Evaluating a new computer-based system for face recall

The Computer-Aided Design Centre has developed a new face recall system which enables a witness to construct the image of a suspect on a television screen. The system currently uses the Photofit library and method of construction. Likenesses of target faces produced by subjects (n=36) using the new system were compared to conventional Photofits produced by a second group of subjects (n=36) of the same targets. There was no difference on any of the three measures employed in the assessed quality of the likenesses produced by Photofit and the CADC System. Both systems tended to produce more recognizable composites when construction was carried out in the presence of the target compared to its absence. Composites made from memory on the initial trial were marginally better with the CADC System than with Photofit. While both systems showed low absolute accuracy, they enjoyed a measure of success in transmitting a likeness of a facial type. It is concluded that while the CADC system shows promise, psychological constraints upon witnesses' powers of recall may place limits on the absolute accuracy achievable with any system of face reconstruction.These experiments were financed by grant No. POL/73 1675/24/3 from the Police Scientific Development Branch of the U.K. Home Office. Points of view expressed in this article are solely those of the authors and do not necessarily reflect the opinions or positions of the funding body. Our gratitude and thanks are due to Tony Kitson and Clive Holloway for their many hours spent developing and operating the CADC system; Jean Shepherd for her skills as a Photofit operator; the Computer-Aided Design Centre; and the MRC Applied Psychology Unit, Cambridge for facilities and subjects.     

The relationship of abuse and witnessing violence on the Child Abuse Potential Inventory with black adolescents

A number of social and psychological factors present in most adolescent parents place them at high risk for abusive behavior toward their children. However, current child abuse potential measures do not include adolescent samples as part of the psychometric data base. Consequently, the purpose of this study was twofold: (1) to investigate whether a Black adolescent sample would perform differently than an adult nonabusing sample on the Child Abuse Potential Inventory (CAP), and (2) to examine the relationship between history of abuse or witnessing violence to scores on the CAP. Results revealed that Black adolescents scored significantly higher than the adult normative sample on the CAP; consequently, cutoff scores need to be empirically established for adolescents. Additional analyses indicated that a history of abuse, as well as a history of witnessing violence, are associated with a high potential for abuse.