To Whom Does One Aspire? The Effect of Same Gender STEM Role Models on Aspirations and Performance.

2017-03-14

2017-04-17

Survey measures of educational aspirations and growth mindset;
performance in national standardised Maths tests at the end of primary school.

Educational aspirations are measured as the academic degree a participant wishes to achieve in their life time.
Growth mindset is characterised by embracing mistakes as part of learning. I measure this as children's attitude towards important exams.

I study the effect of role models on Maths aspirations and performance, and how the effect differs for same gender and same ethnicity role models relative to "non-relatable" role models. To investigate this question, I run a randomised controlled trial in English primary schools. Treatment schools are randomly assigned to one version of a TED-like video featuring a STEM professional. The content of all versions of the video is very similar, but gender and ethnicity of the speaker vary across different versions. I compare Maths aspirations and test scores in end of primary school standardised exams of pupils who receive a same gender or same ethnicity role model to pupils who receive a speaker of different gender and ethnicity, and to pupils in the control group. This allows to trace out the effect of role models on aspirations and performance, and further the difference in impact of same gender and same ethnicity role models relative to role models of different gender and ethnicity.

The experiment took place in 42 state-funded primary schools across England. The participants are year 6 pupils, on average aged between 10-11. Half of the schools in the sample were randomly assigned to the treatment group, the other half to control. Randomisation took place at the school level, rather than class level, to avoid contamination of the control group. The control group is a pure control group with no experimental intervention. All treatment schools were randomly assigned one version of the treatment video, featuring a TED-like speaker of varying gender and ethnicity. All year 6 classes in one school watched the same video.

The treatment videos are identical in the first part and different in the second: The first part highlights the importance of Maths in various careers, ranging from robotics engineering to graphic design. This part is animated and identical across all six versions of the video. The aim is to get pupils' attention and introduce the message that Maths is useful in a range of careers in playful, non-stereotyped manner. The animation uses snaps from popular animation films and features a friendly, animated commentator to minimise the educational feel of the video.

The second part of the video varies in that each version features a speaker's personal Maths story, the relevance of Maths in their career, and difficulties and successes they experienced studying and applying Maths. This is to raise aspirations and strengthen children's self-efficacy for Maths, ie. their belief in their own abilities to develop Maths skills. Focus groups prior to the experiment have shown that the idea of either having or not having a Maths brain is very prevalent amongst pupils. Highlighting speakers' difficulties with Maths in the past aims at fighting this belief and promoting a Growth Mindset. Each speaker gives a genuine account of their experience while all versions follow the same protocol of questions. The six speakers are either female or male, and belong to one of the three largest ethnic minorities in English primary schools: White British, African / Black Caribbean or South Asian. Each version of the video features only one speaker.

I administered a baseline survey in all 42 participating schools during December 2016 and January 2017. Video screenings in all but three schools took place between 14th and 31st of March 2017, the three remaining interventions took place during the week commencing 17th of April -- this delay is owed to spring term break taking place during the first two weeks of April. I collected follow up survey data during this time in all control schools, and in treatment schools after their respective screening dates. At the end of spring term, in the week from 8th to 12th of May 2017, all year 6 pupils take national standardised tests (SATs) in Maths, Reading, Writing and Grammar, Punctuation and Spelling (GPS). Test scores pupils obtained in the Maths section of SATs will serve as my main outcome measure of Maths attainment.

Randomisation stratified by prior Maths attainment. Done by a computer.

Randomisation into treatment and control group by school. Same gender and ethnicity treatment within treatment schools by pupil.

Yes

42 schools

1500 pupils

21 control schools, 21 treatment schools. 50% of pupils in treatment schools receive same gender treatment, 25% same ethnicity.

0.4 standard deviations.