Children left behind: self-confidence of pupils in competitive environments

ABSTRACT

To sort pupils at the end of primary school, some early-tracking systems apply a mechanism that unwittingly divides classes into two groups: students preparing for exams to enter better schools and everyone else, who decide not to compete for selective schools. Utilizing TIMSS data and a follow-up study in the Czech Republic, we show that this environment has a detrimental effect on pupils’ self-confidence in mathematics, particularly among girls, who do not apply for selective schools but have peers in their classroom who do apply. Our results imply that gender gaps in self-confidence can result from school competitive environments.

KEYWORDS:

• Early tracking
• gender differences
• inequality aversion
• mathematics
• self-confidence

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Michael L. Smith http://orcid.org/0000-0001-8125-2282

Notes

1 Bandura (1997) posited that self-confidence in one's own capacities was the result of observed performance and other social-psychological factors (Usher and Pajares 2009). The previous literature also discussed the problem of nature vs. nurture in gender differences in competitiveness (Niederle and Vesterlund 2010), but not self-confidence as such.

2 Previous evidence and our data suggest that parental background and 4th grade GPA are the main predictors of application decisions. It is therefore reasonable to assume that 11 year olds have a rather passive role in the decision to apply, and no decision to apply would be made without the support of their parents.

3 A sub-sample of the 5th graders – i.e. 2837 students – were further interviewed in subsequent school years. To examine the long-term effects of tracking, we provide also the main descriptive statistics for 6th graders.

4 Here, we refer to the loss of classmates to selective academic junior high schools. Pupils can also experience the loss of classmates e.g. when they move to a different city. Because such changes are few and heterogeneous, we do not include them in our analysis.

5 The math test score from TIMSS is normalized to a mean of 0 and a standard deviation of 1.

6 The variable for GPA is constructed as an average of final grades from mathematics, Czech language and a foreign language. In the Czech Republic, grades are distributed on a 5-point scale from 1 (the best grade) to 5 (the worst grade).

Additional information

Funding

This research was supported by the Czech Science Foundation (Grantová agentura České republiky) [grant number P402/12/G130 for the first author, and grant number 14-36154G for the second and third authors].