The Effect of Simulator Platform Motion on Pilot Training Transfer: A Meta-Analysis

Abstract

We used a meta-analysis (MA) approach to generate an estimate of true mean effect size (d) for simulator motion with regard to fixed-wing pilot training transfer. We based the analysis on the techniques developed by Hunter and Schmidt (1990). We used a d statistic for effect size calculations based on information available in the included sources. We reviewed and considered 11 studies for analysis, but only 7 of these included the information necessary for calculating effect size and we included these in the study. The result of the MA suggest a small, positive effect for motion, d = .16. No credibility interval could be built around this estimate of population mean effect size because the resulting sampling error variance was larger than the observed variance in d across the assessed studies. This led to a negative variance estimate for d and subsequently an estimated SD_d of 0. These results suggest that simulator motion has a small, positive effect on pilot training transfer and contradict an earlier MA on the same subject. The small sample size (few studies) and methodological shortcomings within the included studies require that the findings be interpreted cautiously. We discuss alternative interpretations and their implications for the aviation training community.