2025/11/05 by D. M. Bialer, C. J. Brainerd · 1 voice
Neuroscience · Psychology · #Face Recognition and Perception #Memory Processes and Influences #Deception detection and forensic psychology
paper · pdf · doi:10.3758/s13421-025-01805-x
openalex created_date 2025/11/05 · openalex publication_date 2025/11/05 · openalex updated_date 2026/07/27
Deep distortions occur when a person remembers multiple incompatible reality states as true, producing logical impossibilities. Those impossibilities include violations of basic logical principles such as additivity, countable additivity, and universal event. Previously, deep distortions have been detected in experiments using word lists and using real-world facts. We report two experiments in which we extend these findings to realistic pictures that figure prominently in legal cases. In Experiment 1, we presented participants with images of celebrity faces and asked them true or false questions about their identities, which were either conventional probes (e.g., "This is Alexander Hamilton") or disjunctive probes (e.g., "This is Alexander Hamilton or James Madison"). In Experiment 2, we presented participants with images of locations around their local community, which were also paired with true or false probes about the identity of the locations that were either conventional probes (e.g., "This is Mann Library") or disjunctive probes (e.g., "This is Uris Library or Mann Library"). In both experiments we found violations of additivity, countable additivity, and universal event. We also found that the incompatible identities of the faces and locations were not compensatory (i.e., as the acceptance rates for one incompatible identity increased, the acceptance rates for the other incompatible identities did not decrease). Additionally, we conducted conjoint recognition analyses and found that deep distortions were rooted in gist memory rather than verbatim memory. Overall, these finding suggest that deep distortions are also stimulated by realistic images that are highly relevant to legal cases.