Congruity effects in the survival processing paradigm

Five experiments were conducted to investigate a proposal by Butler, Kang, and Roediger (2009) that congruity (or fit) between target items and processing tasks might contribute, at least partly, to the mnemonic advantages typically produced by survival processing. In their research, no significant survival advantages were found when words were preselected to be highly congruent or incongruent with a survival and control (robbery) scenario. Experiments 1a and 1b of the present report show that survival advantages, in fact, generalize across a wide set of selected target words; each participant received a unique set of words, sampled without replacement from a large pool, yet significant survival advantages remained. In Experiment 2, we found a significant survival advantage using words that had been preselected by Butler et al. to be highly unrelated (or irrelevant) to both the survival and control scenarios. Experiment 3 showed a significant survival advantage using word sets that had been preselected to be highly congruent with both scenarios. Finally, Experiment 4 mixed congruent and incongruent words in the same list, more closely replicating the design used by Butler et al., and a highly reliable main effect of survival processing was still obtained (although the survival advantage for the congruent words did not reach conventional levels of statistical significance). Our results suggest that the null effects of survival processing obtained by Butler et al. may not generalize beyond their particular experimental design.