Policy Brief Issue 5 | November 2018 Why Do Indonesian Adolescent Boys have Poorer Schooling Outcomes than Girls? EAST ASIA AND PACIFIC GENDER INNOVATION LAB KEY FINDINGS The East Asia and • In a sample of 8th grade Indonesian students, boys have on Pacific Gender Innovation Lab average lower grades and school attendance than girls. (EAPGIL) carries out • Differences in mindsets, socioemotional skills, and other behavioral factors impact evaluations and inferential research to may drive gender differences in schooling outcomes at this level. generate evidence on what works in closing gender gaps in assets, CONTEXT economic opportunities, and agency, and how Indonesian secondary students perform worse academically than their peers in other closing these gaps countries, especially boys. In the 2015 Program for International Student Assessment can help achieve other (PISA) tests, Indonesia ranked among the worse of the 72 participating countries. More than development outcomes. half of 15-year-olds could read a text but could not answer simple questions related to it; Ultimately, EAPGIL that was only the case of 14 percent of students in high-performing Vietnam and 20 percent seeks to increase the welfare of women in member countries of the Organization of Economic Cooperation and Development and men in East Asia (OECD).1 While Indonesian boys and girls had similar average scores in math and science, and the Pacific by girls outperformed boys in average scores of reading.2 promoting the uptake of effective policies and Adolescent boys may learn less because they have lower socioemotional programs identified skills, mindsets that demotivate them, and other behavioral factors. Students based on evidence. with more of a growth mindset, namely who believe that they can increase their abilities in school through dedication and hard work, are more motivated, indeed 1 OECD. 2016. PISA 2015 Results (Volume I): Excellence and Equity in Education. Paris: OECD Publishing. Indonesian boys’ average scores in math, reading, and science, were respectively 385, 386, and 401. Girls’ were 2  387, 409, and 405. work harder, and get higher test scores in diverse outcomes and long-term wellbeing, such as future contexts such as Chile and the United States.3 job opportunities and income.7 To our knowledge, Students with higher socioemotional skills, the set of this is the first attempt to measure socioemotional attitudes and behaviors to manage personal and social skills of Indonesian adolescents at such a scale.8 situations, also have better educational achievement We use these data to explore the relationships and cognition. Adolescence is a crucial period in 4 between mindsets, perceptions, socioemotional which mature mental abilities and socioemotional skills, schooling outcomes, and gender. While we skills develop and influence both learning and future cannot establish causal relationships, the data outcomes.5 Other behavioral factors that could provide some patterns which may be policy- negatively influence learning — and gender differences relevant. — are students’ perceptions of a negative classroom environment for learning, low educational aspirations, WHAT DID WE FIND? and behaviors that are not conducive to studying. Eighth-grade boys have lower grades and miss more classes than girls. In our sample, WHAT DID WE DO? boys’ average grades are below girls’ in all seven A team of researchers from the World Bank and the subjects, with differences ranging between 2 and University of California, Davis, in cooperation with 3 grade points, the equivalent of 0.34 and 0.55 the Ministry of Education and Culture of Indonesia of a standard deviation (Figure 1, panel A).9 Most as well as local education authorities, collected data girls, rather than just a few excellent students, on mindsets and socioemotional skills, studying drive these differences since girls not only have practices, education and career aspirations, grades, higher grade averages but have better grades all and family socioeconomic characteristics from 56,000 along the distribution. Besides, while secondary 8th grade students studying in 2,100 public schools school-aged boys and girls are enrolled in a on the Indonesian islands of Java and Sumatera. 6 similar proportions, enrolled boys in the survey These islands are home to 70 percent of the are more likely to miss school: 55 percent of boys population. The data are the baseline for an impact miss at least one day per month compared to 43 evaluation of a pilot intervention aimed at improving percent of girls.10 Apart from being sick, boys are students’ mindsets and socioemotional skills. more likely to miss school because of all the other Through support to improved socioemotional skills, reasons, including being suspended, being bored, the intervention may also positively affect schooling and working for the family (Figure 1, panel B). 