Cornell University
Memory for Faces, Places, and Pictures: An Examination of Issues in Memory and Law
Abstract
dc:description.abstractMemory is integral to legal institutions. In numerous cases, eyewitness testimony plays an essential evidentiary role. Therefore, an understanding of memory in the legal context is important. This dissertation explores three different applications of memory research to forensic settings: the presence of deep distortions in memory for faces and locations, the effect of framing on memory for pictures and video clips, and the effect of instructions and lineup presentation types on eyewitness identification performance. In Paper 1, we examined whether deep distortions could be detected in memory for forensically relevant stimuli. A deep distortion is a memory distortion in which a person remembers two or more mutually exclusive events, leading to violations of logical axioms. We examined whether deep distortions were present in memory for pictures of famous people and photographs of locations around the participants’ community. In each experiment, we presented participants with a mix of conventional probes, which were basic true/false statements about the identity of the person/location, and disjunctive probes, which were true/false statements that included two plausible identities for the subject of the image. We detected deep distortions in both experiments, suggesting that deep distortions are likely present in forensic settings. In Paper 2, we examined the effect of question framing on memory for photographs and video clips. In typical recognition experiments, participants are asked whether a test item was previously observed (old? probe). Logically, asking instead whether the item is new (new? probe) should produce equivalent results. Research with word stimuli has found that accuracy is highest when the reality state of an item mismatches the framing of the probe. We examined the effects of framing on memory for pictures and video clips. We partially replicated past results in our experiments with picture stimuli, finding that accuracy for new items was significantly higher for old? probes than for new? probes. We, however, did not replicate these results with video clips. These findings have implications for the administration of showups. In Paper 3, we examined the effect of instructions on performance in simultaneous and sequential lineups. Whether simultaneous or sequential lineups are preferable is a source of contention in the eyewitness identification literature. Recently, it has been proposed that the difference between simultaneous and sequential lineups may be the result of sequential lineups that require eyewitnesses to commit to their first yes response. We sought to examine this finding. Additionally, we examined the role of instructions explicitly encouraging conservative or liberal bias in choosing on lineup performance. We found that differences in discriminability and bias were only present when a “first yes” sequential methodology was used. Additionally, we found that instructions encouraging conservative bias induced a conservative criterion shift without sacrificing discriminability. These results have implications for practitioners choosing whether to use simultaneous or sequential lineups and for practitioners wishing to induce conservative bias without sacrificing discriminability.
Degree
thesis:*- Name thesis:degree_name
- Ph. D., Psychological Sciences and Human Development
- Level thesis:degree_level
- Doctor of Philosophy
- Discipline thesis:degree_discipline
- Psychological Sciences and Human Development
- Grantor
- Cornell University
- Year dc:date.issued
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Bialer, Daniel
- Chair dc:contributor.chair
-
- Brainerd, Charles
- Committee members dc:contributor.committeemember
-
- Ceci, Stephen
- Hans, Valerie
Rights
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Dc Identifier Bibid
- 16611681
- Dc Identifier Other
-
ProQuest Submission ID: 14479
ProQuest Publication ID: 31485073