Evaluation of Breakthrough’s school-based attitude change programme in India

Publication Details

Jayachandran, S, Jain, T and Dhar, D, 2018, Evaluation of breakthrough’s school-based attitude change programme, Haryana, India, 3ie Grantee Final Report. New Delhi:  International Initiative for Impact Evaluation (3ie)

Link to Source
Seema Jayachandran, Tarun Jain and Diva Dhar
Institutional affiliations
Northwestern University, Indian School of Business, Bill & Melinda Gates Foundation
Grant-holding institution
Institute for Financial Management and Research
South Asia
None specified
Gender analysis
None specified
Gender analysis
Equity Focus
None specified
Evaluation design
Randomised Control Trials (RCT)
3ie Final Grantee Report
3ie Funding Window
Policy Window Round 2


This evaluation assessed the impact of a school-based programme in changing girls’ aspirations, gender attitudes, and behaviours among adolescent boys and girls in Haryana, India. Approximately 300 government schools across four districts participated in the evaluation. Haryana, a state in north India, has one of the country’s worst sex ratios stemming from deep-seated gender discrimination against girls.


Home to the world’s largest youth population, it is important that India targets adolescents to change gender norms and make them more equitable for girls and women.

Breakthrough, a non-profit human rights organisation, worked with the Education Department in Haryana to enroll government schools in Sonepat, Panipat, Rohtak and Jhajjar districts into the programme.

The programme included multiple activities. In participating schools, Breakthrough facilitated the setting up of a youth club, Taaron ki Toli (Cluster of Stars). Students would participate in its activities, which would emphasize messages on gender equality through the use of posters and street performances. In turn, students received an activity workbook, and branded materials, such as caps and badges with the club’s insignia on them.

Trained facilitators visited schools once every 2-3 weeks and conducted 45-minute long sessions based on the curriculum developed by Breakthrough. Spread across 28 interactive classroom sessions, it covers gender identity, values, aspirations, goals, roles and stereotypes, and recognition of and tolerance towards discrimination. It also included interpersonal skills, such as public speaking, communication and social interaction between the sexes, negotiation, presentation, assertiveness, leadership, self-efficacy and trust-building. Interactive activities included writing letters and stories, recording observations, street theatre, games, sports, video vans, short school campaigns and dialogue with families. Using these approaches, students explored gender identity and stereotypes to better understand gender inequities and their consequences, understand their rights and entitlements, and were encouraged to communicate and act on what they had learnt.

Breakthrough also conducted two community mobilisation activities by running video vans across all intervention districts in 2015 and 2016.  The vans showcased the work that the students had done as a part of the youth club.

Research questions

The main research question was whether a gender attitude change programme has an impact on attitudes, educational and occupational aspirations, and gender-related behaviours, and if so by how much.

To understand the mechanisms through which the impact takes place, the study also analysed which gender attitudes were affected the most, as well as how the effects varied by students’ sex, parents' gender attitudes and other characteristics.


The evaluation used a randomised design, with schools as the unit of randomisation.

Schools were selected from a total of 607 government-run secondary schools that offered grades six through nine in the four districts. From this group, schools with medium to high enrolment based on DISE (2011) data and with low dropout rates between grades were shortlisted. In villages with multiple schools, only one school per village was randomly selected. Visits to schools revealed very low attendance in some schools, despite high enrolment. These schools were also dropped. After this screening, 314 schools qualified to be a part of the sample, with 150 in the treatment arm and 164 in control arm.

By targeting government schools, the final sample had disproportionately high participation from girls and students from poorer families. In Haryana, children from better off households, and boys in general, are more likely to attend private schools.

The team collected baseline data in 2013-2014, followed by two midlines in subsequent years and the endline in 2016-2017. The baseline data included a survey of the students, parents and the schools. Additionally, an Implicit Association Test was administered to students. They were asked to associate images of boys and girls with ‘good’ and ‘bad’ characteristics reflecting gender norms. These surveys were also repeated at the endline.

Students in the classroom activities were also observed through facilitated discussions and poster making to understand students’ comfort level in interacting with the opposite sex and beliefs in girls’ competence held by boys and the girls. This was done for 197 schools for which permission could be secured.

They carried out a qualitative study, in which 15 schools participated during the baseline and 12 schools during the endline.

Main findings

The study looked at three primary outcomes: changes in gender attitudes, changes in gendered behaviours and changes in girls’ aspirations.

The study found that students in treatment schools were likely to have more progressive gender attitudes as measured by attitude towards opportunity for education, employment outside home, women’s role in society and fertility behaviour.

The intervention increased girls’ aspirations for future education and non-traditional occupations, although the effect is much smaller and is not robust when the sample is restricted to respondents below the median social desirability score.

The intervention was successful in reducing gender-biased and gender-stereotyping behaviour among both boys and girls. Measurement of gender behaviour is derived from interaction with the opposite sex, participation in household chores, supporting the aspiration of girls and women in the household, girls’ decision-making and girls’ mobility.

The intervention generated more interaction with the opposite sex for both boys and girls, with a larger impact for girls. It also led to greater mobility (i.e. walking to school alone) for girls, but had no impact on girls’ decision-making power. Boys in the treatment arm helped out more with household chores, though there was no change for girls. They continued to help at the same rates. There are also large increases in boys encouraging females in the family to pursue higher education and careers.

Given that self-reported measures are subject to concerns about social desirability bias, the team observed activities in schools to measure girls’ participation in classroom discussions, students’ belief in girls’ knowledge and ability (based on whom they vote for to represent the class in a quiz competition) and interaction with age peers of the opposite sex. Interestingly, for both participations in class discussions and voting for girls to participate in the quiz competition, girls are not underrepresented in the control group. On interaction with the opposite sex, while it cannot be rejected that the intervention did not affect this outcome given the extremely low means in the control group, it is difficult to say anything conclusive.

Scroll to Top