Brian R. Clifford

Voice identification by human listeners: On earwitness reliability

This paper reviews what is currently known about voice identification by human listeners. Our own experimental data from a four-year research program into this topic is used to elucidate, support, and in some cases to contradict published work into the effects on voice identification of such factors as speech sample size and quality, voice disguise, delay in holding voice identification sessions, incidental as opposed to intentional memory for voices, the effects of the age of the witness, training in specific modes of encoding voices, and the relationship between objective accuracy and subjective feelings of certainty of correctness. It is concluded that the caution and suspicion currently accorded to visual identification must be extended also, and perhaps more so, to voice identification.

The Effects of the Cognitive Interview on Recall, Identification, Confidence and the Confidence/Accuracy Relationship

The effects of two interview techniques, the Cognitive Interview (CI) and the Structured Interview (SI), were examined in relation to recall, recognition, confidence and the confidence/accuracy relationship. Volunteer subjects witnessed a live staged event, and at set time delays (48 and 96 hours) were interviewed by means of either the CI or SI and made lineup identifications from both blank and filled lineup presentations. The CI significantly improved total recall (p<0.001) and correct recall (p<0.001) relative to the SI. There were no differences in either accuracy or error rates for the two conditions. The CI produced significantly higher ‘pre-post interview’ confidence ratings than the SI (p<0.05). Within-, between- and event-confidence ratings were not predictive of recall or recognition accuracy. Failure to find a significant consistent confidence/accuracy relationship was suggested to have occurred because of the operation of an accuracy assessment heuristic. The CI did not reliably improve identification from a filled lineup presentation relative to the SI, but did produce better performance on a blank lineup presentation (p<0.001). Confidence and the confidence/accuracy relationship involving identifications were not found to differ as a function of interview condition. Throughout the study multiple regression analyses failed to reveal consistent predictors of recall or recognition accuracy, confidence or the confidence/accuracy relationship of these two domains.

Children's understanding of the earth in a multicultural community: Mental models or fragments of knowledge?

Children’s understanding of properties of the earth was investigated by interviewing Asian and white British classmates aged 4 - 8 years (N = 167). Two issues were explored: whether they held mental models of the earth (Vosniadou & Brewer, 1992) or instead had fragmented knowledge (di Sessa, 1988); and the influence of the children’s different cultural backgrounds. Children selected from a set of plastic models and answered forced choice questions. Using this methodology, there were no significant differences in the overall performance of Asian and white children after language skills were partialled out. Even young children showed an emerging knowledge of some properties of the earth, but the distributions of their combinations of responses provided no evidence that they had mental models. Instead, these distributions closely resembled those that would be expected if children’s knowledge in this domain were fragmented. Possible reasons for the differences between these findings and those of previous research are discussed.

Inoculation or Antidote? The Effects of Cognitive Interview Timing on False Memory for Forcibly Fabricated Events

This study examined whether a cognitive interview (CI) can counteract the effects of suggestive interviews involving forced fabrication. College students witnessed a filmed event and were later forced to fabricate answers to misleading questions about the event. All witnesses were interviewed with a non-leading CI or free recall (FR) either before or after the forced fabrication phase. A week later participants completed a recognition and source monitoring (SM) test of video content. Relative to FR, the CI administered before the forced fabrication interview increased reports of correct details and reduced false assents to fabricated items. A CI after resulted in false memory rates comparable to the FR group. Early interviews using CI techniques may protect against memory loss and misinformation effects.

The Voice-Recognition Accuracy of Blind Listeners

A research programme has been carried out that concerns the accuracy with which listeners can identify a speaker heard once before. The present study examined the voice-recognition abilities of blind listeners, and it was found that they could more accurately select target voices from the test arrays than could sighted people. However, the degree of blindness, the age at onset of blindness, and the number of years of blindness all failed to relate to voice-recognition accuracy.

The effects of delay on voice recognition accuracy

Two experiments were conducted in which 176 listeners heard male and female objectively defined “high-” and “low-recognition” voices and then attempted to identify these voices from a “voice parade” containing 20 distractors after either 10, 40, 100, or 130 minutes (experiment 1), or 10 minutes, one day, seven days, or 14 days (experiment 2). In experiment 1 delay had no overall effect, although further analysis revealed that the shortest delay did produce better performance than all other delay conditions. Further, “high-recognition” voices were better identified than “low-recognition” voices. In experiment 2 delay had an overall effect, with the shortest delay interval again being significantly better than all other conditions, which did not differ among themselves. “High-” and “low-recognition” voices, however, did not exhibit a statistically significant difference, although these two factors entered into a marginally significant interaction. Theoretical speculation and forensic implications were drawn.

