caregiver/Head of Household Characteristics  The majority of caregivers interviewed in the caregiver survey were female (72.27%) (see appendix Table A7). When asked about the highest grade level they completed in school, the most common response was none (24.00%). The next most commonly-reported level was grade 12 (11.09%). Most of the rest of the caregivers (58.85%) fell somewhere between grade 1 and grade 11, with a relatively even spread across grades (see appendix Table A8). 17 Figure 3. Caregivers' Highest Completed Grade Level (%) 30 25 24.00 20 15 11.09 10 6.85 6.5 7.54 7.54 5.2 5.55 5.11 5.81 5 3.55 3.64 1.56 2.17 3.03 0.17 0.61 0.09 0 In the majority of cases (71.92%), the caregiver interviewed was the head of household. In cases where the caregiver was not the head of household, the head of household was most commonly the father of the girl (58.02%) (see Table A9 in appendix). Table 8. Caregiver survey Item B5: Are you the head of  this household?  Percent  Number  No  28.08%  324  Yes  71.92%  830     Total  1154  Among household heads who were not the caregiver, the most common highest level of educational attainment was no education (21.30%), followed by grade 12 (25.00%) (see appendix Table A10). Among all household heads (regardless of whether or not they were the caregiver interviewed), the most common type of primary employment was agriculture/fishing (self-employed) (69.84%), followed by business/self-employed (e.g. market vendor) (9.79%). Table 9. Caregiver survey Item B8: What is the main thing the  head of the household is doing for living?  Percent  Number  Unemployed  4.07%  47  Agriculture/fishing (self‐employed)  69.84%  806  Agriculture/fishing (i.e. paid by someone else to farm/fish)  5.55%  64  18 Non agriculture wage labor (i.e. paid by someone else on an  hourly/daily basis)  1.65%  19  Business/self‐employed (e.g. market vendor)  9.79%  113  Salaried/ professional  7.63%  88  Domestic worker (e.g. housewife)  0.69%  8  Student  0.09%  1  Other  0.69%  8     Total  1154  Girl Survey  Perceived Ebola Impact  The GE baseline assessment took place just as the 2014-2015 Ebola crisis ended in Liberia. As a result, Ebola-specific questions were added to the girl survey to gauge the impact of Ebola on the girls interviewed. Among all of the girls interviewed, 4.19% reported that Ebola had killed one or more of their family members. Within that group, the most commonly reported family member killed was an aunt (38.78%) or an uncle (32.65%) (see appendix Table A11). Most girls (80.59%) reported that the Ebola crisis had no continuing impact on their lives at the time of the interview. Among those who did report a continued impact, the most common was that the girl or her family members who used to go to school were no longer going to school (12.01%) or that the girl was afraid or less confident (7.24%). 19 Table 10. Girl survey Item C1: What are all of the ways that  the Ebola crisis is still affecting your life now and the life of  your family in terms of income, schooling, living conditions  and loss of family or friends?     Percent  Number  No impact now  80.59%  980  Me or my family members that used to go to school are no  longer going to school   12.01%  146  We are afraid, less confident   7.24%   88  It stopped my family from going to work   6.74%   82  Family members are hungry more often  5.43%   66  It killed some of my family members   4.19%   51    Me and my family members cannot afford transportation  costs, and we don’t travel as much as we used to   3.37%   41  The family had to change houses   0.58%   7  It is difficult for my family to pay for housing   0.33%   4  Family members get more sick   0.33%   4  Other   0.33%   4  Don't Know  0.00%  0  No Response  0.00%  0  *multiple responses allowed  Respondents:  1,216  School Participation  The GE baseline assessment spanned two school years and was fielded during a time of irregularity in the academic calendar because of the Ebola crisis. Among all interviewed girls, 60.20% reported that the 2014/2015 school year was still in session at the time they were interviewed, whereas 32.24% reported that school was not currently in session, while 5.26% reported that the next (2015/2016) school year was in session. This outcome is presented because school breaks and holidays may impact upon girls’ vulnerability to sexual violence. Table 11. Girl survey item B4. What school year is  currently in session?  Percent  Number  2014/2015 school year  60.20%  732  2015/2016 school year  5.26%  64  School not currently in session  32.24%  392  Don't know  2.22%  27  No response  0.08%  1     Total  1216  20 The vast majority of girls (85.03%) had been enrolled during the 2014/2015 school year (the one that began before the Ebola outbreak). The most commonly-cited reason for lack of enrollment that year was that the family could not afford it (48.89%) or the Ebola epidemic (21.11%). Table 12. Girl survey item B5: Let’s talk about the  2014/2015 school year that was shortened due to  Ebola. Were you enrolled in that school year, OR if that  school year is still in session are you currently enrolled?  Percent  Number  Yes, I am/was enrolled  85.03%  1034  No, I am/was not enrolled  14.80%  180  Don't know  0.16%  2     Total  1216  Girl survey item B5a. if you were not enrolled (“no” to  B5), what is the first main reason you are (were) not  enrolled in the 2014/2015   Percent  Number  Family could not afford  48.89%  88  Result of Ebola epidemic  21.11%  38  Too many domestic responsibilities  5.00%  9  Finished school  4.44%  8  Got pregnant  2.78%  5  No school places available  (school full)  1.67%  3  School too far / no school in vicinity  0.56%  1  Family does not approve/see benefit  0.56%  1  Other  15.00%  27     Total  180  Most girls (83.55%) were not enrolled in the 2015/2016 school year at the time they were interviewed, mainly (95.96% of cases) because registration had not yet opened (see appendix Table A12). Girls reported missing an average of 2.67 days of school in the month before the interview (or the last month school was in session). Table 13. Girl survey item B5b: How many days did you miss school last month or in the  last month that school was in session?  Observations  Mean  Median  SD  Min  Max  1010  2.67  2  3.41  0  23  21 Among girls who had missed school during that period, 54.89% reported that the primary reason was that they were sick. The second most commonly-cited reason for missing school was having housework or farm work to do (15.95%). Table 14. Girl survey item B5c: If missed school (item  B5b>0), what was the main reason you missed school  last month or in the last month school was in session  Percent  Number  Sick  54.89%  382  Had housework/farm work to do  15.95%  111  Tuition outstanding  7.18%  50  Had to take care of a child or sick relative  4.60%  32  Too tired  4.31%  30  Did not want to go to school  3.74%  26  Had to go to work for money  1.15%  8  Other  8.19%  57     Total  696  Sexual History  The GE baseline assessment asked girls a variety of questions about their sexual history. When asked about how much peer pressure they were under to have sex, 38.24% reported they were under no pressure, 21.22% reported they were under a little pressure and 32.57% reported they were under a lot of pressure (see appendix Table A13). Figure 4. Girl survey item O7: How much peer pressure is there on  people your age to have sex? (%) 45 40 35 30 25 20 15 10 5 0 None (465) A little (258) A moderate A lot (396) A great deal Don't know (6) amount (47) (44)   22 However, the majority (82.07%) reported never having “kissed a boy in a loving or sexy way” (see appendix Table A14). When asked whether or not they had ever had sex, 20.72% of girls reported that they had (see appendix Table A15). Figure 5. Girl survey item O8: Have you ever had sex, whether this was something you  wanted to do or not? This includes when a man’s penis  enters someone’s vagina or anus. 0% 21% No (963) Yes (252) Don't Know (1) 79% Among girls who had previously had sex, 97.62% reported having had vaginal sex and 3.97% reported having had anal sex. Table 15. Girl survey item O9: If have had sex (“yes” to  item O8), what kind of sex have you had?    Percent  Number  Vaginal  97.62%  246  Anal  3.97%  10  Don't Know  0.00%  0  No Response  0.00%  0  *multiple responses allowed  Respondents:  252  Among girls who had ever had sex, the average age at which they first had sex was 12.3 years old and the median age was 13. Table 16. Girl survey item O10: If have had sex (“yes” to item O8), how old were you  when you first did man and woman business, by which I mean vaginal or anal  intercourse?  Observations  Mean  Median  SD  Min  Max  247  12.3  13  2.62  1  14  The average number of sexual partners these girls report having had in their lifetime was 1.43 and the median number was 1. During the 12 months before the interview, these girls had an average of 1.24 sexual partners and a median of 1 sexual partner. 23 Table 17. Girl survey item O12: If have had sex (“yes” to item O8), can you tell me how  many sexual partners you have had in your lifetime?  Observations  Mean  Median  SD  Min  Max  251  1.43  1  0.76  1  5  Table 18. Girl survey item O13. If have had sex (“yes” to item O8), in the past 12 months,  how many different sexual partners have you had?  Observations  Mean  Median  SD  Min  Max  250  1.24  1  0.5  1  3  Consent for First Sexual Act  As shown in Figure 5 below (and appendix Table A16), among girls who had ever had sex, 71.43% reported that they had sex for the first time because they wanted to. In contrast, 13.89% reported that the first time they had sex, they were physically forced, and 14.68% reported that they were tricked. Figure 6. Girl survey item O11:  The first time had sex, would you say you did it because you  wanted to do it or because you were forced or tricked into  doing it against your will? 15% Wanted to (180) 14% Physically forced (35) Tricked (37) 71%   24 Condom use  Girls interviewed as part of the GE baseline assessment reported low rates of condom-usage. Among girls who had ever had sex, the majority (68.65%) reported that they had never used a condom during intercourse in the 12 months before the survey. Just 7.54% reported that they had always used a condom and 23.41% reported they had sometimes used a condom. Table 19. Girl survey item O14:  If have had sex (“yes”  to item O8), in the past 12 months, how often did you  or your partners use a condom (male or female) during  vaginal or anal intercourse?  Percent  Number  Always  7.54%  19  Sometimes  23.41%  59  Never  68.65%  173  No Response  0.40%  1     Total  252  Asked whether or not they had used a condom the last time they had sex, 87.30% (of girls who had ever had sex) reported that they had not. Table 20. Girl survey item O15: If have had sex (“yes” to  item O8), the last time you had sex, did you use a  condom?  Percent  Number  No  87.30%  220  Yes  12.70%  32     Total  252  Pregnancy  Among girls who had ever had sex, 7.94% (20 girls) had been pregnant (see appendix Table A17). Each of these had been pregnant only once. Table 21. Girl survey item O17: If ever pregnant (“yes” to O16), how many times have you  ever been pregnant?  This includes all pregnancies, even if the pregnancy did not end  with a live birth.  Observations  Mean  Median  SD  Min  Max  20  1  1  0  1  1  Most of the girls who had been pregnant reported having wanted to wait (until later) to get pregnant (65.00%) or that they had not wanted children at all (25.00%). Only 10.00% of girls who had been pregnant (n=2) reported that they wanted to get pregnant (see appendix Table A18). Among the girls who had ever been pregnant, 45% were pregnant at the time of the interview, 30.00% had given live birth to the child, 15% had aborted and 10% had miscarried 25 (see appendix Table A19). Among those who gave live birth to their child, the child was most commonly living with the respondent (66.67%) or with the father (20.00%) (see appendix Table A20). Marriage and Relationships  Almost all of the girls interviewed during the GE baseline survey (99.18%) were not married or living with a man/boyfriend as if married. Table 22. Girl survey item O2. Are you married now or  you living with someone just like you are married?    Percent  Number  No  99.18%  1206  Yes married but not living with husband  0.16%  2  Living with a man/boyfriend as if married  0.58%  7  Don't Know  0.08%  1     Total  1216  Among those who were married or living with a man/boyfriend as if married (9 girls), 88.89% reported that they got married or started living together willingly (see appendix Table A21). The average age of these girls’ husbands/partners was 18.43 years old and the median age was 18 years old (see appendix Table A22a). Among girls who were not married, 20.31% reported that they had a boyfriend (see appendix Table A22b). The average age of the girl’s boyfriend was 15.83 years old, whereas the median age was 17 years old (see appendix Table A22c). Physical Violence  A significant proportion of girls interviewed as part of the GE baseline assessment had been victims of physical abuse. Just under half, 49.18%, of girls reported that someone had ever hit, slapped, kicked or done anything bad to hurt them (see appendix Table A23). 26 Figure 7. Girl survey item P1: Has anyone ever hit, slapped,  kicked, or done anything bad to hurt you before? 0% No (617) 49% Yes (598) 51% Don't know (1) Among those who reported having ever been physically abused in these ways, 4.85% reported that the abuse had happened often in the 12 months before the interview and 79.93% reported that it had sometimes happened. Table 23. Girl survey item P2:  If ever abused (“yes” to  P1), in the last 12 months, how often have you been  hit, slapped, kicked, or physically hurt by this/these  person(s)?  