3 C laro, S., D. Paunesku, and C. S. Dweck. 2016. “Growth mindset tempers the effects of poverty on academic achievement.” Proceedings of the National Academy of Sciences 113(31): 8664–8668. West, M. R., M. A. Kraft, A. S. Finn, R. E. Martin, A. L. Duckworth, C. F. Gabrieli, and J. D. Gabrieli. 2016. “Promise and paradox: Measuring students’ non- cognitive skills and the impact of schooling.” Educational Evaluation and Policy Analysis 38(1): 148–170. OECD. 2015. Skills for Social Progress: The Power of Social and Emotional Skills. OECD Skills Studies. Paris: OECD Publishing. 4  Farrington, C. A., M. Roderick, E. Allensworth, J. Nagaoka, T. Seneca Keyes, D. W. Johnson, and N. Beechum. 2012. “Teaching Adolescents to Become Learners: The Role of Noncognitive Factors in Shaping School Performance.” Chicago, IL: University of Chicago Consortium on School. 5 Guerra, N., K. Modecki, and W. Cunningham. 2014. “Social-Emotional Skills Development across the Life Span: PRACTICE”. Policy Research Working Paper 7123. Washington, DC: World Bank. 6  The sample excludes Islamic schools and technical or remedial schools, which approximately 13 percent of students attend. However, it is representative of students attending public, secular junior secondary schools in Java and Sumatera. 7 For the details of the intervention, please visit www.worldbank.org/eapgil. 8  A pilot study of the currently-evaluated intervention collected measures of socioemotional skills for a smaller sample of 114 students. 9 As a benchmark, grades can range from 0 to 100, with 75 being the passing grade in most subjects. Girls’ higher 2–3 grade points compared to boys is a large difference given that only 15 percent of students have grades below 75 — the passing grade — and that three quarters of students have grades within a 10-point range between 75 and 85. Boys also miss more days on average: boys miss 1.5 days on average in the past month of the survey while girls miss 1.0 days. Boys and girls who miss at 10  least one day of class miss an average of 2.7 and 2.3 days, respectively. Boys lower grades and higher absenteeism may suggest that there is a strong positive association be partially related to lower socioemotional skills, between students’ mindsets, perceptions, and demotivating mindsets, and how they feel in their socioemotional skills, and their grades.12 The classroom environment. Apart from self-esteem, association is especially strong for growth mindset boys in our sample have significantly lower levels and learning orientation (the perception that effort in all the mindsets, perceptions of their classroom to learn is valued in their classroom): an increase environment, and socioemotional skills measured in of one standard deviation in growth mindset and the survey (see table A.1. for the definitions). Data learning orientation correlate on average with an FIGURE 1. BOYS LAG GIRLS IN SCHOOLING OUTCOMES A. Average grades of 8th grade students, by gender 85 Boys Girls 84 83 82 81 Grade (0-100) 80 79 78 77 76 75 Mathematics English Science Social Science Citizenship Bahasa Religious Average Language Education Indonesia Education B. Share of 8th grade students who missed school in the past month by the main reason for missing school, by gender 90 Boys Girls 80 70 Percentage of students who missed school last month 60 50 40 30 20 10 0 School Boredom Missing Natural Transport Work to Bring Taking Care of a Help with Sickness Suspension Teacher Disaster Issues Money Home Family Member Family Work STUDENT’S BEHAVIOR SCHOOL AND WEATHER FACTORS FAMILY’S DEMANDS HEALTH Notes: Bars represent the average for both gender. In panel A, grades range from 0 to 100, with an average of 79 and a standard deviation of 5. Panel B only includes students who missed school at least one day in the past month (48 percent of all students). Students who miss several days may respond several reasons. All differences between boys and girls in both panels are significant. Source: Survey “Getting to Know 8th Grade Students”, February-March 2018.11 11 As all the data presented in the graphs are from the “Getting to Know 8th Grade Students,” references to the source will be omitted henceforth. The association between grades and performance-avoidance orientation is negative given that the later reflects students’ propensity to hide mistakes 12  and pretend they know material (see definitions in table A.1.). Hiding mistakes is not conducive to learning and one may consider that lower performance- avoidance orientation is desirable. increase of 0.7 and 0.6 grade points of the average mindset and higher learning orientation are also highly grade (increases of 0.9 and 0.8 percent, respectively), associated with lower absenteeism. This