The emotional eyewitness: the effects of emotion on specific aspects of eyewitness recall and recognition performance

The present set of experiments aimed to investigate the effects of negative emotion on specific aspects of eyewitness recall and recognition performance. The experience of emotion was manipulated between subjects, with participants either viewing a crime scenario (a mugging) or a neutral scenario (a conversation). Eyewitness recall was categorized into descriptions of the perpetrator, critical incident, victim, and environmental details. The completeness and accuracy of eyewitness recall across categories of detail were measured in Experiment 1. A significant main effect of negative emotion was found for the completeness of recall. Furthermore, a significant main effect of the completeness of eyewitness statements was found, but not for their accuracy. However, these main effects were qualified by a significant interaction between emotion and category of detail recalled. Specifically, emotional participants provided a more complete description of the perpetrator than neutral participants; however, they were less able than their neutral counterparts to describe what the perpetrator did to the victim. In light of these findings, Experiment 2 investigated whether enhanced completeness of perpetrator descriptions during recall translated into an enhanced ability to recognize the perpetrator from a photographic lineup by emotional compared with neutral participants. Results from Experiment 2 suggest that while emotional participants again provide a more complete description of the perpetrator, they are less able than their neutral counterparts to recognize the perpetrator from a photographic lineup. Results are discussed in terms of a retrieval motivation hypothesis of negative emotional experience and the possible consequences for eyewitness testimony.

Suggestibility and state anxiety: How the two concepts relate in a source identification paradigm

Source identification tests provide a stringent method for testing the suggestibility of memory because they reduce response bias and experimental demand characteristics. Using the techniques and materials of Maria Zaragoza and her colleagues, we investigated how state anxiety affects the ability of undergraduates to identify correctly the source of misleading post-event information. The results showed that individuals high in state anxiety were less likely to make source misattributions of misleading information, indicating lower levels of suggestibility. This effect was strengthened when forgotten or non-recognised misleading items (for which a source identification task is not possible) were excluded from the analysis. Confidence in the correct attribution of misleading post-event information to its source was significantly less than confidence in source misattributions. Participants who were high in state anxiety tended to be less confident than those lower in state anxiety when they correctly identified the source of both misleading post-event information and non-misled items. The implications of these findings are discussed, drawing on the literature on anxiety and cognition as well as suggestibility.

The effects of the cognitive interview and other methods of context reinstatement on identification

Abstract Can reinstatement of encoding context aid eyewitness identification? Two experiments are reported in which participants were asked to identify, from both a Blank and a Filled lineup, a target seen 1-week (Experiment 1) or 3-months (Experiment 2) earlier in a staged live interaction. Identifications were made following either a no context reinstatement (NCR), a CI-type reinstate context (CI-CR), a mental and physical (M&PCR) context reinstatement or a multiple reinstatement of context (Multi-CR) manipulation. In Experiment 1 in the Blank lineup condition, correct rejection (CR) and false identification (FID) rates did not differ between the four context manipulation conditions. However, within the different conditions only Multi-CR showed a significant difference between CR and FID. In the Filled lineup condition, neither correct identification (CID), FID, nor non-identification (NID) rates differed between context conditions. Within the four context conditions only Multi-CR produced significantly...

Memory for televised information: a problem for applied and theoretical psychology

Several approaches to variations in memory for informational television are reviewed. Although broad sociological approaches have not yet proved very fruitful, recent cognitively-oriented research has indicated important effects on learning and remembering of detailed aspects of programmes, such as visual format, sequencing of material, vision-text relations and varied recapitulation of information. The evidence for the importance of these factors has derived mainly from studies of TV news material, but Educational Television research is also beginning to show promising results as more attention is given to the detailed ways in which information is organized within programmes. Individual differences are also important, however, and need to be more effectively studied. By and large, there is little support for the more pessimistic views about the value of television as a source of information and instruction, and research findings point to many aspects of practice that might be improved to make programmes more effective. Some of the problems of applying research findings and of the relation of psychological theory to applied experimentation are discussed.