Percent  Number  Often  4.85%  29  Sometimes  79.93%  478  Never  15.22%  91     Total  598  Sexual Abuse and Exploitation  Ever experienced sexual abuse or exploitation  One of the primary objectives of the GE program is to reduce the incidence of sexual abuse and exploitation. In order to gauge this at baseline, the GE girl survey included a section focused on the topic. Girls were first asked if they had experienced a particular type of sexual abuse or exploitation and then whether or not it had happened in the past 12 months (see appendix Table A24). As shown in the Figure 7 below, over one-third (38.32%) of all girls had experienced at least one of the four types of sexual abuse. The most common type of abuse girls reported was sexual touching: 28.87% of girls reported having ever been touched in sexual way without their permission, and among those girls, 86.89% reported that it had happened within the past 12 months. 27 The second-most common form of abuse girls reported was unsuccessful attempts to force them to have sex: 24.67% of girls reported that someone had tried to have sex with them against their will, but had not succeeded, and 85.33% of those girls reported that this had happened to within the past 12 months. The third most common type of sexual abuse was being pressured in a non-physical way (coerced or persuaded), which 8.39% of girls reported. The fourth most common experience was being physically forced to have sex, which 7.81% of girls reported. The majority of girls who experienced these two types of abuse said that the abuse happened in the 12 months prior to the interview. Figure 8. Prevalence of ever experienced sexual abuse  (%, n=1216) 40 35 30 25 20 15 10 5 0 Sexually touched Unsuccessful Non‐physically Physically forced "Yes" to any type attempted sex pressured 28 With regard to sexual exploitation, 7.48% of girls reported being given money to have sex; 7.89% reported being given gifts or favors to have sex; and 10.65% had been given either money or gifts or favors to have sex. The majority of girls who experienced sexual exploitation said that it occurred in the 12 months before the survey. Figure 9. Prevalence of ever experienced sexual exploitation  (%, n=1216) 15 10 5 0 Received money for sex Received goods for sex "Yes" to either money or goods for sex Experienced sexual abuse in the past 12 months  The prevalence of sexual abuse in the past 12 months among all girls surveyed (not just among those who answered “yes” to ever experienced of each type), closely mirrors the figures for ever- experienced. This implies that much of the sexual abuse experienced by girls has begun rather recently in their lives. Figure 10. Prevalence of experience of sexual abuse in past 12 months   (unconditional % among all survey respondents, n=1216) 40 35 30 25 20 15 10 5 0 Sexually touched Unsuccessful Non‐physically Physically forced "Yes" to any type attempted sex pressured   29 Post‐traumatic Stress  If a girl reported having been the victim of physical or sexual abuse or exploitation (“yes” to any of one of girl survey questions Q1-Q6), she was asked a series of questions related to post- traumatic stress. The questions in this section were taken from the Children’s Revised Impact of Event Scale (CRIES-8). The CRIES-8 scale is a tool used by international programs3 to screen children at risk for post-traumatic stress symptoms. The CRIES-8 questions ask respondents how often they have particular reactions related to their experience of abuse. Respondents are asked to select between: Not at all, Rarely, Sometimes or Often. These responses are scored as 0, 1, 3 and 5 respectively. The scores for each question are then summed to create intrusion and aversion subscales as well as a composite scale. The average CRIES aversion subscale score was 10.67 (of a possible 20), the average CRIES intrusion subscale was 7.59 (of a possible 20) and the average CRIES composite scale (the sum of the two subscales) was 18.25 (of a possible 40) (see appendix Table A25). Table 24. CRIES‐8 Composite Scale (range: 0‐40)  Observations  Mean  Median  SD  Min  Max  465  18.25  18  7.77  0  40  Table 25. CRIES Intrusion Subscale (range: 0‐20)  Observations  Mean  Median  SD  Min  Max  465  7.59  8  4.75  0  20  Table 26. CRIES Aversion Subscale (range: 0‐20)  Observations  Mean  Median  SD  Min  Max  466  10.67  10  5.17  0  20  Social Networks and Safety  As part of the baseline assessment, girls were asked a series of questions about their social networks and sense of safety. Asked whether or not there was a place they could go to sleep in an emergency situation, most (51.89%) girls reported there was not (graph below and appendix Table A26). 3 http://www.childrenandwar.org/measures/children%E2%80%99s-revised-impact-of-event-scale-8-%E2%80%93- cries-8/ 30 Figure 11. Girl survey item E1: Percentage of Girls with a Safe Place Away from Home to  Sleep during an Emergency 48% No (631) 52% Yes (585) Among those who reported that there was such a place, 79.66% indicated that place was a relative’s home and 22.74% reported it was a friend’s home. Table 27. Girl survey item E2: If yes to safe place outside of  home to sleep (yes to E1), where could go to sleep if there was  an emergency or something happened that made you feel  unsafe where you usually sleep?     Percent  Number  Relative’s home   79.66%   466  Friend’s home   22.74%   133  Home of boyfriend  2.74%  16  Other   1.54%   9  Shelter   0.34%   2  Don't Know  0.00%  0  No Response  0.00%  0  *multiple responses allowed  Respondents:  585  With regard to safety around schooling, most girls (93.12%) who attended school indicated that they felt safe there. Most (91.40%) also reported that they felt safe traveling to school (see appendix Table A27 and Table A28). The majority of girls (80.92%) indicated that there was a safe place in the community outside of school or home where they felt free to meet and talk freely with other girls. Among those who reported that such a place existed, 72.97% said they rarely went to that safe place and 22.15% said that they went there every day (see appendix Table A29 and Table A30). When asked about their relationships with adults, just 35.36% of girls reported that there was a woman in the community outside of their own household who they could usually go to with problems (see appendix Table A31). In terms of friendships with other girls, 95.72% of girls reported that they had girl friends around their age (outside their own household). Among those 31 who had girl friends around their age, the average number of friends was 3.19 and the median number was 3. Figure 12. Girl survey item E7: Availability of an Local  Adult Female Source of Social Support  outside Girl's Household 35% No (786) Yes (430) 65% Table 28. Girl survey item E8. Do you have any girl friends  around your age group outside of your household?  Percent  Number  No  4.28%  52  Yes  95.72%  1164     Total  1216  Table 29. Girl survey item E9. If any friends outside of household (“yes” to E8), how many?  Observations  Mean  Median  SD  Min  Max  1164  3.19  3  1.93  1  15  Self‐Esteem  The GE baseline assessment included a section aimed at measuring girls’ level of self-esteem using the Rosenberg Self Esteem Scale. Girls were asked questions regarding their feelings and attitudes towards themselves, and could respond with either strongly disagree, disagree, agree or strongly agree. Responses were coded on a scale of 1 to 4. For questions where agreement indicates higher self-esteem, disagree strongly was coded as 1, disagree as 2, agree as 3 and strongly agree as 4. For questions where agreement indicates lower self-esteem, the answer choices were reverse-scored. The values for each of the questions included in the scale were then summed to produce a self-esteem scale for each girl. The mean self-esteem score among the girls interviewed in the GE baseline assessment was 31, as was the median (see appendix Table A32). Table 30. Girl survey:  Rosenberg Self‐Esteem Scale (range: 0‐40)  Observations  Mean  Median  SD  Min  Max  1212  31  31  4.28  15  40  32 Aspirational Attitudes  As part of the interview process, girls were asked about their hopes and plans for the future. Most of the girls indicated that they would like to reach a high educational level: 55.92% said they hoped to reach grade 12 (the completion of secondary school) and 28.78% said they hoped to have some level of formal post-secondary education (see appendix Table A33). Table 31. Girl survey item G1: What grade in school would you hope  to complete before leaving school?  Percent  Number  None  0.58%  7  Grade 9 or lower   5.33%  65  Grade 10‐11  8.88   108  Grade 12  55.92%  680  Post‐Secondary Formal  28.78%  350  Don’t know  0.49%  6     Total  1216  Girls were also asked to report the age at which they hoped to get married and the age at which they hoped to have their first baby. The average age at which girls hoped to get married was 25.77 years old (the median was 25) and the average age at which girls hoped to have their first baby was slightly lower, at 25.55 years old (the median was 24). In addition to these questions, girls were also asked whether or not they hoped to get a job outside of the home when they grew older. The majority (82.07%) reported that they did hope to get a job outside of the home (see appendix Table A34). Table 32. Girl survey item G2: At what age would you hope to get married?  Observations  Mean  Median  SD  Min  Max  1158  25.77  25  8.18  3  92  Table 33. Girl survey item G3: At what age would you hope to have your first baby?  Observations  Mean  Median  SD  Min  Max  1154  25.55  24  7.85  12  84  33 Figure 13. Girl survey item G4: Would you hope to have a job  outside of the home even after marriage? 0% 18% Job outside the home (998) No work outside the home (217) Don't know (1) 82% Attitudes Regarding Gender Roles and Violence  The girl survey asked girls a series of questions about their perception of gender roles, gender relations and gender-based violence. One of those questions asked girls to describe what characterizes a good, healthy male-female relationship. Girls could mention as many attributes as they wanted. The most commonly-cited attributes were hardworking (53.13%), financial support (49.26%) and general respect (36.60%). Table 34. Girl survey item J1. Please tell me some things that  describe a good, healthy male‐female relationship   Percent  Number  Hardworking   53.13%   646  Financial support   49.26%   599  General respect  36.60%  446  Faithfulness  21.55%  262  Respectful of partner’s feelings   16.20%   197  Emotional Support   15.79%   192  Listens to partner   14.80%   180  Respectful for partner desires  10.69%  130  Don’t know   3.04%   37  Other   1.07%   13  No Response  0.25%  3  *multiple responses allowed  Respondents:  1,216  In addition to being asked to describe the characteristics of a healthy male-female relationship, girls were also asked a series of questions about their beliefs about such relationships and about gender roles in general. These questions consisted of the Gender Equitable Attitudes subscale within the Gender Relations Scale (Stephenson et al., 2012), and in each, girls were asked to agree or disagree with a given statement. Agree responses were coded as 1 and disagree as 0. 34 The values of each of the questions were then summed to produce a gender relations score for each girl. Two questions, “You don’t talk about sex, you just do it” and “A man should know what his partner likes during sex,” were omitted because of the relative sexual inexperience of girls in the GE survey. The total possible score was therefore 14 (instead of the usual 16). The average score among interviewed girls was 7.52, and the median score was 7. A relatively high percentage of girls gave answers to the gender relations scale questions indicating agreement with unequal roles between men and women: 48.68% of girls, for example, agreed that men need sex more than women, 51.73% agreed that it is a woman’s duty not to get pregnant, 54.69% agreed that women should accept violence to keep the family together and 34.46% agreed that a man can beat his wife if she does not agree to have sex with him. The majority (84.13%) of girls did, however, believe that a husband and wife should agree if they want to have children. The distribution of responses for the remaining questions in the gender relations scale can be found in Table A35 of the appendix. Table 35. Girl survey:  Gender Equitable Attitudes  Scale results (range: 0‐14)  Observations  Mean  Median  SD  Min  Max  1197  7.52  7  3.03  0  14  Relationship and Communication with Adults  The GE program aims to improve outcomes for girls both through direct interventions with girls and through their caregivers. As a result, it is important to assess baseline indicators of girls’ relationships with adults. To do this, the survey asked girls a series of questions from the Monitoring, Conflict, Emotional Support and Financial Support (MCEF) scales (Bingenheimer et al, 2014). The first two questions asked the girl whether or not there was an adult in her life who knew where she was during the day and at night (response options were yes or no). If the girl answered yes to either of those questions, she was then asked to identify the person. In the remaining scale questions, girls were read a statement about the adults in their lives and were asked to respond with: not true at all, somewhat true or very true. In two of the questions, the girl was also asked to identify who the adult was that fit the description in the statement. For questions with the not true at all/somewhat true/very true response options, not true at all was coded as 1, somewhat true was coded as 2 and very true was coded as 3. For the questions that had yes/no response options, yes was coded as 3 and no as 1. To construct the final composite scale, the values for all of these variables were summed. Sub- sets of the questions were also summed to create subscales for the monitoring, conflict, emotional support and financial support components of the composite scale. The average composite MCEF score for girls interviewed was 24.75 with a median score of 25. Average and median values for the subscales are given in Table A36 in the appendix. The majority of girls reported that they had an adult in their lives who knew where they were during the night (79.85%) and day (89.47%) (see appendix Table A37 and Table A38). The majority also indicated that it was “very true” that they had an adult in their lives who provided for their daily necessities (87.91%) and gave them money (71.13%). In each case, girls most 35 commonly reported that their mother served in these roles. In each case, the next most commonly-identified adult in these roles was their father (see appendix Table A39). Table 36. Girl survey item D13: There is a big person in your  life who gives you money (MCEF component)  Percent  Number  Very true   71.13%   865  Not true at all   14.47%   176  Somewhat true   14.31%   174  Don't Know  0.08%  1     Total  1216  D14. If “yes” to D13, who usually gives you money?   