As these proceedings illustrate, interest in the theoretical mechanisms underlying face recogniti... more As these proceedings illustrate, interest in the theoretical mechanisms underlying face recognition has increased dramatically in recent years. The study of face recall, by contrast, has been somewhat neglected. Yet any comprehensive theory of face memory must take account of recall and recognition. Most work on recall has been couched in terms of the practical problems surrounding the identification of criminals.
Marked discrepancies exist between levels of face recognition and accuracy of face recall registe... more Marked discrepancies exist between levels of face recognition and accuracy of face recall registered by Photofit. One of the reasons for this may be that Photofit demands that subjects initially identify features in isolation and such a strategy is incompatible with normal coding processes. Experiment 1 compared the normal feature-based approach to face building with a condition which permitted subjects to work from an established face (‘Mr Average’). No superiority for the latter condition could be established. Experiment 2 contrasted the standard building method with one in which subjects were offered a variety of facial types as potential starting points. Again, accuracy was no higher under the novel procedure. It was concluded that there were no grounds advocating for a wholesale change in current tactics of Photofit building which emphasize feature-selection prior to amending a total facial composite.
Twenty white and 20 black encoders each made Photofit constructions from memory of one of a sampl... more Twenty white and 20 black encoders each made Photofit constructions from memory of one of a sample of 10 white faces and one of a sample of 10 black faces. Then, to provide accuracy indices, two different groups of 20 white decoders matched the resulting Photofits to photographs of the original 20 faces. Both white and black encoders made more accurate white than black Photofits; white encoders made more accurate white Photofits than did black encoders; however, there were no racial differences in accuracy among encoders for black faces. Results were explained by cross-racial memory and by Photofit Kit differences.
Abstract Children aged approximately 8.5–9.5 years were shown either names (NP), or pictures (PP)... more Abstract Children aged approximately 8.5–9.5 years were shown either names (NP), or pictures (PP), of objects, for each of which they located the corresponding likenesses among an array of alternatives. Written recall and identification accuracy were assessed at 5 hr and 3 days after presentation. As predicted, NP produced significantly superior recall to PP. The rate of identification for recalled items among S s in PP was significantly higher than that for NP. This result was interpreted as supporting the view that the superiority of NP in recall derives from the fact that S s may rely upon retrieval from either name or image of each item whereas those in PP must depend upon image alone. This interpretation was extended to take two unexpected findings into account: (i) the superior overall identification accuracy engendered by PP, particularly at the longer time interval; (ii) the relatively weak relationship between identification accuracy and recall scores among matching S s.
Page 1. CHU.DRI;N & SOCIETY (1991) 53, 254-265 Children's credibility on camera: the inf... more Page 1. CHU.DRI;N & SOCIETY (1991) 53, 254-265 Children's credibility on camera: the influence of age and production factors HELEN L. WESTCOTT, BRIAN R. CLIFFORD AND GRAHAM M. DAVIES SUMMARY. An experiment ...
Page 1. CHILDREN & SOCIBTY (1991) 5:2, 123-135 Adults' perceptions of children's vi... more Page 1. CHILDREN & SOCIBTY (1991) 5:2, 123-135 Adults' perceptions of children's videotaped truthful and deceptive statements HELEN L. WESTCOTT, GRAHAM M. DA VIES and BRIAN R. CLIFFORD SUMMARY. This article ...
British journal of psychology (London, England : 1953), Jan 27, 2017
The review of 'Recognizing faces' by Hadyn Ellis, published in the British Journal of Psy... more The review of 'Recognizing faces' by Hadyn Ellis, published in the British Journal of Psychology in 1975, marked the genesis of a distinct field of research. This seminal review sprang from a broader programme of research on face recognition conducted at the University of Aberdeen, whose influence continues to be felt in what has become an internationally important research area. We discuss the background to the Aberdeen research, summarize some of its achievements, and offer reasons why it proved so successful. These reasons include the synergy between theory and practice-based studies, the value of combining different perspectives and sources of evidence, sound techniques that led to easily reproducible findings, and the emphasis on testing even the most 'common sense' assumptions.
Http Dx Doi Org 10 1080 14640747608400591, May 29, 2007
Schoolchildren, aged 16 years showed poorer incidental learning for nouns if the orienting task w... more Schoolchildren, aged 16 years showed poorer incidental learning for nouns if the orienting task was one of finding rhymes for the items compared to assigning each word to a semantic category. On a recognition task including homophone and synonym foils, rhyming subjects made significantly more phonemic errors, whereas categorizing subjects produced a significantly greater number of semantic confusions. Subjects instructed to construct images for each of the nouns showed a performance pattern similar to the categorizing group. The results are interpreted in terms of Herriot's (1974) view that different levels of processing are associated with the coding of different clusters of stimulus attributes.
Http Dx Doi Org 10 1080 13218710802101605, Nov 1, 2008
... We are satisfied that in cases which depend wholly or mainly on eye-witness evidence of ... w... more ... We are satisfied that in cases which depend wholly or mainly on eye-witness evidence of ... with its emphasis upon the manipulation of small numbers of variables, real-world eyewitness performance is ... The procedure used in video parades involves the accused first making a 15 s ...
... other criteria, such as Pardoning the Accused are rarely reported (Lamb et al., 1997). Some c... more ... other criteria, such as Pardoning the Accused are rarely reported (Lamb et al., 1997). Some criteria appear to be difficult to apply with any consistency (Horowitz et al., 1997). ... Alonzo-Quecuty was among the first researchers to apply reality moni-toring to eyewitness testimony. ...
