Differential mnemonic consistency differs between experienced and fabricated incidents

ABSTRACT

The degree of consistency of reports across several interviews is often taken as indicating whether a report is experience-based or fabricated. However, due to forgetting some aspects of mnemonic reports will be inconsistent across interviews independent of whether a report is experience-based or fabricated. The concept of differential mnemonic consistency implies that certain elements of experienced incidents such as core activities are remembered relatively well whereas other elements of experienced incidents such as peripheral activities are more easily forgotten. This difference in mnemonic consistency has been claimed to be higher for experienced-based than for fabricated reports. We tested this prediction. Participants experienced or fabricated an ‘examination’ in which their blood pressure, body height, and body weight were measured. N = 326 participants reported as many details as possible immediately and four weeks after the incident. Mnemonic consistency between reports was indeed higher for elements with high than for elements with low expected mnemonic consistency. Most importantly, this difference in mnemonic consistency was larger in experience-based than in fabricated reports. The results suggest that differential mnemonic consistency could serve as a supplemental criterion for distinguishing between experience-based and fabricated reports but further research is necessary to determine its usefulness in forensic practice.

KEYWORDS:

• Statement credibility
• deception detection
• eyewitness testimony
• credibility assessment
• criteria-based content analysis

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The data underlying the analyses reported here are available at https://osf.io/pef6a/

2 As mentioned above, elements were scored as ‘recalled correctly’, as ‘recalled incorrectly’, or as ‘omitted’ such that the probabilities of recalling an element correctly, of recalling an element incorrectly, and of omitting an element must necessarily add to 1. There were extremely few elements that were recalled incorrectly (only 3.7 % on average) so that this category of responses could not be meaningfully analyzed. As a consequence of the negligibly few incorrectly recalled elements, the data pattern for the omitted elements is basically a mirror image of the data pattern for the correctly recalled elements displayed in Figure 2. Therefore, the analysis of the omitted elements necessarily yields results that are completely redundant given the results of the analysis of the correctly recalled elements.