Percent  Number  Mother  55.44%  576  Father  51.49%  535  Brother   20.31%   211  Sister   18.09%   188  Uncle   13.86%   144  Aunt  12.90%  134  Boyfriend   3.56%   37  Other adult   2.12%   22  Play mom  0.10%  1  Husband   0.10%   1  Don't Know  0.00%  0  No Response  0.00%  0  *multiple responses allowed  Respondents:  1,039  The distribution of responses for the remaining MCEF scale questions are listed in appendix Table A40. The proportion of girls who responded “very true” was highest for the questions that reflected positive adult-child relationship qualities: 83.8% of girls, for example, indicated that it was “very true” that they had an adult in their lives who praised them. The percentage of “very true” responses, however, was somewhat lower for questions that reflected a negative adult-child relationship. 40.54% of girls, for example, indicated that it was “very true” that there was an adult in their lives who made them look small (put them down). In addition to the questions used to construct the MCEF scale, girls were also asked whether or not they had a mentor in their life and whether or not they had an adult they would ask to take them to important events: 95.64% of girls reported that they did have a mentor. Among those girls who had a mentor, the mentor was most commonly their mother (39.86%) or their father (14.93%). The most common topics girls reported discussing with their mentor were plans for the future, school lessons and health problems (see appendix Table A41). Almost all of the girls interviewed (99.67%) reported that they had an adult in their life who they would ask to escort them to important events. As with the mentor, the person girls most often identified in this role was their mother (50.00%) or their father (20.21%). As with the mentor question, the most common topics girls reported discussing with this person were plans for the future, health problems and school lessons (see appendix Table A42). This question was intended to help 36 identify the adult the girl would be most likely to ask to attend caregiver meetings if she were part of the GE program. Whenever possible, the person identified in this question was interviewed during the caregiver portion of the GE baseline assessment. Table 37. Girl survey item D15: Is there a big person in your life  who you regard as your mentor? By mentor, I mean a person that  respects you, listens to you and advises you to become a good  person.     Percent  Number  Yes  95.64%  1163  No  4.36%  53     Total  1216              D16. If “yes” to D15, who is that person?  Percent  Number  Mother  39.86%  462  Father  14.93%  173  Aunt  10.79%  125  Sister  9.58%  111  Grandmother  8.80%  102  Brother  6.90%  80  Uncle  5.18%  60  Other adult  2.59%  30  Grandfather  1.12%  13  Boyfriend  0.17%  2  Pastor  0.09%  1     Total  1159  37 Table 38. Girl survey item D18. Is there a big person in your life  who you would ask to escort or carry you to important events, like  to go see a doctor, a meeting with a teacher or a meeting about  helping girls like you?       Percent  Number  Yes  99.67%  1212  No  0.33%  4     Total  1216                   D19. If “yes” to D18, who is that person?  Percent  Number  Mother  50.00%  606  Father  20.21%  245  Aunt   9.41%   114  Sister  5.78%  70  Brother  2.81%  34  Uncle  2.64%  32  Other adult   1.57%  19  Boyfriend   0.08%  1     Total  1212  Mental Health   In order to evaluate girls’ overall psychological wellbeing, the assessment asked a series of questions about girls’ moods and feelings. The questions were taken from the Short Mood and Feelings Questionnaire (SMFQ) scale (Angold et al., 1995). For each of the 13 questions, girls were read a statement and asked to respond with: not true, sometimes, or true. “Not true” responses were coded as 0, “sometimes” responses were coded as 1 and “true” responses were coded as 2. The values for each of the questions were then summed to produce a Short Mood and Feelings Questionnaire score for each girl. The average SMFQ score among girls interviewed was 8.5 and the median was 8. Among the scale questions, the statement with the highest rate of “true” responses (33.47%) was “In the past two weeks, I was very restless”. This was followed by “In the past two weeks, I felt miserable or unhappy” (29.85%) and “In the past two weeks, I didn’t enjoy anything at all” (29.03%). The distribution of responses to all questions in the scale is listed in Table A43 in the appendix. Table 39. Girls’ Mood and Feelings Questionnaire (SMFQ) Scale (range: 0‐26)  Observations  Mean  Median  SD  Min  Max  1214  8.5  8  6.58  0  26  38 Financial Literacy  Part of the GE program is aimed at increasing participants’ financial literacy so that they can make better long-term choices. In order to gauge girls’ level of financial literacy, the assessment asked a series of questions related to financial management knowledge and practices. When asked to identify from a list all the items that could be called something you need, the most common choices were food for your family (88.90%) or an extra pair of shoes (69.33%). Asked to identify from a list all of the items that could be something you need instead of just something you want, the most commonly selected response was food for the family (72.29%). Nearly one third of girls (30.18%), however, also indicated that a soft drink was a need rather than just a want (see appendix Table A44). When asked to select from a list all of the reasons one might save money, the most commonly- cited reason was “To have money to meet an emergency (like an accident or illness)” (71.88%) (see Table A45 in appendix). Approximately two thirds of girls (67.85%) reported that they had ever saved money for something, however, only about one third (32.97%) of the girls who answered affirmatively indicated that they were currently saving for something special. When asked to report all the things a girl can do to save money, the majority (88.24%) cited doing extra work for pay. The second most common response (37.17%) was to ask someone for money. The most common locations they reported that girls could save money were at home in a cash box (90.38%) or giving it to someone trusted (43.83%); few girls had access to secure methods of saving money. Table 40. Girls Saving Money  Response options  Girl Survey Questions  No  Yes  Total  I4. Have you ever put money aside (saved) money for  391  825  1216  something before?  (32.15%)  (67.85%)  I5. If “yes” to I4, are you putting money aside (saving)  553  272  825  right now for something special?  (67.03%)  (32.97%)  Table 41. Girl survey item I6.   What things can a girl do to save more money?   Percent  Number  Do extra work for pay  88.24%  1073   Ask someone for money   37.17%   452   Sell sex for money   15.13%   184   Other   11.68%   142   Spend less on needs   7.81%   95   Reduce unnecessary spending   7.73%   94  Don't Know  0.41%  5  No Response  0.16%  2  *multiple responses allowed  Respondents:  1,216  39 Table 42. Girl survey item I7. What are some different  places where girls can put money aside (save)?   Percent  Number  At home, cash box  90.38%  1099  Give it to someone you trust   43.83%   533  Bank, credit union, cooperative or microfinance institution  10.94%  133  Savings group   7.40%   90  Through a cell phone account   1.15%   14  Others  0.74%  9  Don't Know  0.08%  1  No Response  0.00%  0  *multiple responses allowed  Respondents:  1,216  Only a small minority of girls (4.36%) had ever kept a record of the spending they did. Among those who had kept such a record, the majority (81.13%) only kept it some of the time. Table 43. Spending Records  Girl survey item I8.   Have you ever written down the spending you do?  Percent  Number  No  95.64%  1163  Yes  4.36%  53     Total  1216       I9. How often do you do this?  Percent  Number  Rarely  7.55%  4  Some of the time  81.13%  43  Most of the time  3.77%  2  Always  7.55%  4     Total  53  When asked to list all of the sources of the money they spent over the past year (appendix Table A46), the majority of girls obtained money from their mothers and fathers (56.09% and 43.83%, respectively, as multiple responses were permitted. Roughly 12% obtained the money from another guardian with whom they live, and roughly 20% reported obtaining it from another relative. Very few indicated spending money earned from a small job or savings, and roughly 13 percent indicated that they simply do not purchase the item(s) they wish to buy. Nearly 65% indicated that they do not own anything that helps them earn money, while 22.04% indicated owning a basket or bucket and 11.84% indicated owning a garden or farm (see appendix Table A47). 40 Health Literacy  During the survey, girls were asked several questions about their knowledge of health and physical maturation. When asked to indicate what menstruation is, nearly one-third (29.69%) reported that they did not know. More than half (60.61%), however, reported that menstruation was something to be ashamed of. Asked to list all of the physical changes that happen as a girl grows in to a woman (see appendix Table A48), the most commonly cited changes were the development of breasts (92.02%), growing taller (68.75%), growing hair in new areas of the body (47.86%) and menstruation (41.28%). In addition to these questions about physical development, girls were also asked what the side-effects could be of participating in rituals where a girl’s body is cut (female genital mutilation). More than half of girls (60.61%) reported that they did not know. The next most common response was getting an infection from the procedure (24.67%) and trouble in childbirth later (13.24%) (see Table A49 in appendix). It should be noted, however, that ritual practices in Liberia related to female genital mutilation are often secretive. As a consequence, the high rate of “don’t know” responses to this question may in part be due to girls’ reluctance to discuss the topic. Table 44. Girl survey item K1:   Can you tell me what menstruation, or your period is?   Percent  Number  It is when you “receive” (monthly female bleeding)  40.63%  494  It happens once a month  19.33%  235  It is when blood comes out of the vagina  47.04%  572  It usually lasts about 3 to 7 days  21.79%  265  Don't Know  29.69%  361  No Response  0.41%  5  *multiple responses allowed  Respondents:  1,216  Figure 14. Girl survey item K2: Do you think your monthly period is something to be ashamed of? 13% 26% No (323) Yes (737) Don't know (154) 61% 41 Table 45. Girl survey item K4: Do you know of any  health problems that can result for a girl after  participating in the ritual where a girls’ body is cut  (Sande society)?  Percent  Number  Don't Know  60.61%  737  Infection from the procedure  24.67%  300  Trouble in childbirth later  13.24%  161  Risk for sexually transmitted infections or HIV  6.09%  74  Other problem  5.92%  72  No Response  4.11%  50  *multiple responses allowed  Respondents:  1,216  Knowledge about Family Planning Methods  As part of the focus on knowledge about sexual health and pregnancy, the survey asked girls several questions about family planning methods. Girls were first asked to list all of the methods of contraception that they had heard of. Two-thirds (67.02%) had heard of the oral contraceptive pill, more than half (53.45%) had heard of hormonal injection and nearly one-third (32.32%) had heard of the hormonal implant. When girls were then asked which method was best for young people, 41.83% chose the oral contraceptive pill, 29.91% chose the hormonal injection and 14.22% chose the hormonal implant. When girls were subsequently asked to list all of the family planning methods they had used in the past 12 months, 96.06% indicated that they had not used any family planning methods (this included girls who had not had sex in the past 12 months), 1.47% reported using a hormonal injection, 1.01% reported using a male condom, and 0.99% reported using a hormonal implant (see appendix Table A50 for full list of items, and table below for abridged list). Table 46. Girl survey item L1: Can you tell me some methods of  contraception or preventing pregnancy that have you heard of?   Percent  Number  Oral contraceptive pill  67.02%  815  Hormonal injection (i.e. Depo Provera, Nur Isterate)  53.45%  650  Hormonal Implant (Implanon or Nexplanon)   32.32%   393  Male condom   17.02%   207  Don’t Know   9.95%  121  Hormonal implant (Implanon or Nexplanon)   4.85%   59  Female condom  3.21%  39  Diaphragm  1.48%  18  Periodic abstinence / rhythm (a couple can avoid sex on days when  pregnancy is most likely to occur)   0.74%   9  No Response   0.74%   9  Other   0.66%   8  Non‐vaginal sex (Anal or oral sex, thigh sex)  0.33%  4  *multiple responses allowed  Respondents:  1,216  42 Table 47. Girl survey item L2: Which one of the ways do you think  is best for young people to use?  Percent  Number  Oral contraceptive pill  41.83%  456  Hormonal injection (i.e. Depo Provera, Nur Isterate)  29.91%  326  Hormonal implant  (Implanon or  Nexplanon)  14.22%  155  Male condom  9.36%  102  Don't Know  1.47%  16  Intra‐uterine device (IUD or Loop)  1.10%  12  Other  0.64%  7  Periodic Abstinence/Rhythm  (a couple can avoid sex on days when  pregnancy is most likely to occur)  0.37%  4  Diaphragm  0.46%  5  Female condom  0.28%  3  No Response  0.28%  3  Non vaginal sex (anal or oral sex, thigh sex)  0.09%  1     Total  1090  Table 48. Girl survey Item L3: I am now going to read out a list of  contraceptive/family planning methods.  Which of these have you  used in the past 12 months?   