Response to Pashler et al. (2016)

ABSTRACT

Pashler, Rohrer, Abramson, Wolfson, and Harris (2016/this issue) in their critique of Chatterjee, Rose, and Sinha (2013) data argue that (a) large effect sizes in Study 1 and Study 2 undermine the credibility of the data; (b) in a lexical task that is part of Study 3, a concentration of participants in (5,0) and (0,5) benefit/cost word data points and the similarity of 9 neutral word responses at these points are extremely unlikely; and (c) there are apparent errors in the execution of Study 3. In this response, I examine these issues in detail. A recent review of money prime literature (Vohs, 2015) notes many effect sizes that are as big as or bigger than ours. Although 8 coding errors were discovered in Study 3 data and this particular study has been retracted from that article, as I show in this article, the arguments being put forth by the critics are untenable. For instance, my analysis shows that results hold even without the concentration of (5,0) and (0,5) participants, and I offer statistical simulations to counter critics’ arguments. Regarding the apparent errors in Study 3, I find that removing the target word stems SUPP and CE do not influence findings in any way. I also report findings from MacDonnell and White (2015), who replicate the basic finding of Chatterjee et al. in a different context.