Economic preferences, school performance and social background: How do they relate? Preliminary results from an incentivized experimental survey in Hungarian schools
(Dániel Horn, Hubert János Kiss, Tünde Lénárd)
In this presentation, we show preliminary results from a grand scale ongoing survey that uses incentivized experimental tasks to measure students’ time, risk, competitive and social preferences, henceforth non-cognitive skills, and link these with previous school test scores (cognitive skills) and family background variables.
In 2018 and 2019, we have visited 53 school groups (classes) in 9 schools to measure the non-cognitive skills of 1108 students. Using vouchers for the school buffet, we incentivize the experiments. Following Falk et al (2018), we use the staircase (or unfolding brackets) method to measure time preferences, and deduct the individual β and δ of students (Laibson, 1997). To test risk preferences we use the bomb-risk elicitation task of Crosetto and Filippin (2013). For testing competitive preferences we use a simple, gender neutral, real effort task (counting zeros) and use the three round measure of competition suggested by Niederle and Vesterlund (2007). We use three different games to test social preferences: the dictator game, the trust game and a simple public good game.
We then, take these anonymously measured preferences and connect them to the administrative panel of the National Assessment of Basic Competencies (NABC) using the hash codes provided by the Education Authority. We gain all additional information on gender, parental background and school performance from the NABC data.
Our sample of students is far from being representative, as the ratio of low status, disadvantaged students in our sample is small. First glance on the data, however, show that most of our measures are in line with the most robust observations in the literature.
In this talk we will introduce our measures, and the research in general, and highlight three potential avenue for further research, sharing preliminary results in each topic.