Percent  Number  Has not used any  96.06%  1047  Hormonal injection (i.e. Depo Provera, Nur Isterate)  1.47%  16  Male condom  1.01%  11  Oral Contraceptive Pill  0.99%  12  Don't Know  0.92%  10  Other  0.28%  3  *multiple responses allowed  Respondents:  1,090  Knowledge about Condoms  As additional measures of knowledge about sexual health, the assessment presented girls with a series of questions about condoms. Girls were then asked to agree or disagree with each statement: 62.17% of girls agreed that condoms are a good way of preventing pregnancy; 74.34% agreed that condoms are a good way of protecting against HIV/AIDS and 73.68% agreed that condoms are a good way of protecting against sexually transmitted infections. In addition, the majority of girls disagreed with both the statement that condoms can slip off the man and get lost inside the woman’s body (60.44% disagreed) and the statement that condoms can be used more than once (63.9% disagreed). For each of these questions, between 6% and 12% of girls responded “don’t know”, which was a higher rate than in most other sections of the survey (see appendix Table A51). 43 Knowledge about HIV/AIDS and Sexually Transmitted Infections  During the baseline assessment interview, girls were asked specific questions about their knowledge of HIV/AIDS and other sexually transmitted infections. The majority of girls (82.15%) reported that they had heard of HIV/AIDS. Among girls who had heard of HIV/AIDS, one-third (33.33%) believed it was possible to cure HIV/AIDS. Just under half (49.55%), however, believed that people living with HIV/AIDS can manage and survive.  Table 49. Knowledge of HIV/AIDS  Response options  Question  No  Yes  DK  Total  192  999  25  1216  M1. Have you heard of HIV or AIDS?  (15.79%)  (82.15%)  (2.06%)  664  333  2  999  M2. Do you think it is possible to cure HIV/AIDS?  (66.47%)  (33.33%)  (0.2%)  M3. Do you think it is possible that people with HIV/AIDS  500  495  4  999  can manage and survive?  (50.05%)  (49.55%)  (0.4%)  When asked to list all of the signs and symptoms of sexually-transmitted infections in women (see appendix Table A52), the most common responses were pain during urination (48.52%) and discharge from the vagina (32.81%). In addition, 30.92% of respondents reported that they did not know what any of the signs and symptoms were. Caregiver Survey  Attitudes about Girls: Gender Role Norms  The GE program aims to assist girls both directly through mentorship meetings and indirectly through their caretakers. In order to gauge caretaker gender norm attitudes at baseline, the assessment asked caretakers a series of questions related to the topic. The questions were adapted from the Gender Norm Attitudes Scale (Waszak et al., 2000). For each question, caretakers were read a statement and asked whether they agreed or disagreed with it. For questions where agree indicated a more egalitarian attitude, agree was scored as 2 and disagree as 1. For questions where agree indicated a less egalitarian attitude, agree was scored as 1 and disagree as 2. The scores for each of the questions were then summed to create a composite Gender Norm Attitude Scale score for each caregiver. The scores for sub-sets of questions were then summed separately to create Rights and Privileges of Men, and Equity of Girls subscales. The average Composite Gender Norm Attitudes Scale score was 18.87 and the median was 19. For all questions except one, the majority of caregivers selected the option that was scored as more egalitarian: 89.95% of caregivers, for example, disagreed with the assertion that it is better 44 boys to go to school than girls, and 72.1% of caregivers agreed with the assertion that daughters should have just the same chance to work outside the home as sons. The majority of caregivers (68.37%) did, however, agree with the statement that every woman needs a man to protect her because she cannot protect herself (see Table A53 in appendix). Table 50. Composite Caregiver Gender Norm Attitudes Scale (range: 11‐22)  Observations  Mean  Median  SD  Min  Max  1145  18.87  19  2.2  12.00  22.00  Caregiver Rights and Privileges of Men Sub‐Scale (range: 8‐16)  Observations  Mean  Median  SD  Min  Max  1145  13.61       14.00       2.07      8.00      16.00  Caregiver Equality of Girls Sub‐Scale (range: 3‐6)  Observations  Mean  Median  SD  Min  Max  1154  5.26  5.00  0.85  3.00  6.00  Attitudes about Girls: Child Rearing  The monthly caregiver meetings that will take place as part of the GE program will focus on parenting skills. As a result, it was important to gauge caregiver attitudes towards parenting at baseline. To do this, the caregiver portion of the assessment asked caregivers a series of questions about their views on raising their daughters. Nearly half (44.63%) of caregivers believed that to raise girls properly, parents need to physically punish them (graph below and appendix Table A54) and 91.59% agreed that it is important to know where their 13-14 year-old is at all times of day and night. Figure 15. Caregiver survey item D1. Do you believe that in order to  bring up girls correctly, you need to physically punish them? 45% No (639) 55% Yes (515) 45 Table 51. Caregiver survey item D2: How true is this  statement? It is important that you know where a 13‐ 14‐year‐old girl can be at all times of the day or night?  Percent  Number  Agree  91.59%  1057  Disagree  8.41%  97     Total  1154  In addition, caregivers had relatively high aspirations for the level of education they hoped their daughters would reach before leaving school. The majority of caregivers (59.10%) reported that they hoped their daughters would reach some level of post-secondary education, and 38.21% reported that they hoped their daughters would complete the 12th grade. The fact that caregivers wanted their daughters to reach a relatively high level of education is consistent with girls’ own aspirations: 97.1% of caregivers hoped their daughters would reach 12th grade or post-secondary schooling and 84.7% of girls reported the same. Caretaker preference for their daughters to reach a post-secondary level of education, however, was significantly higher than the girls themselves reported (59.10% of caretakers vs. 28.78% of girls). Table 52. Caregiver survey item D3: What grade do you  want the girl(s) to finish before leaving school?  Percent  Number  None  0.09%  1  Grade 6  0.09%  1  Grade 9  0.26%  3  Grade 10  1.21%  14  Grade 11  0.78%  9  Grade 12  38.21%  441  Post‐Secondary formal  59.10%  682  Don’t know  0.26%  3     Total  1154  Just as with the girls, caretakers were also asked at what age they hoped their girl would get married and have her first baby. The average age caretakers reported hoping their girl would get married at was 26.06 years old (with a median age of 25) and the average age caretaker reported hoping their girl would have her first baby was 25.87 years old (with a median age of 25). These ages were relatively consistent with the average ages the girls themselves reported: 25.77 years old for getting married and 25.55 years old for having a first baby. In addition to the questions about marriage and having babies, caretakers were also asked whether or not they wanted their girls to get a job outside of the home after marriage. Nearly all (92.98%) of the caregivers reported that they would like their daughters to get a job outside of the home after marriage, consistent with girls’ reports. Table 53. Caregiver survey item D4. At what age do you want the girl(s) to get married?  Observations  Mean  Median  SD  Min  Max  1117  26.06  25  5.46  13  55  46 Table 54. Caregiver survey item D5: At what age do you want the girl(s) to have her/their  first baby?  Observations  Mean  Median  SD  Min  Max  1118  25.87  25  5.53  14  56  Table 55. Caregiver survey item D6: Do you want the  girl(s) to have a job after marriage or to stay at home?  Percent  Number  Job outside the home  92.98%  1073  No work outside the home  7.02%  81     Total  1154  Discussion  In this section we examine the results from the GE baseline survey and put them into the context of existing literature. Although we used well-known, validated measures and scales, it was difficult across the board to find results to serve as comparators. This was the case for several reasons: first, GE respondents are much younger than most respondents in studies that assess similar outcomes; second, for many of the scales, studies that have employed them seldom report the outcome score for the scale itself - the score is instead often used as a covariate for predicting some other outcome of interest. A primary finding is that although the percentage of GE girls who had sexually debuted as of the baseline survey was not noticeably higher for their age than available comparable sources for the region indicate, their levels of experiencing sexual abuse appear to in fact be quite high. Details of these key outcomes, and others in the survey, are discussed below. SEXUAL DEBUT  The level of sexual initiation among GE respondents was 20.72 percent. Within the UNICEF Violence against Children Surveys (VACS) from Sub-Saharan Africa – the program of surveys most comparable to GE with regard to age group and sexual violence survey question style and content - the percentage of respondents having sexually debuted was usually not reported. In Swaziland, however, 17.1 percent of 13-17-year-olds girls had debuted. In the DHS surveys the closest comparable reported outcome is the percent of 15-19-year-olds who had sex before age 15 years. In Liberia the level was 22 percent in 2013, and in West and Central Africa the regional average was 11.5 percent (McCarthy et al., 2016). FIRST SEX UNWILLING  Among GE respondents who have experienced penetrative sex, 29 percent report the first act as having been unwilling (15 percent reported being tricked; 14 percent reported being physically 47 forced). Analogous data on the unwilling nature of first sex are rare for 13-14-year-olds, but the VACS provide some relevant information. For the VACS undertaken in Sub-Saharan Africa, among all respondents (aged 13-24 years) who had their sexual debut before age 18, the experience was described as unwilling by 53 percent, 29 percent, 25 percent, 41 percent in Swaziland, Tanzania, Kenya and Zimbabwe, respectively. The GE mean is within this range. Data on the prevalence of first sex acts that were unwilling among 13-17-year-olds who have sexually debuted was reported only for Swaziland and Zimbabwe, where the levels were 63 percent and 43 percent, respectively.4 EVER EXPERIENCED ANY SEXUAL ABUSE   Ever experience of sexual abuse is defined as a “yes” response to any of the four types of experiences: physically forced sex, coerced/persuaded into having sex, attempted unwanted sex, or being sexually touched. Among all GE respondents, 37.3 percent have experienced one or more of these. Available figures to compare this outcome to are difficult to find, even in the VACS surveys from the continent, as reporting varies by country and often focuses on an older age group. The only country for which there is a similar age group reported on is Swaziland, where 28 percent of 13-17-year-old girls had ever experienced sexual abuse. The remaining available estimates from VACS for this outcome are defined as any experience before age 18, among 18-24-year-olds. In Swaziland, Kenya and Zimbabwe, respectively, the levels are 37.8, 31.9, and 32.5 percent. In Tanzania, the level for 13-24-year-olds was 27.9 percent. The level in GE is within the range of these figures but at the high end, particularly given the relatively young age of GE respondents compared with those reported by the VACS. Table 56. GE and VACS prevalence of ever experienced sexual abuse  AMONG ALL SURVEY RESPONDENTS  Percentage  Ever experienced    GE Liberia  (13‐14‐year‐olds)  37  Swaziland VACS  (13‐17‐year‐olds)  28  Tanzania VACS  (13‐24‐year‐olds)  28  Ever experienced before age 18 years    Swaziland VACS   (18‐24‐year‐olds)  38  Kenya VACS  (18‐24‐year‐olds)  32  Zimbabwe VACS  (18‐24‐year‐olds)  33  Retrospective questions are available in the DHS for 12 Sub-Saharan African countries on the percentage of adolescent girls (aged 15–19) who were ever subjected to sexual abuse (UNICEF 2014). Among the countries with these data, ever experience of sexual abuse among this age group ranges between 8 percent (Mozambique) to 22 percent (Cameroon). In the Liberia 2007 4 Sexual abuse questions in DHS data differ greatly from those in the GE survey and the VACS. In the Liberia 2007 DHS, 9.9 percent of respondents age 15-49 who had had sex reported “physically forced” to the following question: “The first time you did men business, would you say you did it because you wanted to or because you were physically forced to do it against your will?” The DHS report did not provide figures for 15-19-year-olds. This outcome was not included in the 2013 Liberia DHS survey report. 48 DHS, 13.1 percent of 15-19-year-olds report having ever experienced sexual abuse. This outcome was not described in the 2013 Liberia DHS report. It should also be noted that the DHS question is more limited in scope compared to those included in the GE survey and focuses mainly on physical forced sex. (DHS for Liberia: “At any time in your life, as a child or as an adult, has anyone physically forced you in any way to do men business or perform any other sexual acts?”) NATURE OF SEXUAL ABUSE EVER EXPERIENCED  Among those who have ever experienced any type of sexual abuse, GE respondents reported the following four possible types: physically forced, non-physically pressured (coerced/persuaded), someone unsuccessfully attempting to have sex with them, or being touched in a sexual way. The prevalence of each component of sexual abuse among the GE respondents appears to be quite high for this age group compared with the levels reported in the VACS from Sub-Saharan Africa. Table 57. GE and VACS prevalence of types of sexual abuse ever experienced (%)  AMONG ALL SURVEY  Forced  Coerced or  Attempted  Sexual touched  RESPONDENTS  persuaded  unwanted  Ever experienced          GE Liberia   7.8  8.4  24.7  28.9  (13‐14‐year‐olds)  Swaziland VACS  2.3  5.7  16.8  nr  (13‐17‐year‐olds)  Ever experienced before age          18 years  Swaziland VACS  7.2  12.1  20.5  14.1  (13‐24‐year‐olds)  Tanzania VACS  5.5  3.1  14.6  16.0  (13‐24‐year‐olds)  Kenya VACS  7.1  8.3  15.3  20.7  (18‐24‐year‐olds)  Zimbabwe VACS  9.0  7.4  15.0  20.2  (18‐24‐year‐olds)    EXPERIENCE OF ANY SEXUAL ABUSE IN PAST 12 MONTHS  Given the young age and recent sexual debut of the GE respondents, the level of sexual abuse in the past 12 months, 32.7 percent, is nearly equal to the ever experienced sexual abuse outcome. Contrasting the GE results with those from the VACS provides further perspective on girls’ recent sexual abuse experiences. Among 13-17-year-old girls in Swaziland, Tanzania, Kenya and Zimbabwe, respectively, 28, 14, 11 and 9 percent, respectively, had experienced any unwanted sexual experience in the past 12 months. The GE prevalence is very high in comparison, particularly given that GE girls are younger than the age group reported on in the VACS surveys. 