I already have the recommended algorithm to analyse IAT data. It allows to compare within-subject differences in response latency in the congruent versus non-congruent condition. However, in addition to analyzing the within-subject differences, I also need to know if the differences between two groups are statistically significant. How should I go about it?
I assume IAT gives you a score for each person.
I imagine you could just do an independent groups t-test to assess whether the means of the two groups significantly differ.