Recent research has indicated that diffusion model analyses allow the user to decompose the traditional IAT effect (D measure) into three newly developed IAT effects: IAT v , which has already been shown to be significantly related to the construct-related variance of the IAT effect, and IAT a and IAT t0 , both of which have been assumed to provide an indication of faking. But research on the impacts of faking on IAT v , IAT a , and IAT t0 is still warranted. By reanalyzing a data set containing both faked and unfaked IAT effects, we investigated whether diffusion model analyses could be used to separate construct-related variance from faking-related variance on the IAT. Our results revealed that this separation is not yet possible. As had already been shown for the traditional IAT effect, IAT v was affected by faking. Interestingly, it was affected by faking only under more difficult faking conditions (i.e., when participants were asked to fake without being given recommended strategies for how to do so, and when they were requested to fake high scores). By contrast, IAT a was affected by faking only in the comparably easy faking condition (i.e., when participants had been informed about possible faking strategies and were asked to fake low scores). IAT t0 was not affected by faking at all. Our results show that although diffusion model analyses cannot yet provide a clear separation between construct-and faking-related variance, they allow us to peer into the black box of the faking process itself, and thus provide a useful tool for analyzing and interpreting IAT scores.

The Implicit Association Test (IAT) is a popular and frequently used measure in research on implicit associations. However, an important drawback of the traditional computation of IAT results with the so-called D measure is that the D measure may verifiably include more than just indications of the implicit associations that should be measured. It can also be contaminated by faking and other sources of variance. The D measure does not differentiate between different sources of variance. With the help of diffusion model analyses, IAT results can be analyzed and interpreted in a more detailed manner because three separable IAT effects (i.e., IAT v , IAT a , and IAT t0 ) can be computed from the parameters from diffusion model analyses. These effects have been assumed to separate faking-and construct-specific variance from each other. Thus, a possible advantage of using diffusion model analyses instead of the traditional IAT effect is that less contaminated and more interpretable IAT effects are produced (i.e., IAT v , which captures the construct-related variance; IAT a and IAT t0 , which capture the faking-specific variance). This paper was written to demonstrate how to use the software fast-dm to compute these three newly developed IAT effects and to describe how to interpret them.