49 Table 58. GE and VACS prevalence of any sexual abuse sex in past 12 months  AMONG ALL SURVEY RESPONDENTS  Percentage  GE Liberia  (13‐14‐year‐olds)  33  Swaziland VACS  (13‐17‐year‐olds)  28  Tanzania VACS  (13‐17‐year‐olds)  14  Kenya VACS  (13‐17‐year‐olds)  11  Zimbabwe VACS  (13‐17‐year‐olds)  9  In the DHS (UNICEF, 2014), prevalence of “forced sex” among 15-19-year-olds in the past 12 months ranges from 3 percent (Gabon) to 10 percent (DRC). In the Liberia 2007 DHS, 5 percent of 15-19-year-olds report it happening in the past 12 months. This outcome was not described in the 2013 Liberia DHS report. These levels are much lower than that found in GE, but it should be noted that the DHS question was much more narrowly focused on physical force. (“In the last 12 months has anyone forced you to do men business against your will?”) NATURE OF SEXUAL ABUSE EXPERIENCED IN PAST 12 MONTHS  As many of the sexual experiences reported by the 13-14-year-old GE respondents in the 12 months before the survey are likely to be among their earlier ones, the 12 month recall and “ever” prevalence rates for sexual abuse are similar. Among all girls, in the past 12 months, 7 percent had been physically forced to have sex, 8 percent had been non-physically pressured, 21 percent had sex someone unsuccessfully attempt to have sex with them, and 25 percent had been touched in a sexual way. The closest analogous outcomes to compare with are the VACS reports for 13-17-year-old girls in Kenya and Zimbabwe, where the levels are much lower than those reported in GE. Table 59. GE and VACS prevalence of types of sexual abuse experienced in past 12 months (%)  AMONG ALL RESPONDENTS  Forced  Coerced or  Attempted  Sexual  persuaded  unwanted  touched  GE Liberia   6.6  7.8  21.1  25.1  (13‐14‐year‐olds)  Kenya VACS  0.0  1.1  3.3  8.5  (13‐17‐year‐olds)  Zimbabwe VACS  0.9  1.9  3.3  5.4  (13‐17‐year‐olds)    PERSPECTIVES ON OTHER KEY OUTCOMES   School enrollment  Most GE girls (85.03 percent) report having been enrolled during the 2014/2015 school year (the academic year before the survey). This level is not atypical for the setting. DHS data indicate that 50 school attendance for females aged 10-14 years was 87 percent for Liberia (2013), and 71 percent regionally for West and Central Africa (McCarthy et al., 2016). Parental co‐residence and survival  Among GE respondents, 74.59 percent reside with mothers and 58.63 percent reside with fathers. The intersection of these outcomes indicates that approximately half reside with both parents, one-fifth with mother only, 5 percent with father only, and one-fifth with neither. A Population Council (2009) special report on adolescent outcomes in the Liberia 2007 DHS, shows a higher percentage of girls aged 10-14 residing with father only or neither parent relative to GE girls. Parental survival in the GE survey does not appear to be significantly different from that reported for 10-14-year-old girls in the Liberia 2007 DHS (Population Council, 2009). Table 60. Parental co‐residence and survival in Liberia    Parental co‐residence with girl  Parental survival    GE baseline  NorthCentral  GE baseline  NorthCentral  survey  Region (DHS 2007,  survey  Region (DHS 2007,  10‐14‐year‐olds)  10‐14‐year‐olds)  Mother and father  53  44  86  89  Mother only  21  16  10  5  Father only   5  15  3  4  Neither  20  25  1  2    Not enrolled in school and not living with either parent  Population Council cross-country analysis of DHS data (McCarthy et al., 2016) examines a category of potential vulnerability among 10-14-year-olds: girls not enrolled in school and not residing with either parent. In the GE survey this was the case for 4 percent of girls. In the Liberia 2013 DHS the rate nationally was also 4 percent. In the West and Africa Central region the rate was 7 percent. Pregnancy  Pregnancy is another outcome on which precisely comparable data are not readily available for 13-14-year-olds. DHS results are available, however, for the percentage of women aged 15-19 who gave birth before age 15 years. The figure is 3.0 percent in Liberia nationally and 3.3 percent for the region of West and Central Africa (McCarthy et al., 2016). In the GE survey, 6 of 1,216 girls (0.5 percent) report having had a live birth, 20 of 1,216 (1.7 percent) report having ever been pregnant – percentages that are lower than the regional and national averages for 15- 19-year-olds in the DHS. Marriage  Less than 1.0 percent of GE girls reported they were married or living with a man as if married. The DHS figures for females age 15-19 years show that 9.1 percent in West and Central Africa, 51 and 4.0 percent in Liberia (2013), were married before age 15 (McCarthy et al., 2016). Twenty percent of GE respondents reported having a non-co-resident boyfriend. Age‐disparate relationships  Of the 21% of GE girls who report having a husband or boyfriend, only 141 of 258 reported valid information to calculate partner age difference (most did not know their partner’s age). Among those who reported this information, the median husband age was 18 years; the median boyfriend age was 17 years. Sixteen girls reported a partner 5 or more years older; two reported a partner 10 or more years older. By contrast, Population Council analysis of the Liberia 2007 DHS (Population Council, 2009) indicates that among 15-24-year-old females residing with a sexual partner, 30 percent were living with a man who was 10 or more years their senior. Clearly GE girls will be at risk for such inter-generational partnerships in the two years between baseline and endline. Although the GE baseline survey did not ask the age of girls’ first or most recent sexual partner, this should be included in the GE endline survey. In the Liberia 2007 DHS, 38.4 percent of sexually debuted 15-24-year-old females report their first sex was with a partner 10 or more years older than themselves. In the Liberia 2013 DHS, the only partner age differences that were assessed were for 15-19-year-olds who had “higher risk” sex in the 12 months before the survey (defined as sex with someone they were not residing with as a partner). This is not a good comparator to the GE statistics; nonetheless, 10.7 percent of these females reported having a partner 10 or more years older; in the Liberia 2007 DHS the percentage was 13.7. HIV knowledge   HIV knowledge among girls in the GE survey is very low; only 82 percent had ever heard of HIV or AIDS. Of these, one-third believed that it was possible to cure AIDS – a shockingly high percentage for the region. While the GE baseline survey did not include each element of the comprehensive HIV knowledge question set (the endline survey should include these and be compared with levels in UNICEF’s State of the World’s Children most recent annual report), only 74 percent of GE girls believed that condoms were an effective way of protecting against HIV. (A similar percentage believed that condoms could protect against STIs.) Only half believed a person with HIV/AIDS can manage and survive. HIV knowledge, attitude and behavior data are not readily available for very young adolescents (defined as 10-14-year-olds) in Sub-Saharan African. In the 2013 Liberia DHS, however, 72 percent of females aged 15-24 years believed a healthy looking person can have HIV. Condom use   Condom use is extraordinarily low among girls in the GE survey who reported having sex in the past 12 months: only 31 percent say they used a condom at any time in the past year and only 8 percent said they always did. Condom use at last sex was lower still, at only 13 percent of respondents. Sexual behavior self-reports are almost non-existent among 10-14-year-olds in LMICs so putting these results in context is difficult. The DHS, however, collects data starting at 52 age 15 years. In the Liberia 2013 DHS, among never married women aged 15-19 years who had had sex in the past 12 months, 24 percent report having used a condom at last sex. Among 15-24- year-old females nationally, rural residents have condom use rates half those of urban residents (12 versus 26 percent). Across Sub-Saharan Africa, the rate of condom use at last sex among 15- 19-year-old females is 35 percent (Santhya and Jejeebhoy, 2015). Low condom utilization rates among sexually experienced respondents is a particular concern, especially in light of the potential long-term health questions surrounding the Ebola epidemic. Gender norms   The Gender Equitable Attitudes Subscale (derived from Pulerwitz and Barker, 2008) of the Gender Relations Scale (Nanda, 2011; Stephenson et al., 2012) was included to assess girls’ gender equitable norms. Two questions, “You don’t talk about sex, you just do it” and “A man should know what his partner likes during sex” were omitted because of the relative sexual inexperience of girls in the GE baseline survey. The total possible score was therefore 14 (instead of the usual 16). GE girls’ gender equity attitudes were relatively progressive, with a mean of 7.5. This scale is fairly new and has not been applied extensively with adolescents. The original application of the scale by Stephenson and colleagues (2012) among adults in Ethiopia and Kenya revealed that out of a possible 16, women’s scores were 6.0 and 10.1, respectively; men’s scores were 9.0 and 13.0, respectively. Rosenberg Self‐Esteem Scale (RSES)  In the GE sample, the mean and median score on the Rosenberg Self-Esteem Scale RSES was 32 (SD=4.28), which is considered within the normal range (https://www.wwnorton.com/college/psych/psychsci/media/rosenberg.htm) and somewhat higher than found in primarily college student samples in 5 African countries (Schmidt and Allick, 2005): Botswana, DRC, Ethiopia, Tanzania, Zimbabwe, and the United States (for comparison) (see appendix Table A55). The higher mean score in our sample of Liberian adolescent girls age 13-14 than in the other African samples could be age related, however, as self-esteem scores among girls tend to decrease as age increases during adolescence. A study among Irish girls found a drop in mean scores between the ages 10-12 (mean=28, SD=4.3) and 13-17 (Age=13-14: mean=27.1, SD=3.5; age=15-17: mean=27.1, SD=3.3) (Nic Gabhainn and Mullan, 2003). In a study looking at the RSES in 53 different countries, the factor analysis found that question #8 (“I wish I could have more respect for myself.”) had unexpected negative loadings for the DRC and Tanzania. Thus, looking at the Cronbach’s alpha, if items were to be deleted it might be worth exploring before the endline survey to see if this (or any other) items should be excluded in the Liberian sample (Schmidt and Allick, 2005). Short Mood and Feelings Questionnaire (SMFQ): Assessment for Depression  The average SMFQ score among the GE survey girls was 8.5 and the median was 8.0 (SD=6.58). There is no consensus about the cut-off for classifying individuals as depressed and there is little information about the use of this measure among adolescent girls in Africa. A study among a racially diverse group of 6th graders (boys and girls, mean age of 11.5 years) in the USA found that the mean score on the SMFQ was 3.8 among non-depressed children and 8.2 among children 53 diagnosed with depression. A cut off of 4 was assessed for diagnosis of depression and found to have fairly low sensitivity and specificity of 66% and 61% respectively (Rhew et al., 2010). However, another study among male and female youth in the juvenile justice system in the USA used the SMFQ and found it to be reliable (alpha=0.87) and that a cut off of >=10 for depression maximized sensitivity and specificity at 100% and 72% respectively (Kuo et al., 2005). When assessed in a sample of male and female 8th graders in South Africa, the SMFQ was found to be reliable (alpha=0.853 overall), with little variation in reliability across major racial groups (black, white, colored). The test-retest reliability was assessed with a kappa of agreement and scores were mostly in the fair range (25%-49%) (Rothon et al., 2011). A study among adolescents (mean age 17 years) in Norway found that girls scored significantly higher than boys (mean for girls=7.4, SD=6.1; mean for boys=4.1, SD=4,9, p<0.001) (Lundervold et al., 2013). Thus while the girls in our sample exhibited a higher mean score on the SMFQ than that in other western populations, the gender make-up may partially explain this higher score and warrant using the higher cut-of (e.g. >=10) for a depression classification. Table 61. SMFQ comparison scores  Mean  Depression cutoff  score  GE Liberian females (13‐14 years)  8.5    US 6th graders (depressed)  8.2    US 6th graders (non‐depressed)  3.8    US juvenile justice youths    ≥10  Norwegian girls (mean age 17 years)  7.4    Norwegian boys (mean age 17 years)  4.1    Children’s Revised Impact of Event Scale (CRIES‐8): Assessment for PTSD  Overall 475 girls reported experiences of physical abuse or sexual abuse or exploitation and were administered the CRIES-8. The mean combined score on the CRIES-8 was 18.25 (median=18.0, SD=7.77), and 7.59 (median=8.0, SD=4.75) and 10.67 (median=10.0, SD=5.17) on the intrusion and aversion subscales respectively. A combined score of 17 is considered fairly accurate for identifying PTSD in children (http://www.childrenandwar.org/measures/children%E2%80%99s- revised-impact-of-event-scale-8-%E2%80%93-cries-8/). A study in the USA among 63 children age 10-16 seen in the emergency department (ED) and 52 children age 7-18 seen in a specialized PTSD clinic found a mean CRIES-8 score of 12.9 and 23.9 respectively. A cut-off score of 17 was found to maximized sensitivity and specificity at 100% and 71% in the ED sample and at 94% and 59% respectively in the clinic sample (Perrin et al., 2005). Thus using a cut-off of 17, over half of the girls in our sample who reported physical abuse or sexual abuse or exploitation may be suffering from PTSD from this experience. This appears very high and is cause for concern; culturally-specific trauma-informed care may be appropriate for these girls. Access to such care as well as broader sexual violence prevention and treatment services should be investigated during the endline survey.   