Charles Explorer logo
🇬🇧

Verbal Description and Accuracy in a Cross-Race Effect Study

Publication at Faculty of Mathematics and Physics, Faculty of Arts |
2019

Abstract

Eyewitnesses' verbal descriptions of the perpetrator are used in identification tasks. Laboratory studies in this field are often focused on a) the verbal over-shadowing effect, b) the amount of information provided vs. the accuracy c) the way lineups should be assembled (based on the perpetrator's description or based on the similarity of the suspect) and in a limited amount d) the differences in verbal descriptions of other ethnicities.

In our study, we focused on these differences when describing a Caucasian ethnicity perpetrator and an Asian ethnicity perpetrator. The number of details provided was one of the variables which was also further examined in relation to participants' cognitive and verbal abilities.

Forty Caucasian participants of different ages (24 males, mean age: 30, sd = 6) and educational levels (university degree = 23, secondary education = 14, completed GED = 2, without GED = 1) took part in this study. The stimuli used for this research included two 20-second videos (one with an Asian and the second with a Caucasian perpetrator) of a mock robbery and 8-person photolineups for each condition (target present x target absent, Caucasian x Asian race).

In the first part of the research, participants were shown the video with Caucasian offender and provided his description. As a distraction task between the stimuli and identification task, they filled a 30-minute cognitive test which evaluated their basic cognitive abilities (attention, language, visuospatial/constructional abilities, immediate and delayed memory), and later were presented with the lineup (target present or absent).

This procedure was repeated an hour later with a perpetrator of Asian ethnicity. Only a small correlation was found between the total amount of information given and their identification accuracy (r = 0,231).

The average number of information given was 8.05 (Avg Caucasian = 8.7; Max = 15; Min = 4; Total = 349; Avg Asian = 7.4; Max = 11; Min = 3; Total = 295). On average, accurate witnesses provided more information (MCaucasian = 9.4; MAsian = 7.8) than inaccurate witnesses (MCaucasian = 8.3; MAsian = 7.2).

Differences were also found in the type of provided details. When describing a suspect of their own ethnicity, participants provided more detail about facial features (62.3 %), the weapon used (60.6 %), and other equipment used during the crime (58.8 %).

Racial characteristics were described least often (42.9 %). When describing a suspect of another ethnicity, respondents focused more on the clothing of the perpetrator rather than his face.

No significant results were found between the description length and cognitive abilities of our participants. The lack of our capability to recognize other ethnicities understandably has an influence on our ability to describe the features of its members.

The description length cannot be solely used to determine a witness' accuracy, but it may be a contributing factor. This work was supported by the Czech grant GAUK-232217.