This article presents collaborative work of proponents (T.P., J.N.M.) and a skeptic (A.B.) of the recognition heuristic. Although minor disagreements remain on aspects of interpretation and emphasis, the text represents compromise statements that are acceptable to all authors.
The recognition heuristic in memory-based inference: is recognition a non-compensatory cue?†
Article first published online: 30 OCT 2007
Copyright © 2007 John Wiley & Sons, Ltd.
Journal of Behavioral Decision Making
Volume 21, Issue 2, pages 183–210, April 2008
How to Cite
Pachur, T., Bröder, A. and Marewski, J. N. (2008), The recognition heuristic in memory-based inference: is recognition a non-compensatory cue?. J. Behav. Decis. Making, 21: 183–210. doi: 10.1002/bdm.581
- Issue published online: 26 MAR 2008
- Article first published online: 30 OCT 2007
- recognition heuristic;
- noncompensatory strategies;
The recognition heuristic makes the strong claim that probabilistic inferences in which a recognized object is compared to an unrecognized one are made solely on the basis of whether the objects are recognized or not, ignoring all other available cues. This claim has been seriously challenged by a number of studies that have shown a clear effect of additional cue knowledge. In most of these studies, either recognition knowledge was acquired during the experiment, and/or additional cues were provided to participants. However, the recognition heuristic is more likely to be a tool for exploiting natural (rather than induced) recognition when inferences have to be made from memory. In our study on natural recognition and inferences from memory, around 85% of the inferences followed recognition information even when participants had learned three cues that contradicted recognition and when some of the contradictory cues were deemed more valid than recognition. Nevertheless, there were strong individual differences in the use of recognition. Whereas about half of the participants chose the recognized object regardless of the number of conflicting cues—suggestive of the hypothesized noncompensatory processing of recognition—the remaining participants were influenced by the additional knowledge. The former group of participants also tended to give higher estimates of recognition's validity. In addition, we found that the use of recognition for an inference may be affected by whether additional cue knowledge has been learned outside or within the experimental setting. Copyright © 2007 John Wiley & Sons, Ltd.