is the case even when holding fixed other mindsets, skills, for a higher sense of belonging to the school as well: perceptions, and confounding factors such as their an increase of one standard deviation in the score gender, age, and where they live (Figure 2, panel A). of sense of belonging is associated with missing 0.1 Each of the skills is also associated with a reduction fewer days of school per month (a reduction of 8 in absenteeism (Figure 2, panel B). More of a growth percent of the 1.2 days that are missed on average).13 FIGURE 2. BETTER MINDSETS, PERCEPTIONS, AND SOCIOEMOTIONAL SKILLS, ARE ASSOCIATED WITH HIGHER GRADES AND LOWER ABSENTEEISM Panel A. Change in average grade point associated with a change in mindsets, perceptions, and socioemotional skills 0.9 0.8 mindset, perception, and socioemotional skill Changes in grade points associated with one standard deviation increase in the score of 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 -0.1 -0.2 -0.3 -0.4 Performance Self-esteem Perserverance Effort Beliefs Sense of Learning Growth Mindset Avoidance Orientation Belonging Orientation Panel B. Change in number of missed school days associated with a change in mindsets, perceptions, and socioemotional skills 0.08 Changes in the number of missed school days associated 0.06 with one standard deviation increase in the score of mindset, perception, and socioemotional skill 0.04 0.02 0.0 -0.02 -0.04 -0.06 -0.08 -0.10 -0.12 -0.14 Performance Effort Beliefs Perserverance Self-esteem Sense of Growth Mindset Learning Avoidance Orientation Belonging Orientation Notes: Bars are coefficients from a regression of an outcome of interest — average of grades in panel A, number of missed school days in the past month in panel B — on standardized measures of mindsets, perceptions, and socioemotional skills that keep constant other measures and four demographic factors (age, sex, location in urban or rural areas, and location in Java or Sumatra). Slim bars are confidence intervals for the estimates at the 95-percent level; Coefficients are statistically significantly different from zero if they do not overlap the zero line. Standardized measures of mindsets, perceptions, and socioemotional skills have a mean of 0 and a standard deviation of 1. In panel A, grades can range from 0 to 100, with an average of 79 and a standard deviation of 5. All coefficients are significantly different from zero at the 99-percent level but that of self-esteem and perseverance, which are not significant, even at the 90-percent level. In panel B, all coefficients are significantly different from zero at the 99-percent level but that of effort belief, which is not significant, even at the 90-percent level. 48 percent of students in the survey missed at least one school day in the past month. Those who missed school missed 2.5 days on average. 13  Lower educational aspirations may also contribute compared to less than 10 percent of girls). Consistent to lower levels of boys’ learning. Educational with evidence from other countries, our survey results aspirations are the future years of education students suggest that higher aspirations are associated with wish to achieve and that drive their effort to reach better schooling outcomes. Students who, in the that goal.14 Adolescent students aspiring to complete absence of other constraints, wish to complete secondary and tertiary education in Ethiopia and India, postsecondary school have higher average grades respectively, ended up completing more schooling by 1.6 points and lower absenteeism by 0.3 days than those with lower educational aspirations. Given 15 compared to those who only wish to complete high- the strong link between educational aspirations, effort school or a lower level of education, when controlling and school attendance, it is possible that higher for demographic factors. aspirations may also be linked with greater learning Boys are less likely to engage in behaviors while school and better grades. conducive to learning. Girls report spending more In our survey, 84 percent of 8th grade girls state that time studying outside of school than boys: more they would like to complete a general tertiary degree, than five hours per week compared to less than four and more than half state that they would like to get a hours for boys. Girls also consistently receive more PhD if they had no constraints. Among 8th grade boys, homework help at this education level from family just 64 percent aspire to general tertiary education members and others: 83 percent of girls receive help, and 37 percent to a PhD. Boys are more likely to state compared to 78 percent of boys (Figure 3). These that they wish to complete technical or vocational high differences in behaviors may emerge from different school with no further formal education (20 percent social expectations of boys and girls. FIGURE 3. 