ABSTRACT
The level of significance was set at ex = .05 unless otherwise noted. Given the preliminary exploratory nature of this experiment, alpha levels for post hoc analyses remained at .05; this has the effect of increasing the probability of committing a Type-I error, but this procedure was considered more acceptable than committing a Type-II error and missing future potential avenues of study. When performing chi-square analyses, Fisher's Exact Test was used when proportions were more extreme than 80-20 or when cell membership for any given cell was ::;5. RT data was analyzed for the presence and elimination of outliers using Cook's D procedure. Due to the L-shaped distribution, RTs were converted to a logarithmic scale, but this failed to produce a normal distribution. The RTs were then inverted (i.e., l/X), which satisfactorily approximated a normal distribution. Subsequent analyses on RTs were conducted with the inverted RT data. RTs for practice trials and trials during which a blue word (PM cue) was presented were not included in the analyses.
