Correcting for Self-selection Bias in Business Ethics Research

Abstract

Self-selection bias occurs when there is non-random sampling of membership within a group or category, such as employment status, that is hypothesized to affect a variable of interest, such as ethical attitudes or behaviors. Self-selection bias is germane to a variety of important business ethics questions, such as how the business environment affect personal ethics or whether business students are more or less ethical than non-business students. This paper describes an empirical technique to control for self-selection bias. The technique is applied to the empirical question of whether the employment status of workers is correlated with ethical proclivities. Correcting for self-selection bias shows that an observed negative correlation between employment and ethics is the result of self-selection rather than factors associated with employment, other things being equal.