8TH GRADE GIRLS RECEIVE MORE HELP FOR STUDIES AND STUDY MORE Help for studies received by students from family members and other people 85 Boys Girls 80 75 70 65 Percentage of Students 60 55 50 45 40 35 30 25 Father Other Relatives Non-relatives Mother Brothers and At Least Sisters Someone Note: All differences are statistically significant between boys and girls.  ay, D. 2006. “Aspirations, poverty and economic change.” in A. Banerjee, R. Benabou, and D. Mookherjee (eds.). Understanding Poverty. pp. 409–22. 14 R Oxford: Oxford University Press. Beaman, L., E. Duflo, R. Pande, and P. Topalova. 2012. “Female Leadership Raises Aspirations and Educational Attainment for Girls: A Policy Experiment in 15  India.” Science 335(6068): 582–586. Favara, M. 2017. “Do Dreams Come True? Aspirations and Educational Attainments of Ethiopian Boys and Girls.” Journal of African Economies 26(5): 1–23. POLICY IMPLICATIONS Our data suggest some possible contributing factors driving Indonesian ACKNOWLEDGMENTS boys’ underachievement in grade 8. We observe gender gaps in mindsets, This brief was prepared by socioemotional skills, perceptions of the classroom environment, aspirations, and Noël Muller and Elizaveta help-seeking behaviors. While we cannot establish causal links, our descriptive Perova with inputs from findings point to important directions for future research and policy work. Specifically: colleagues in EAPGIL, 1. There is a need to develop and test further interventions aimed at eMBeD, University increasing mindsets, socioemotional skills, and aspirations. In Indonesia, of California Davis, such interventions could focus on increasing boys’ low aspirations, mindsets, the Poverty & Equity and socioemotional skills, and provide tools to bolster the academic Global Practice, and the performance of both boys and girls. Such interventions are generally low Education Global Practice. cost and easily scalable. The impact evaluation associated with the baseline EAPGIL is supported survey used in this brief will test the effectiveness of two such interventions. through the World Bank The first is a 6-week growth mindset based curriculum that includes comics Group’s Umbrella Facility to introduce the topic of the week, guided classroom discussions, student for Gender Equality (UFGE) reflection, small group discussions, and other interactive activities. The in partnership with the second includes additional tools and activities that classroom teachers may Australian Department of integrate into their daily interactions with students in order to reinforce the Foreign Affairs and Trade. impacts of the 6-week curriculum. The UFGE has received 2. There might be some social expectations that discourage help-seeking generous contributions outside of school and lower academic motivation for boys at this from Australia, Canada, educational level. A better understanding of the gender specific impacts of Denmark, Finland, social expectations on students’ behaviors is needed. Germany, Iceland, the Netherlands, Norway, TABLE A.1. DEFINITION OF MEASURES Spain, Sweden, OF MINDSETS, PERCEPTIONS, AND Switzerland, the SOCIOEMOTIONAL SKILLS IN THE SURVEY United Kingdom, and Mindset, perception, skill Definition the United States. The extent to which students believe that exerting Effort beliefs effort will lead to improved ability The extent to which students believe that they can improve their Growth mindset abilities and become smarter through dedication and hard work The extent to which student perceive that classmates and teachers Learning Orientation value hard work, learning progress, and learning from mistakes FOR MORE INFORMATION Performance Avoidance The extent to which student perceive that classmates and Elizaveta Perova (EAPGIL) Orientation teachers disparage mistakes and lack of knowledge eperova@worldbank.org The extent to which students keep up effort to achieve Perseverance their goals despite difficulty, delays, and failure Renos Vakis (eMBeD) The degree of positive self-evaluation of the student’s rvakis@worldbank.org Self-esteem own worth http://www.worldbank.org/eapgil The extent to which students feel they are accepted Sense of belonging members of the school community Note: Socioemotional skills are a set of learned attitudes and behaviors that allow people to manage personal and social situations effectively. Mindsets are ways in which students perceive their own abilities, in particular regarding how smart they are, and whether they can change it or not. Perceptions of the classroom environment refer to students’ perceptions of what teachers and classmates value in the learning process. Measured constructs that are featured in the table fall into one category or overlap over several ones.