Adults’ Memories of Traumatic Childhood Experiences

• In Event: Improving Eyewitness Reporting: Novel Approaches Across Development

Abstract

Claims of decades-old child maltreatment raise questions of how to obtain accurate memories from adults about negative childhood experiences. Our research adds theoretical and empirical knowledge to national and international debates about adults’ memory for childhood traumatic events, specifically, child sexual abuse (CSA).

In a longitudinal study, children (3- to 16-year-olds, M = 98.00 months, SD = 40.2) experienced at Time 1 (1990s) a documented child maltreatment investigation. As adults (N = 105; 67% females), they were interviewed two decades (20 years) later (Time 2). Participants were predominately African American (70%), with 11% White (non-Latino/a), 10% Latino/a, and 9% “other,” living in an urban Midwestern city. As adults, they were randomly assigned to: a standard forensic interview; the Cognitive Interview (CI), with mental reinstatement (e.g., close your eyes, think back to that time); or the CI, with mental and context reinstatement (i.e., watching a videotaped recording, from the 1990s, of the forensic unit). We predicted that older Time 1 age, Time 2 CI protocol (especially with both mental and context reinstatement), and Time 2 trauma-related psychopathology would be associated with greater memory accuracy.

Consistent with the interview protocols, participants were asked, “What brought you to the forensic unit?” Memories were coded against Time 1 records, using an 11-item checklist (e.g., relation to the alleged perpetrator, CSA was one time or chronic, drugs or alcohol involved), reliably completed by “naïve” coders (ICCs > .70). Proportions were calculated for correct responses, omission errors (e.g., omitting the alleged perpetrator was a relative, when our records indicated he was an uncle), and commission errors. A square-root transformation was applied to proportions of correct responses and omission errors, due to skew. The Maltreatment Classification System (Barnet et al., 1993) was first applied: Then coders blind to hypotheses rated (1 = least severe, 5 = most severe) the severity of the CSA.

Regression analyses revealed that older Time 1 age and being female; higher Time 1 CSA severity (with age and gender statistically controlled); and higher Time 2 trauma-related symptom scores (e.g., PTSD symptoms and Trauma Symptom Checklist scores, with Time 1 age, gender, ethnicity [dichotomous], and CSA severity statistically controlled) were associated with higher proportion correct scores. A higher Time 1 CSA severity score and lower trauma-related psychopathology score predicted a higher proportion of omissions. (All βs > |.220|, ps < .021; Model Fs > 3.804, ps < .013.) A floor effect emerged for proportion of commission errors (M = .04, SD = .09).

Overall, childhood age played a central role in adults’ abilities to remember the negative childhood experiences. CSA severity was associated with increased correct responses but also greater omission of information. Whereas general expectations could drive memory after a long delay, Time 2 trauma-related psychopathology symptoms were associated with adults’ memory accuracy. Analyses continue on interview type: The CI predicted greater memory accuracy for a part of the forensic unit experience. Theoretical implications for memory and its development regarding negative, traumatic events, as well as clinical and legal implications, will be discussed.

Authors

• Gail S. Goodman, University of California, Davis
  Presenting Author

• Deborah Goldfarb, University of California, Davis

• Yan Wang, University of California - Davis

• Yuerui Wu, University of California - Davis

• Miranda Kramer, University of California, Davis

• Daisy E. Vidales, University of California - Davis

• Ronald Fisher

• Mitchell L. Eisen, California State University