54 CAREGIVER SURVEY  Gender Norm Attitudes  Most caregivers in the survey displayed gender egalitarian attitudes; for only one outcome did caregivers show a more male-favored attitude – that every women needs a man to protect her because she cannot protect herself. Nearly half (47 percent) also agreed that a good woman never refuses anything her husband says. Both of these questions were among the eight that comprised the “Rights and Privileges of Men Sub-scale”. For the three questions in the “Equality of Girls Sub-scale,” 69 percent or more of caregivers showed attitudes indicating gender equality. The only direct comparison in the literature were questions was Waszak et al. (2000) who used these (and several other) questions with adult Egyptian women. In the vast majority of cases, GE caregivers showed much more equitable attitudes, with the exception of the two “Rights and Privileges of Men Sub-scale” questions mentioned above (women needing men for protection, and not refusing what husband says). For the “Equality of Girls Sub-scale” questions, GE caregivers were similar to Egyptian women, with the exception of being more equitable about girls being able to work outside the home even after having children. Table 62. Gender Norm Attitudes  Percent who agree    GE caregivers  Egypt women  (2015)   (2001)  Rights and Privileges of Men Sub‐scale      It is better for boys to go to school than girls.  10  30  Girls should be sent to school only if they are not needed to help at home.  9  20  The main reason why boys/sons should be more educated than girls is for  34  50  them to take care of their parents when they are older.  If there is a very small amount of money to pay for school fees it is better to  15  30  send the boys to school first.  Women should leave politics or government business to the men.  23  80  Every woman needs a man to protect her because she cannot protect herself.  68  82  The only thing a woman can depend on in her old age is her sons.  32  59  A good woman never refuses anything her husband says.  47  56  Equality of Girls Sub‐scale      Daughters should have just the same chance to work outside the homes as  72  67  sons.  Girls should be told that if they have plenty of children they will not be able  69  44  work outside the home and make money.  I would like my girl child to work outside the home so she can support herself  85  75  and her family  Attitudes about Child Rearing  Despite caregivers’ high aspirations for their girls to gain a high level of education (post- secondary) and then marry and begin child-bearing in their mid-twenties, almost half also agreed 55 that girls need to be physically punished in order to raise them correctly. Such a high prevalence of this attitude is of great concern. While this practice may be viewed as an appropriate child disciplinary method, it condones acceptance of violence and girls may therefore view violence as an expected norm in their relationships with adults and sexual partners. Such acceptance may reduce both their inclination to reject violence in sexual partnerships, as well as to seek help/treatment if/when it does occur. Girls’ attitudes and behaviors around recognizing, reporting and seeking help for violence should be investigated during the endline survey. 56 References   Angold A, Costello EJ, Messer SC, Pickles A, Winder F, & Silver D. 1995. The development of a short questionnaire for use in epidemiological studies of depression in children and adolescents. International Journal of Methods in Psychiatric Research, 5, 237 - 249.http://devepi.duhs.duke.edu/instruments/MFQ%20Child%20Self-Report%20- %20Short.pdf. Accessed 15 July 2014. Kuo ES, Stoep AV, Stewart DG. 2005. Using the short mood and feelings questionnaire to detect depression in detained adolescents. Assessment. Vol. 12, No. 4, 374-83. Liberia Institute of Statistics and Geo-Information Services (LISGIS) [Liberia], Ministry of Health and Social Welfare [Liberia], National AIDS Control Program [Liberia], and Macro International Inc. 2008. Liberia Demographic and Health Survey 2007. Monrovia, Liberia: Liberia Institute of Statistics and Geo-Information Services (LISGIS) and Macro International Inc. Liberia Institute of Statistics and Geo-Information Services (LISGIS), Ministry of Health and Social Welfare [Liberia], National AIDS Control Program [Liberia], and ICF International. 2014. Liberia Demographic and Health Survey 2013. Monrovia, Liberia: Liberia Institute of Statistics and GeoInformation Services (LISGIS) and ICF International. Lundervold AJ, Breivik K, Posserud MB, Stormark KM, Hysing M. 2013. Symptoms of depression as reported by Norwegian adolescents on the Short Mood and Feelings Questionnaire. Frontiers in psychology. Vol 4, 613. McCarthy K, Brady M, and Hallman K. 2016. Investing When It Counts: Reviewing the evidence and charting a course of research and action for very young adolescents. New York: Population Council. http://www.popcouncil.org/uploads/pdfs/2016PGY_InvestingWhenItCounts.pdf Nanda G, Compendium of Gender Scales. 2011: Washington, DC: FHI 360/C-Change. Nic Gabhainn S and Mullan E. 2003. Self-esteem norms for Irish young people. Psychological reports. Vol. 92, No. 3-Pt 1, 829-30. Perrin S, Meiser-Stedman R, Smith P. 2005. The Children’s Revised Impact of Event Scale (CRIES): Validity as a Screening Instrument for PTSD. Behavioural and Cognitive Psychotherapy. Vol. 33, 487–98. 57 Population Council. 2009. The Adolescent Experience In-Depth: Using Data to Identify and Reach the Most Vulnerable Young People: Liberia 2007. New York: Population Council. Pulerwitz J and Barker G. 2008. Measuring attitudes toward gender norms among young men in Brazil. Men and Masculinities, Vol. 10, No. 3, 322338. Rhew IC, Simpson K, Tracy M, Lymp J, McCauley E, Tsuang D, et al. 2010. Criterion validity of the Short Mood and Feelings Questionnaire and one- and two-item depression screens in young adolescents. Child and adolescent psychiatry and mental health. Vol. 4, No. 1, 8. Rothon C, Stansfeld SA, Mathews C, Kleinhans A, Clark C, Lund C, et al. 2011. Reliability of self-report questionnaires for epidemiological investigations of adolescent mental health in Cape Town, South Africa. Journal of child and adolescent mental health. Vol. 23, No. 2, 119-28. Rosenberg, M., Society and the adolescent child. 1965, Princeton, NJ: Princeton University Press. Santhya, KG and Shireen J. Jejeebhoy. 2015 Sexual and reproductive health and rights of adolescent girls: Evidence from low- and middle-income countries. Global Public Health, Vol. 10, No. 2, 189–221, http://dx.doi.org/10.1080/17441692.2014.986169 Schmitt DP and Allik J. 2005. Simultaneous administration of the Rosenberg Self-Esteem Scale in 53 nations: exploring the universal and culture-specific features of global self-esteem. Journal of personality and social psychology. Vol 89, No. 4, 623-42. Stephenson R, Bartel D and Rubardt M. 2012. Constructs of power and equity and their association with contraceptive use among men and women in rural Ethiopia and Kenya. Global Public Health, Vol. 7, No. 6, 618–634. United Nations Children’s Fund. 2007. A National Study on Violence Against Children and Young Women in Swaziland. Mbabane: UNICEF. United Nations Children’s Fund, U.S. Centers for Disease Control and Prevention, Muhimbili University of Health and Allied Sciences. 2011. Violence against Children in Tanzania: Findings from a National Survey 2009. Dar es Salaam: UNICEF. United Nations Children’s Fund-Kenya Country Office; Division of Violence Prevention, National Center for Injury Prevention and Control, U.S. Centers for Disease Control and Prevention; and the Kenya National Bureau of Statistics. 2012. Violence against Children 58 in Kenya: Findings from a 2010 National Survey. Summary Report on the Prevalence of Sexual, Physical and Emotional Violence, Context of Sexual Violence, and Health and Behavioral Consequences of Violence Experienced in Childhood. Nairobi: UNICEF. United Nations Children’s Fund. 2014. Hidden in Plain Sight: A statistical analysis of violence against children. New York: UNICEF. Waszak, C, JL Severy, L Kafafi, and I Badawi. 2000. Fertility behavior and psychological stress: The mediating influence of gender norm beliefs among Egyptian women. Psychology of Women Quarterly 25:197–208. Zimbabwe National Statistics Agency (ZIMSTAT), United Nations Children’s Fund (UNICEF) and Collaborating Centre for Operational Research and Evaluation (CCORE). 2013. National Baseline Survey on Life Experiences of Adolescents, 2011. Harare: ZIMSTAT.   59 Appendix   Supplementary Survey Tables  Table A1  Household listing item A15: Where do the people  living in this house get their drinking water from?  Percent  Number  Water is piped into the yard/plot  0.43%  5  Water is from a Tube well of borehole in the  yard/plot  2.43%  28  Water is from a dug well with a hand pump in the  yard/plot  49.05%  566  Water is from an unprotected (no pump covering  opening) well in the yard/plot  4.59%  53  Water is from public tap or standpipe  4.51%  52  Water is from a public tube well, borehole, or dug  well with  a pump  29.46%  340  Water is from a public unprotected well (no pump)  5.98%  69  Water is surface water  from spring, lake or river  3.03%  35  bottled water is used  0.09%  1  Other water source  0.43%  5     Total  1154  Table A2  Household listing item A16: What type of toilet do  people in your house use?  Percent  Number  Flush or pour toilet in house  5.46%  63  Ventilated improved pit latrine  2.60%  30  Pit latrine with slab (cement floor)  22.96%  265  Pit latrine without slab (open pit)  15.08%  174  Composting toilet  0.26%  3  Hanging toilet/hanging latrine (i.e. built hanging  over a body of water)  5.89%  68  No facility/bush/open field  47.57%  549  Other  0.17%  2     Total  1154  60 Table A3   Household listing item A8b: How many families  live in this house?    Percent  Number  1  70.47%  7701  2  13.16%  1438  3  9.32%  1019  4  4.49%  491  5  1.65%  180  6  0.69%  75  7  0.13%  14  8  0.09%  10     Total  10928  Table A4  Girl survey item B1a: What is your year of birth?    Percent  Number  Don't know  10.61%  129  No response  0.16%  2  1984  0.08%  1  1995  0.08%  1  1997  0.08%  1  1998  0.25%  3  1999  0.08%  1  2000  13.82%  168  2001  38.65%  470  2002  35.69%  434  2003  0.49%  6     Total  1216  Table A5  Girl survey item B2: Have you ever attended  school?    Percent  Number  No  2.06%  25  Yes  97.94%  1191     Total  1216  61 Table A6  Girl survey item B3a: What is the highest grade  that you have completed in school?   Percent  Number  None  0.25%  3  ABC  0.84%  10  K1  4.03%  48  K2  9.57%  114  Grade 1  17.13%  204  Grade 2  20.99%  250  Grade 3  18.39%  219  Grade 4  15.20%  181  Grade 5  7.98%  95  Grade 6  3.78%  45  Grade 7  1.26%  15  Grade 8  0.50%  6  Grade 9  0.08%  1     Total  1191  Table A7  Caregiver survey item B3: Caregiver’s gender  Percent  Number  Female  72.27%  834  Male  27.73%  320     Total  1154  62 Table A8  Caregiver survey item B4: Highest grade caregiver  has completed (passed) in school?  Percent  Number  None  24.00%  277  ABC  1.56%  18  K1  0.17%  2  K2  0.61%  7  Grade 1  2.17%  25  Grade 2  3.03%  35  Grade 3  5.20%  60  Grade 4  6.85%  79  Grade 5  6.50%  75  Grade 6  7.54%  87  Grade 7  7.54%  87  Grade 8  5.55%  64  Grade 9  5.11%  59  Grade 10  5.81%  67  Grade 11  3.55%  41  Grade 12  11.09%  128  Post‐secondary formal  3.64%  42  Don’t know  0.09%  1     Total  1154  Table A9  Caregiver survey item B6: Who is the head of the  household in relationship to the girl(s)?  Percent  Number  Mother  8.95%  29  Father  58.02%  188  Grandmother  6.17%  20  Grandfather  5.25%  17  Stepmother  0.31%  1  Stepfather  4.01%  13  Sister  1.23%  4  Brother  2.78%  9  Aunt  1.54%  5  Uncle  7.72%  25  Other  4.01%  13     Total  324  Caregiver survey item B6b: Specify gender of  “other” person  Percent  Number  Female  23.08%  3  Male  76.92%  10     Total  13  63 Table A10  Caregiver survey item B7: What is the highest  grade that the head of household has completed  (passed) ?  Percent  Number  None  21.30%  69  Grade 1  0.62%  2  Grade 2  1.23%  4  Grade 3  2.16%  7  Grade 4  1.54%  5  Grade 5  2.47%  8  Grade 6  6.17%  20  Grade 7  4.32%  14  Grade 8  5.25%  17  Grade 9  6.17%  20  Grade 10  5.56%  18  Grade 11  5.56%  18  Grade 12  25.00%  81  Post‐secondary formal  6.17%  20  Don’t know  6.48%  21     Total  324  Table A11  Girl survey item C2: What family members died  from Ebola?    Percent  Number  Mother  2.04%  1  Father  2.04%  1  Grandmother  2.04%  1  Grandfather  4.08%  2  Sister  18.37%  9  Brother  16.33%  8  Aunt  38.78%  19  Uncle  32.65%  16  Husband  0.00%  0  Boyfriend  0.00%  0  Play mum  0.00%  0  Other family member  14.29%  7  Don't Know  0.00%  0  No Response  0.00%  0  *multiple responses allowed  Respondents:  49  64 Table A12  Girl survey item B6: Are you enrolled in the  2015/2016 school year? I am referring to the  2015/2016 school year that begins in September  2015.  Percent  Number  No, I am not enrolled  83.55%  1016  Yes, I am enrolled  16.37%  199  Don't know  0.08%  1     Total  1216  Girl survey item B6a: What is the primary reason  you are not enrolled in the 2015/2016 school year?  Percent  Number  Registration is not yet opened  95.96%  975  Family could not afford  2.17%  22  Family does not approve/see benefit  0.10%  1  Got pregnant  0.20%  2  Result of Ebola epidemic  0.49%  5  Other  0.79%  8  Don't Know  0.20%  2  No Response  0.10%  1     Total  1016  Table A13  Girl survey item O7: How much peer pressure is  there on people your age to have sex?  