Increasing accuracy, reducing choosing behaviour, and mock juror perceptions of children’s eyewitness identification

  • Julie Dunlevy

Student thesis: Doctoral Thesis


This thesis reports the findings of four experiments that aimed to provide a better understanding of why children, more than adults, tend to choose from a line-up and make more mistaken identifications. Further, this research also aimed to examine the perceptions of mock jurors pertaining to both the age of a witness, either a child or an adult and witness choice for additional viewings of the line-up or line-up members. A total of 288 children and 351 adults participated in one of four experiments. In experiment one, a nonverbal rejection option (the ‘wildcard’ which allowed children to select from a line-up without selecting a face) was found to significantly decrease incorrect identifications of a foil in target absent conditions whilst maintaining a good rate of correct identifications in target present conditions. The encouraging initial findings did not generalise to the second experiment. However, in the second experiment, training with simple practice line-ups, prior to viewing a line-up containing faces, was found to significantly reduce children’s choosing from the line-up, although this was at a cost to correct identifications. In experiment three neither the wildcard procedure nor training with practice line-ups benefitted the young adults. The fourth experiment investigated mock jurors’ perceptions of a police witness to a crime who later identified the police suspect from a line-up. Both the age of the witness (7-years or 29-years) and witness choice for additional viewings were manipulated. No age-related differences were found in credibility ratings for identifying the police suspect but children, compared to adults were perceived as significantly more credible for their account of the crime event. Additionally, both the child and adult witness were perceived as more credible for their account of the event than for their credibility in the identification of the police suspect. Implications and directions for future research are discussed.
Date of Award2019
Original languageEnglish
Awarding Institution
  • University of Portsmouth
SupervisorJulie Cherryman (Supervisor) & Aldert Vrij (Supervisor)

Cite this