In the present study, we used a new approach to establish the smallest effect size of interest (SESOI) for false memory research by asking memory researchers what they considered to be the SESOI in false memory research. They were presented with three hypothetical and three influential paper scenarios. These scenarios depicted studies examining the effects of certain manipulations (e.g., therapy) on false memory formation using well-known false memory paradigms: Deese/Roediger-McDermott, misinformation, and forced fabrication. Subsequently, they were asked for each scenario what they would consider to be the SESOI for practical and theoretical purposes and justify their decisions. We found that there was no clear consensus for the SESOI. However, memory researchers tended to accept smaller SESOIs or "any difference that leads to a p < .05," especially for theoretical ends. We argue that the lack of a general consensus is acceptable as long as proper justification is used. We discuss such rationales and provide recommendations for setting the SESOI.

No abstract

Memory experts are sometimes asked to evaluate the validity of accounts of witnesses, victims, or suspects. In some of these cases, they are asked what effect alcohol has on the validity of such accounts. In this article, we offer a guide on what expert witnesses can reliably say about how alcohol affects memory. We do so by resorting to effect sizes from previous studies and meta-analytic work, and address this novel question: Are these effect sizes meaningful in legal cases? More specifically, we argue that any determination of whether individual studies about alcohol and memory are practically relevant for legal cases, scientists must focus on the smallest effect size of interest. We make the case that a decrease or increase of only 1 detail, especially an incorrect detail, should be regarded as the smallest effect size of interest in this line of research. In line with this idea, we show that effect sizes in the alcohol and memory literature are often larger than this smallest effect size of interest. This finding is important because it implies that alcohol often exerts a practically relevant and meaningful detrimental effect on the reporting of both correct and incorrect details, which in turn negatively affects the validity of witness testimony.