Percent  Number  None  38.24%  465  A little  21.22%  258  A moderate amount  3.87%  47  A lot  32.57%  396  A great deal  3.62%  44  Don't Know  0.49%  6     Total  1216  Table A14  Girl survey item O1: Have you ever kissed a boy in  a loving or sexy way?  Percent  Number  No  82.07%  998  Yes  17.85%  217  Don't Know  0.08%  1     Total  1216  65 Table A15  Girl survey item O8: Have you ever had sex,  whether this was something you wanted to do or  not? This includes when a man’s penis enters  someone’s vagina or anus.  Percent  Number  No  79.19%  963  Yes  20.72%  252  Don't Know  0.08%  1     Total  1216  Table A16  Girl survey item O11: If have had sex (“yes” to item  O8), the first time had sex, would you say you did  it because you wanted to do it or because you  were forced or tricked into doing it against your  will?  Percent  Number  Wanted to  71.43%  180  Tricked  14.68%  37  Physically forced  13.89%  35     Total  252  Table A17  Girl survey item O16: Have you ever been  pregnant? This includes if you are currently  pregnant.  Percent  Number  No  92.06%  232  Yes  7.94%  20     Total  252  Table A18  Girl survey item O18: Think back to your first  pregnancy.  At the time you first became pregnant,  did you want to become pregnant then, did you  want to wait until later, or did you not want  children at all?  Percent  Number  Wanted the pregnancy  10.00%  2  Wanted to wait until later to become pregnant  65.00%  13  Did not want children at all  25.00%  5     Total  20  66 Table A19  Girl survey item O19: What happened with your  last or most recent pregnancy?    Percent  Number  Aborted the pregnancy  15.00%  3  Miscarried  10.00%  2  Live birth  30.00%  6  Currently pregnant AND this is my first pregnancy  45.00%  9     Total  20  Table A20  Girl survey item O20: If the pregnancy resulted in a  live birth, who is raising the child?  Percent  Number  Respondent  (child’s mother)  66.67%  10  Father of the baby  20.00%  3  Other family member  6.67%  1  Other  6.67%  1     Total  15  Table A21  Girl survey item O3: At the time you got  married/started living together, did you agree  willingly?  Percent  Number  No  11.11%  1  Yes  88.89%  8     Total  9  67 Table A22a   Girl survey item O4: If living with husband/partner, how old is your  husband/partner?  Observations  Mean  Median  SD  Min  Max  7  18.43  18  1.62  17  21    Table A22b  Girl survey item O5. If not married, do you have a  boyfriend at this time?*  Percent  Number  No  79.69%  969  Yes  20.31%  247     Total  1216  * Data for this variable contained 8 cases where boyfriend age was listed as 1 and 2 cases where boyfriend age was listed as 2. This summary table excludes those cases. Table A22c  Girl survey item O6: If have boyfriend (yes to O5), how old is your boyfriend?*  Observations  Mean  Median  SD  Min  Max  134  16.93  17  4.56  10  26  * Data for this variable excludes girls who reported “don’t know” for the age of their boyfriend. Table A23  Girl survey item P1: Has anyone ever hit, slapped,  kicked, or done anything bad to hurt you before?  Percent  Number  No  50.74%  617  Yes  49.18%  598  Don't Know  0.08%  1     Total  1216  68 Table A24  Girl survey items Q1‐Q6:  Sexual Abuse or Exploitation   (Ever; and Conditional Upon Ever, Happened in Past 12 Months)  Response options  Question  No  Yes  DK  Total  Q1. Has anyone, male or female, ever touched  863  351  2  1216  you in a sexual way without your permission,  but did not try and force you to have sex?  (70.97%)  (28.87%)  (0.16%)  Q1a. Has this happened in the past 12  46  305  0  351  months?  (13.11%)  (86.89%)  (0.00%)  Q2. Has anyone, male or female, ever tried to  916  300  0  1216  do man and woman business against your will,  but did not succeed?  (75.33%)  (24.67%)  (0.00%)  Q2a. Has this happened in the past 12  44  256  0  300  months?  (14.67%)  (85.33%)  (0.00%)  Q3. Have you ever had sex with anyone, male  1112  102  2  1216  or female, after they pressured you in a non‐ physical way?  (91.45%)  (8.39%)  (0.16%)  Q3a. Has this happened in the past 12  7  95  0  102  months?  (6.86%)  (93.14%)  (0.00%)  Q4. Has anyone, male or female, ever  1120  95  1  1216  physically forced you to have sex with them  without your permission?  (92.11%)  (7.81%)  (0.08%)  Q4a. Has this happened in the past 12  15  80  0  95  months?  (15.79%)  (84.21%)  (0.00%)  Q5. Has anyone ever given you money to do  1123  91  2  1216  man and woman business with them?  (92.35%)  (7.48%)  (0.16%)  Q5a. Has this happened in the past 12  7  84  0  91  months?  (7.69%)  (92.31%)  (0.00%)  Q6. Has anybody ever given you food, gifts, or  1119  96  1  1216  any favors so that you have sex with them?  (92.02%)  (7.89%)  (0.08%)  Q6a. Has this happened in the past 12  6  90  0  96  months?  (6.25%)  (93.75%)  (0.00%)  69 Table A25  CRIES questions (girl survey)  Response options  Question  Not at all  Rarely  Sometimes  Often  NR  Total  R1. Do you think about it  153  133  111  68  1  466  even when you don’t mean  to think about it?  (32.83%)  (28.54%)  (23.82%)  (14.59%)  (0.21%)  R2. Do you try to remove it  68  96  171  131  0  466  from your mind?  (14.59%)  (20.6%)  (36.7%)  (28.11%)  (0.00%)  R3. Do you have waves of  129  111  159  67  0  466  strong feelings about it?  (27.68%)  (23.82%)  (34.12%)  (14.38%)  (0.00%)  R4. Do you stay away from  77  111  126  152  0  466  reminders of it (e.g. places  or situations)?  (16.52%)  (23.82%)  (27.04%)  (32.62%)  (0.00%)  R5. Do you try not to talk  91  82  145  148  0  466  about it?  (19.53%)  (17.6%)  (31.12%)  (31.76%)  (0.00%)  R6. Do pictures about it  126  124  142  74  0  466  appear in your mind?  (27.04%)  (26.61%)  (30.47%)  (15.88%)  (0.00%)  R7. Do other things keep  139  102  169  56  0  466  making you think about it?  (29.83%)  (21.89%)  (36.27%)  (12.02%)  (0.00%)  R8. Do you try not to think  67  121  154  124  0  466  about it?  (14.38%)  (25.97%)  (33.05%)  (26.61%)  (0.00%)  Table A26  Girl survey item E1: Is there a place that you could go to  sleep if there was an emergency situation or something  happened that made you feel you are not safe where you  usually sleep?  Percent  Number  No  51.89%  631  Yes  48.11%  585     Total  1216  70 Table A27  Girl survey item E3: If you attend school, do you feel  safe when you are at school?    Percent  Number  No  6.88%  80  Yes  93.12%  1083     Total  1163  Table A28  Girl survey item E4: When you travel to school, do you feel  safe?    Percent  Number  No  8.60%  100  Yes  91.40%  1063     Total  1163  Table A29  Girl survey item E5: Is there a safe place in the community  outside of school and home where you feel free to meet  and talk freely with other girls?  Percent  Number  No  19.08%  232  Yes  80.92%  984     Total  1216  Table A30  Girl survey item E6: How often do you go to this safe  place?    Percent  Number  Every day  22.15%  218  Once a week  3.86%  38  Once a month  0.20%  2  Rarely  72.97%  718  Never  0.81%  8     Total  984  71 Table A31  Girl survey item E7: Apart from someone in your  household, is there a big woman in the community you  can usually go to with problems?  Percent  Number  No  64.64%  786  Yes  35.36%  430     Total  1216   Table A32  Girl survey: Self Esteem Scale Components  Response options  Strongly  Strongly  Disagree  Agree  DK  NR  Question  disagree  agree  Total  F1. I feel that I am at  125  48  189  853  1  0  1216  least as important as  others  (10.28%)  (3.95%)  (15.54%)  (70.15%)  (0.08%)  (0.00%)  F2. I feel that I have  61  46  192  917  0  0  1216  plenty good things in  me.  (5.02%)  (3.78%)  (15.79%)  (75.41%)  (0.00%)  (0.00%)  F3. All in all, I feel that  664  245  128  179  0  0  1216  I will not make it in  life.  (54.61%)  (20.15%)  (10.53%)  (14.72%)  (0.00%)  (0.00%)  F4. I am able to do  101  66  197  852  0  0  1216  things as well as most  other people.  (8.31%)  (5.43%)  (16.2%)  (70.07%)  (0.00%)  (0.00%)  F5. I feel I do not have  325  160  303  425  2  1  1216  much to make me  proud.  (26.73%)  (13.16%)  (24.92%)  (34.95%)  (0.16%)  (0.08%)  F6. I take a good  28  25  170  993  0  0  1216  attitude toward  myself  (2.3%)  (2.06%)  (13.98%)  (81.66%)  (0.00%)  (0.00%)  F7. All in all, I am  61  45  148  962  0  0  1216  satisfied with myself.  (5.02%)  (3.7%)  (12.17%)  (79.11%)  (0.00%)  (0.00%)  F8. I wish I could have  93  76  248  799  0  0  1216  more respect for  myself.  (7.65%)  (6.25%)  (20.39%)  (65.71%)  (0.00%)  (0.00%)  F9. I feel useless  616  233  200  166  1  0  1216  sometimes.  (50.66%)  (19.16%)  (16.45%)  (13.65%)  (0.08%)  (0.00%)  549  249  223  195  0  0  1216  F10. At times I think I  am no good at all.  (45.15%)  (20.48%)  (18.34%)  (16.04%)  (0.00%)  (0.00%)  72 Table A33  Girl survey item G1: What grade in school would you hope to  complete before leaving school?  Percent  Number  None  0.58%  7  K1  0.08%  1  K2  0.08%  1  Grade 1  0.08%  1  Grade 2  0.16%  2  Grade 3  0.33%  4  Grade 4  0.33%  4  Grade 5  0.33%  4  Grade 6  0.74%  9  Grade 7  0.49%  6  Grade 8  0.49%  6  Grade 9  2.22%  27  Grade 10  5.10%  62  Grade 11  3.78%  46  Grade 12  55.92%  680  Post‐Secondary Formal  28.78%  350  Don’t know  0.49%  6     Total  1216  Table A34  Girl survey item G4: Would you hope to have a job outside of  the home even after marriage or would you prefer to no work  outside the home?  Percent  Number  Job Outside the home  82.07%  998  No Work outside the home  17.85%  217  Don't Know  0.08%  1     Total  1216  73 Table A35  Girl Survey:  Gender Relations Scale components  Response options  Question  Disagree  Agree  DK  NR  Total  J2. Men need to have sex more than  587  592  35  2  1216  women.  (48.27%)  (48.68%)  (2.88%)  (0.16%)  573  629  14  0  1216  J3. It is a woman’s duty to not get pregnant.  (47.12%)  (51.73%)  (1.15%)  (0.00%)  J4. A man should have the final word about  462  754  0  0  1216  decisions in his household.  (37.99%)  (62.01%)  (0.00%)  (0.00%)  600  578  34  4  1216  J5. Men are always ready to have sex.  (49.34%)  (47.53%)  (2.8%)  (0.33%)  J6. A woman should accept violence to keep  550  665  1  0  1216  the family together.  (45.23%)  (54.69%)  (0.08%)  (0.00%)  J7. A man needs other women even if things  713  499  4  0  1216  with his wife are fine.  (58.63%)  (41.04%)  (0.33%)  (0.00%)  J8. A man can beat his wife if she does not  783  419  13  1  1216  agree to have sex with him.  (64.39%)  (34.46%)  (1.07%)  (0.08%)  J9. Husband and wife should agree if they  192  1023  1  0  1216  want to have children.  (15.79%)  (84.13%)  (0.08%)  (0.00%)  J10. Taking care of children is the mother’s  297  919  0  0  1216  duty.  (24.42%)  (75.58%)  (0.00%)  (0.00%)  J11. A woman can suggest using condoms  589  538  80  9  1216  just like a man can.  (48.44%)  (44.24%)  (6.58%)  (0.74%)  J12. A man and a woman should decide  345  834  29  8  1216  together what type of contraceptive to use.  (28.37%)  (68.59%)  (2.38%)  (0.66%)  591  622  3  0  1216  J13. A real man produces a male child.  (48.6%)  (51.15%)  (0.25%)  (0.00%)  J14. Men and women should share  677  539  0  0  1216  household chores.  (55.67%)  (44.33%)  (0.00%)  (0.00%)  673  504  30  9  1216  J15. A woman should not initiate sex.  (55.35%)  (41.45%)  (2.47%)  (0.74%)  74 Table A36  Girl survey:  Monitoring, Conflict, Emotional Support and Financial Support Scale (MCEF)  (range: 0‐30)  Observations  Mean  Median  SD  Min  Max  1212  24.75  25  3.64  11  30  Monitoring Subscale (MCEF) (range: 0‐6)  Observations  Mean  Median  SD  Min  Max  1216  5.39  6  1.26  2  6  Conflict Subscale (MCEF) (range: 0‐9)  Observations  Mean  Median  SD  Min  Max  1214  5.99  6  2.14  3  9  Emotional Support Subscale (MCEF) (range: 0‐9)  Observations  Mean  Median  SD  Min  Max  1215  7.97  9  1.45  3  9  Financial Support Subscale (MCEF) (range: 0‐6)  Observations  Mean  Median  SD  Min  Max  1215  5.4  6  1  2  6  Table A37  Girl survey item D1: Is there a big person in your life who  knows where you are at night? (MCEF component)  Percent  Number  No  20.15%  245  Yes  79.85%  971     Total  1216  Girl survey item D2: If “yes” to D1, who usually knows where  you are at night?*     Percent  Number  Mother  64.68%  628  Father  34.29%  333  Grandmother  12.77%  124  Grandfather  2.47%  24  Sister  28.22%  274  Brother  14.21%  138  Aunt  15.86%  154  Uncle  7.52%  73  Husband  0.00%  0  Boyfriend  0.62%  6  Play mom  0.10%  1  Other adult  2.16%  21  Don't Know  0.10%  1  No Response  0.00%  0  *multiple responses allowed  Respondents:  971  75 Table A38  Girl survey Item D3: Is there any big person who usually  knows where you are during the day? (MCEF component)  Percent  Number  No  10.53%  128  Yes  89.47%  1088     Total  1216  Girl survey item D2: If “yes” to D3, who usually knows where  you are during the day?  Percent  Number  Mother  65.26%  710  Father  33.64%  366  Grandmother  13.88%  151  Grandfather  3.13%  34  Sister  29.41%  320  Brother  15.99%  174  Aunt  17.56%  191  Uncle  7.35%  80  Husband  0.00%  0  Boyfriend  0.46%  5  Play mom  0.18%  2  Other adult  2.57%  28  Don't Know  0.00%  0  No Response  0.00%  0  *multiple responses allowed  Respondents:  1,088  76 Table A39  Girl survey item D11: There is a big person in your life who  provides you with “small, small” things (MCEF component)  Percent  Number  Not true at all  4.11%  50  Somewhat true  7.98%  97  Very true  87.91%  1069     Total  1216  D12. If “yes” to D11, who usually provides for your “small,  small” things?   Percent  Number  Mother  71.61%  835  Father  42.54%  496  Grandmother  8.92%  104  Grandfather  1.37%  16  Sister  19.81%  231  Brother  11.23%  131  Aunt  14.58%  170  Uncle  6.09%  71  Husband  0.17%  2  Boyfriend  1.63%  19  Play mom  0.17%  2  Other adult  1.46%  17  Don't Know  0.00%  0  No Response  0.00%  0  *multiple responses allowed  Respondents:  1,166  77 Table A40   Girl survey:  Relationship and Communication with Adults Indicators (MCEF components)  Response options  Not true  Somewhat  Very  DK  Question  at all  true  true  Total  D5. There is a big person in your life who  570  152  493  1  1216  makes you to look small  (46.88%)  (12.5%)  (40.54%)  (0.08%)  D6. There is a big person in your life who  419  200  597  0  1216  talks to you roughly or harshly  (34.46%)  (16.45%)  (49.1%)  (0.00%)  D7. There is a big person in your life who  577  176  462  1  1216  condemns you a lot  (47.45%)  (14.47%)  (37.99%)  (0.08%)  D8. There is a big person in your life who  135  168  913  0  1216  listens to you  (11.1%)  (13.82%)  (75.08%)  (0.00%)  D9. There is a big person in your life who  66  131  1019  0  1216  praises you  (5.43%)  (10.77%)  (83.8%)  (0.00%)  D10. There is a big person in your life  179  197  839  1  1216  who comforts you  (14.72%)  (16.2%)  (69%)  (0.08%)  Table A41  Girl survey item D17. If there is a big person in your life who  you regard as your mentor (“yes” to D15), what do you talk to  this person about?  