Believability Bias in Judging Memories for Abuse

Participants (N=337) were presented with four vignettes in which an adult confided to a friend about being sexually or physically abused at age 9 by either a stranger or father. The memory was presented as either continuous or recovered. Participants judged report believability, memory accuracy, and rated each incident on a scale of 0=not abuse to 5=definitely abuse. Analyses were conducted using a 2(continuous or recovered memory) x 2 (victim sex) x 2 (physical or sexual abuse) x 2 (stranger or close perpetrator) repeated measures ANOVA. Participants completed the Dissociative Experiences Scale (DES; Bernstein & Putnam, 1986), Ambivalent Sexism Inventory (ASI; Glick & Fiske, 1996), and Brief Betrayal Trauma Inventory (BBTS; Goldberg & Freyd, 2003). The believability bias hypothesis was supported. Continuous memory was believed more (p<.0001) and rated more accurate (p<.0001) than recovered memory, and male victims were believed more than female victims (p=.05). Level of dissociation was positively correlated with likelihood to label "being made to have sex with" or "being beaten with a belt" as abuse (p<.01), and level of sexism was negatively correlated with labeling these actions as abuse (p<.02). Implications are discussed in relation to biased and unscientific public opinion about memory for abuse.Supported in part by the Trauma and Oppression Research Fund at the University of Oregon Foundation