As these proceedings illustrate, interest in the theoretical mechanisms underlying face recogniti... more As these proceedings illustrate, interest in the theoretical mechanisms underlying face recognition has increased dramatically in recent years. The study of face recall, by contrast, has been somewhat neglected. Yet any comprehensive theory of face memory must take account of recall and recognition. Most work on recall has been couched in terms of the practical problems surrounding the identification of criminals.
Marked discrepancies exist between levels of face recognition and accuracy of face recall registe... more Marked discrepancies exist between levels of face recognition and accuracy of face recall registered by Photofit. One of the reasons for this may be that Photofit demands that subjects initially identify features in isolation and such a strategy is incompatible with normal coding processes. Experiment 1 compared the normal feature-based approach to face building with a condition which permitted subjects to work from an established face (‘Mr Average’). No superiority for the latter condition could be established. Experiment 2 contrasted the standard building method with one in which subjects were offered a variety of facial types as potential starting points. Again, accuracy was no higher under the novel procedure. It was concluded that there were no grounds advocating for a wholesale change in current tactics of Photofit building which emphasize feature-selection prior to amending a total facial composite.
Twenty white and 20 black encoders each made Photofit constructions from memory of one of a sampl... more Twenty white and 20 black encoders each made Photofit constructions from memory of one of a sample of 10 white faces and one of a sample of 10 black faces. Then, to provide accuracy indices, two different groups of 20 white decoders matched the resulting Photofits to photographs of the original 20 faces. Both white and black encoders made more accurate white than black Photofits; white encoders made more accurate white Photofits than did black encoders; however, there were no racial differences in accuracy among encoders for black faces. Results were explained by cross-racial memory and by Photofit Kit differences.
Abstract Children aged approximately 8.5–9.5 years were shown either names (NP), or pictures (PP)... more Abstract Children aged approximately 8.5–9.5 years were shown either names (NP), or pictures (PP), of objects, for each of which they located the corresponding likenesses among an array of alternatives. Written recall and identification accuracy were assessed at 5 hr and 3 days after presentation. As predicted, NP produced significantly superior recall to PP. The rate of identification for recalled items among S s in PP was significantly higher than that for NP. This result was interpreted as supporting the view that the superiority of NP in recall derives from the fact that S s may rely upon retrieval from either name or image of each item whereas those in PP must depend upon image alone. This interpretation was extended to take two unexpected findings into account: (i) the superior overall identification accuracy engendered by PP, particularly at the longer time interval; (ii) the relatively weak relationship between identification accuracy and recall scores among matching S s.
Page 1. CHU.DRI;N & SOCIETY (1991) 53, 254-265 Children's credibility on camera: the inf... more Page 1. CHU.DRI;N & SOCIETY (1991) 53, 254-265 Children's credibility on camera: the influence of age and production factors HELEN L. WESTCOTT, BRIAN R. CLIFFORD AND GRAHAM M. DAVIES SUMMARY. An experiment ...
Page 1. CHILDREN & SOCIBTY (1991) 5:2, 123-135 Adults' perceptions of children's vi... more Page 1. CHILDREN & SOCIBTY (1991) 5:2, 123-135 Adults' perceptions of children's videotaped truthful and deceptive statements HELEN L. WESTCOTT, GRAHAM M. DA VIES and BRIAN R. CLIFFORD SUMMARY. This article ...
British journal of psychology (London, England : 1953), Jan 27, 2017
The review of 'Recognizing faces' by Hadyn Ellis, published in the British Journal of Psy... more The review of 'Recognizing faces' by Hadyn Ellis, published in the British Journal of Psychology in 1975, marked the genesis of a distinct field of research. This seminal review sprang from a broader programme of research on face recognition conducted at the University of Aberdeen, whose influence continues to be felt in what has become an internationally important research area. We discuss the background to the Aberdeen research, summarize some of its achievements, and offer reasons why it proved so successful. These reasons include the synergy between theory and practice-based studies, the value of combining different perspectives and sources of evidence, sound techniques that led to easily reproducible findings, and the emphasis on testing even the most 'common sense' assumptions.
Http Dx Doi Org 10 1080 14640747608400591, May 29, 2007
Schoolchildren, aged 16 years showed poorer incidental learning for nouns if the orienting task w... more Schoolchildren, aged 16 years showed poorer incidental learning for nouns if the orienting task was one of finding rhymes for the items compared to assigning each word to a semantic category. On a recognition task including homophone and synonym foils, rhyming subjects made significantly more phonemic errors, whereas categorizing subjects produced a significantly greater number of semantic confusions. Subjects instructed to construct images for each of the nouns showed a performance pattern similar to the categorizing group. The results are interpreted in terms of Herriot's (1974) view that different levels of processing are associated with the coding of different clusters of stimulus attributes.
Http Dx Doi Org 10 1080 13218710802101605, Nov 1, 2008
... We are satisfied that in cases which depend wholly or mainly on eye-witness evidence of ... w... more ... We are satisfied that in cases which depend wholly or mainly on eye-witness evidence of ... with its emphasis upon the manipulation of small numbers of variables, real-world eyewitness performance is ... The procedure used in video parades involves the accused first making a 15 s ...
... other criteria, such as Pardoning the Accused are rarely reported (Lamb et al., 1997). Some c... more ... other criteria, such as Pardoning the Accused are rarely reported (Lamb et al., 1997). Some criteria appear to be difficult to apply with any consistency (Horowitz et al., 1997). ... Alonzo-Quecuty was among the first researchers to apply reality moni-toring to eyewitness testimony. ...
Uploads
Papers by Graham Davies