Percent  Number  Plans for future  60.96%  709  Health problems  38.18%  444  School lessons  44.11%  513  Peers at school or in the community  12.04%  140  Family problems  16.68%  194  Conflict with your parents or guardians  7.31%  85  Conflicts with husband or boyfriends  0.95%  11  Conflicts with friends and neighbors  30.95%  360  Conflicts with siblings  17.71%  206  Conflicts at work  0.86%  10  Financial plans/saving money  2.32%  27  Work or career plans  1.89%  22  Plans you have with your boyfriend or husband  0.52%  6  Sex or sexual health  0.52%  6  Don't Know  0.17%  2  No Response  0.17%  2  *multiple responses allowed  Respondents  1,163  78 Table A42  Girl survey item D20.  If there a big person living in this house  hold who you would ask to accompany you to important  events (“yes” to D18), what do you talk to this person about?   Percent  Number  Plans for the future  55.28%  670  Health problems  68.40%  829  School lessons  51.98%  630  Peers at school or in the community  11.14%  135  Family problems  12.95%  157  Conflict with your parents or guardians  4.04%  49  Conflicts with husband or boyfriends  1.07%  13  Conflicts with friends and neighbors  20.54%  249  Conflicts with siblings  14.11%  171  Conflicts at work  0.17%  2  Financial plans/saving money  2.31%  28  Work or career plans  1.40%  17  Plans you have with your boyfriend or husband  0.41%  5  Sex or sexual health  0.74%  9  Don't know  0.00%  0  No Response  0.00%  0  *multiple responses allowed  Respondents:  1,212  79 Table A43  Girl survey:  Mental Health/Depression (SMFQ) Scale Components  Response options  Not true  Sometimes  True  DK  NR  Question  Total  H1. In the past two weeks, I  621  232  363  0  0  1216  felt miserable or unhappy  (51.07%)  (19.08%)  (29.85%)  (0.00%)  (0.00%)  H2. In the past two weeks, I  658  204  353  0  1  1216  didn’t enjoy anything at all  (54.11%)  (16.78%)  (29.03%)  (0.00%)  (0.08%)  H3. In the past two weeks, I  649  226  341  0  0  1216  felt so tired I just sat around  and did nothing  (53.37%)  (18.59%)  (28.04%)  (0.00%)  (0.00%)  H4. In the past two weeks, I  529  279  407  1  0  1216  was very restless  (43.5%)  (22.94%)  (33.47%)  (0.08%)  (0.00%)  H5. In the past two weeks, I  782  169  265  0  0  1216  felt I was no good anymore  (64.31%)  (13.9%)  (21.79%)  (0.00%)  (0.00%)  H6. In the past two weeks, I  713  188  315  0  0  1216  cried a lot  (58.63%)  (15.46%)  (25.9%)  (0.00%)  (0.00%)  H7. In the past two weeks, I  646  235  335  0  0  1216  found it hard to think properly  or concentrate  (53.13%)  (19.33%)  (27.55%)  (0.00%)  (0.00%)  H8. In the past two weeks, I  829  158  229  0  0  1216  hated myself  (68.17%)  (12.99%)  (18.83%)  (0.00%)  (0.00%)  H9. In the past two weeks, I  903  121  192  0  0  1216  was a bad person  (74.26%)  (9.95%)  (15.79%)  (0.00%)  (0.00%)  H10. In the past two weeks, I  731  216  269  0  0  1216  felt lonely  (60.12%)  (17.76%)  (22.12%)  (0.00%)  (0.00%)  H11. In the past two weeks, I  755  163  298  0  0  1216  thought nobody really loved  me  (62.09%)  (13.4%)  (24.51%)  (0.00%)  (0.00%)  H12. In the past two weeks, I  703  175  338  0  0  1216  thought I could never be as  good as other kids  (57.81%)  (14.39%)  (27.8%)  (0.00%)  (0.00%)  H13. In the past two weeks, I  831  189  196  0  0  1216  did everything wrong  (68.34%)  (15.54%)  (16.12%)  (%)  (%)  80 Table A44  Girl survey item I1:   Which of the choices could be called something you need?   Percent  Number  Food for your family  88.90%  1081  An extra pair of shoes  69.33%  843  Alcohol  6.66%  81  Don't Know  0.00%  0  No Response  0.00%  0  *multiple responses allowed  Respondents:  1,216                     Girl survey item I2:   Which of the choices could be called something you  need instead of just something you want?     Percent  Number  Food for your family  72.29%  879  Money for a place to live  66.94%  814  Soft Drink  30.18%  367  Don't Know  0.41%  5  No Response  0.16%  2  *multiple responses allowed  Respondents:  1,216  Table A45  Girl survey item I3: Which of the choices is a reason to put  money aside (save money)?   Percent  Number  To have money to meet an emergency (like an accident or  illness).  71.88%  874  For personal or family goals, like braiding your hair  51.56%  627  For future plans, like going on vacation  48.36%  588  Don't Know  0.16%  2  No Response  0.08%  1  *multiple responses allowed  Respondents:  1,216  81 Table A46  Girl survey item I10: In the past year, when you spent  money to buy things you need where did you get that  money from?    Percent  Number  Mother  56.09%  682  Father  43.83%  533  Other family member  19.57%  238  Cannot buy thing  12.66%  154  Guardian you live with  11.92%  145  Casual job  6.91%  84  Own savings  5.26%  64  Boyfriend or husband  4.36%  53  Other  2.06%  25  Savings group  1.81%  22  Friend  1.56%  19  Spend less on needs  1.32%  16  Don't Know  0.16%  2  Play Mom  0.33%  4  Steady job  0.08%  1  Cash transfer/remittances  0.08%  1  Sugar daddy  0.08%  1  No Response  0.00%  0  *multiple responses allowed  Respondents:  1,216  Table A47  Girl survey item I11: Do you own anything that helps you  make money?   Percent  Number  No  64.72%  787  Basket/bucket  22.04%  268  Garden/farm  11.84%  144  Other   3.87%  47  Wheelbarrow/cart  0.74%  9  Booth/stall  0.99%  12  Don't Know  0.08%  1  Cell Phone  0.00%  0  Phone charger  0.00%  0  No Response  0.00%  0  *multiple responses allowed  Respondents:  1,216  82 Table A48  Girl survey item K3: As girls grow into women, what  changes happen in their bodies?   Percent  Number  Develop breasts  92.02%  1119  Grow taller  68.75%  836  Grow hair in new areas of body  47.86%  582  Menstruation  41.28%  502  Gain weight  30.59%  372  Hormonal changes  1.32%  16  Other  0.49%  6  Don't Know  0.33%  4  No Response  0.08%  1  *multiple responses allowed  Respondents:  1,216  Table A49  Girl survey item K2: Do you think your monthly period is  something to be ashamed of?  Percent  Number  Yes  60.61%  737  No  26.56%  323  Don't Know  12.66%  154  No Response  0.16%  2     Total  1216  83 Table A50  Girl survey item L3: I am now going to read out a list of  contraceptive/family planning methods.  Which of these have  you used in the past 12 months?   Percent  Number  Has not used any  96.06%  1047  Oral contraceptive pill  0.99%  12  Hormonal injection (i.e. Depo Provera, Nur Isterate)  1.47%  16  Intra‐uterine device (IUD or Loop)  0.00%  0  Hormonal implant  (Implanon or  Nexplanon  0.00%  0  Jelly/foam   0.00%  0  Male condom  1.01%  11  Female condom  0.00%  0  Diaphragm  0.00%  0  Emergency Contraceptive Pills  (morning after pill)  0.00%  0  Withdrawal  (a man can pull out of a woman before climax to  avoid pregnancy)  0.00%  0  Periodic Abstinence/Rhythm  (a couple can avoid sex on days  when pregnancy is most likely to occur)  0.00%  0  Non vaginal sex(Anal or oral sex, thigh sex)  0.00%  0  Other  0.28%  3  Don't Know  0.92%  10  No Response  0.00%  0  *multiple responses allowed  Respondents:  1,090  Table A51  Girls’ Knowledge of Condoms   Response options  Question  Agree  Disagree  DK  NR  Total  N1. Condoms are a good way of  756  369  80  11  1216  preventing pregnancy  (62.17%)  (30.35%)  (6.58%)  (0.9%)  N2. Condoms are a good way of protecting  904  226  80  6  1216  against HIV/AIDS  (74.34%)  (18.59%)  (6.58%)  (0.49%)  N3. Condoms are an effective way of  896  224  83  13  1216  protecting against sexually transmitted  diseases  (73.68%)  (18.42%)  (6.83%)  (1.07%)  N4. Condoms can slip from on the man  325  735  139  17  1216  and get lost inside the woman's body  (26.73%)  (60.44%)  (11.43%)  (1.4%)  N5. Condoms can be used more than one  310  777  114  15  1216  time  (25.49%)  (63.9%)  (9.38%)  (1.23%)  84 Table A52  Girl survey item M4: What are the signs or symptoms of  other (than HIV) sexually transmitted infections when in a  woman?   Percent  Number  Pain during urination  48.52%  590  Discharge from vagina  32.81%  399  Don't Know  30.92%  376  Ulcers/sores in genital area  12.91%  157  Other  5.76%  70  No Response  1.64%  20  *multiple responses allowed  Respondents:  1,216  85 Table A53  Caretaker Gender Role Norm Attitudes   Response options  Question  Agree  Disagree  DK  NR  Total  C1. It is better for boys to go to school than  114  1038  1  1  1154  girls  (9.88%)  (89.95%)  (0.09%)  (0.09%)  C2. Girls should be sent to school only if  102  1050  1  1  1154  they are not needed to help at home.  (8.84%)  (90.99%)  (0.09%)  (0.09%)  C3. The main reason why boys/sons should  398  756  0  0  1154  be more educated than girls is for them to  take care of their parents when they are  (34.49%)  (65.51%)  (0.00%)  (0.00%)  older.  C4. If there is a very small amount of money  177  977  0  0  1154  to pay for school fees it is better to send the  boys to school first.  (15.34%)  (84.66%)  (0.00%)  (0.00%)  C5. Women should leave politics or  274  879  1  0  1154  government business to the men.  (23.74%)  (76.17%)  (0.09%)  (0.00%)  C6. Every woman needs a man to protect  789  364  1  0  1154  her because she cannot protect herself  (68.37%)  (31.54%)  (0.09%)  (0.00%)  C7. The only thing a woman can depend on  367  786  1  0  1154  in her old age is her sons.  (31.8%)  (68.11%)  (0.09%)  (0.00%)  C8. A good woman never refuses anything  537  615  2  0  1154  her husband says.  (46.53%)  (53.29%)  (0.17%)  (0.00%)  C9. Daughters should have just the same  832  322  0  0  1154  chance to work outside the homes as sons  (72.1%)  (27.9%)  (0.00%)  (0.00%)  C10. Girls should be told that if they have  801  353  0  0  1154  plenty children they will not be able work  outside the home and make money.  (69.41%)  (30.59%)  (0.00%)  (0.00%)  C11. I would like my girl child to work  980  174  0  0  1154  outside the home so she can support  herself and her family  (84.92%)  (15.08%)  (0.00%)  (0.00%)  Table A54  D1. Do you believe that in order to bring up girls  correctly, you need to physically punish them?  Percent  Number  No  55.37%  639  Yes  44.63%  515     Total  1154  86 Table A55  Comparison Rosenberg Self‐Esteem Scores  Nation  Alpha   Mean    SD  Botswana   .72   30.85   4.11  Dem. Rep. of the Congo   .45   31.28   2.93  Ethiopia   .64   29.24   3.69  Tanzania   .61   29.52   3.95  Zimbabwe   .75   30.77   4.07  United States (reference)   .75   30.77   4.07    87 Respondent Comprehension and Survey Environment  Caregiver Survey  Questions in the following tables were asked to caregivers during the caregiver survey to assess the respondent’s level of comprehension and the main language used during the interview. For the question about whether or not the respondent appeared to understand the questions, enumerators judged respondent understanding on a scale of 1-5, with 1 indicating the respondent never understood and 5 indicating the respondent always understood. Table A56  Z1. Were the questions you answered:  Percent  Number  Very easy to understand  29.29%  338  Easy to understand  69.15%  798  Difficult to understand  1.47%  17  Very difficult to understand  0.09%  1     Total  1154  Table A57  END1. DID THE RESPONDENT APPEAR TO UNDERSTAND  THE QUESTIONS?  Percent  Number  1 Never understood  18.20%  210  2  25.04%  289  3  5.81%  67  4  17.07%  197  5 Always understood  33.88%  391     Total  1154  Table A58  END2. ENUMERATOR: WHAT WAS THE MAIN  LANGUAGE USED DURING THIS INTERVIEW?  Percent  Number  Liberian English  79.90%  922  Gio  4.07%  47  Mano  15.86%  183  Other  0.17%  2     Total  1154  88 Girl Survey  Questions in the following tables were asked to girls during the girl survey to assess the respondent’s level of comprehension, the enumerator’s perception of the girl’s age and the main language used during the interview. For the question about whether or not the respondent appeared to understand the questions, enumerators judged respondent understanding on a scale of 1-5, with 1 indicating the respondent never understood and 5 indicating the respondent always understood. Table A59  Z1. Were the questions you answered:  Percent  Number  Very easy to understand  14.39%  175  Easy to understand  80.76%  982  Difficult to understand  4.28%  52  Very difficult to understand  0.58%  7     Total  1216  Table A60  END1. DID THE RESPONDENT APPEAR TO UNDERSTAND  THE QUESTIONS?  Percent  Number  1 Never understood  28.37%  345  2  49.26%  599  3  15.46%  188  4  4.44%  54  5 Always understood  2.47%  30     Total  1216  Table A61  END2. WAS ANYONE ELSE PRESENT DURING ANY PART  OF THE INTERVIEW?  Percent  Number  No  97.62%  1187  Yes  2.38%  29     Total  1216  Table A62  END 2C. DID THIS PERSON INTERFERE WITH THE  INTERVIEW?*  Percent  Number  No  75.86%  22  Yes  24.14%  7     Total  29  * The person identified as being present during the interview in the previous question. 89 Table A63  END3. WHAT WAS THE SETTING IN WHICH THE  INTERVIEW TOOK PLACE?  Percent  Number  Quiet, private  94.08%  1144  Some noise, semi‐private  5.59%  68  Noisy, people around  0.33%  4     Total  1216  Table A64  END4. OVERALL, HOW WAS THE RESPONDENT’S  INTEREST IN THE INTERVIEW?  Percent  Number  Very high  23.85%  290  Above average  36.02%  438  Average  36.27%  441  Below average  2.14%  26  Very low  1.73%  21     Total  1216  Table A65  END6. DID THE RESPONDENT APPEAR TO BE WITHIN  THE 13‐14 YEAR OLD AGE RANGE?  Percent  Number  No  1.48%  18  Yes  98.52%  1198     Total  1216  Table A66  END7. ENUMERATOR: WHAT WAS THE MAIN  LANGUAGE USED DURING THIS INTERVIEW?  Percent  Number  Liberian English  78.13%  950  Gio  5.67%  69  Mano  16.04%  195  Other  0.16%  2     Total  1216  90