Minorities are underrepresented in the science workforce yet adequately represented as players of computer games. Findings in career development research suggest that a decision to pursue a science career is directly impacted by a person's investigative self-efficacy. Because minority students choose to spend a significant amount of time playing computer games this study examines the effects of computer games on investigative self-efficacy.

The dissertation is composed of a systematic literature review, the development of a theoretical framework, and an application of the theoretical framework in a quasi-experimental study. In the systematic literature review, the small-to-moderate effect sizes of the 6 systematically identified studies suggest that elements in computer games can potentially affect self-efficacy. Unfortunately, the similarities across the small number of studies makes it difficult to generalize the results to other settings and content areas while variability across the studies makes it difficult to pinpoint which computer game elements or type of computer games affect self-efficacy.

An exploration of theories and empirical research in cognitive psychology, career development, and performance in complex environments led to a theoretical framework. The theoretical framework integrates attention, flow, and self-efficacy theories as well as the results of Berry and Broadbent's (1988) study that compared the effects of implicit and explicit instructions on performance. Using the theoretical framework developed in this dissertation, stealth educational games are proposed as an option for building the investigative self-efficacy of unmotivated or academically struggling learners.

The effect of stealth educational games on minority students' investigative self-efficacy was explored. Based on the statistical results in this study and the differences across each of the schools, the potential value of stealth educational games is still unknown. Future research should employ theory to systematically document and define the context in which the game is delivered, incorporate assessments built into the game instead of using surveys, include incentives for student participation and obedience, and compare the effects of a stealth educational game to an explicitly educational game.