Abstract:

Prejudices play an important role in many aspects of everyday life. They facilitate orientation and help us assess risks and opportunities. On the downside, they often form the basis for (negative) discrimination and are therefore rightfully under scrutiny. In an employment context, managers and recruiters must regularly make inferences about the expected performance of a (potential) employee based on what little information they have on this person, without resorting to (unfair) prejudices. Similarly, in various stages of education, students tasked with selecting partners for a team exercise hope to pick strong individuals but often have nothing but prejudices on which to base their assessment.

This paper empirically investigates the prevalence and determinants of prejudices among n = 90 undergraduate students of Human Resource Management at the University of Hamburg, Germany. For this purpose, we rely on a narrow and neutral concept of prejudice: person A’s expectation about person B’s performance in relation to a task, based on A’s socio-demographic characteristics and A’s observation of B’s characteristics. We also look at the case of A = B, i.e. a person’s performance expectation about herself, though most would consider this not a prejudice but rather an expression of confidence, self-efficacy, or related concepts. The students were given a multiple-choice task with questions relating to prior teaching content. They were then asked to estimate their own performance in relation to the task, as well as the performance of the person sitting next to them – their arbitrarily assigned ‘partner’. Finally, we collected some socio-demographic data on the students, with gender being our main variable of interest.

Amongst other significant effects, the results suggest that:
(1) Students with a migration background achieve lower scores on the task but are not fully aware of this fact, and they overestimate the performance of others.
(2) Children of highly educated parents strongly overestimate the performance of their ‘partners’ (though this may just be an expression of what they consider to be good manners).
(3) The performance of students who appear to be well-off financially is significantly overestimated on average (this may be the rationale behind ‘dress for success’).
(4) Male students perform less well than females but in fact think that the opposite is true, meaning that they strongly overestimate their performance. Yet males also significantly overstate the performance of their partners, irrespective of their gender. A positive non-result is that there is no discernible tendency for members of either sex to disparage the performance of the other sex.

Our discussion focuses on the finding of male overconfidence and the implications it may have – to the extent that it can be generalised – for women who compete with men for rewards, responsibility, jobs, etc. We conclude that, firstly, modesty is not a promising strategy for women. Secondly, an educator, a recruiter, a manager etc. who compares male versus female statements about performance, be it about the person herself or about others, must crucially be aware of these systematic differences in assessment that may not reflect performance differentials.

Keywords:

Performance expectations, prejudices, overconfidence, gender, migration background.