A bias in an identification procedure is any factor— other than recognition—that leads witnesses to select a person. Clothing bias can occur whenever someone is viewed in an identification procedure wearing clothing that matches the witness’s description of the clothing worn during the crime. A witness may mistakenly select the suspect based on the clothing rather than the physical appearance of the person. Although there is limited research to date, clothing bias has been demonstrated to occur with all three commonly used identification techniques: mug-shot searches, lineups, and showups (the presentation of a single suspect to an eyewitness for identification purposes). This research paper will review why clothing bias is a concern for these three procedures and the ways to prevent it.
Findings to date demonstrate that for adult witnesses, clothing bias generally does not affect correct identification rates for mug-shot searches, lineups, or showups. Correct identification rates increase for children in the presence of clothing bias. As with many biases, clothing bias dramatically increases the rate of false positive choices (i.e., identifications of innocent people). This increase in false identifications has been demonstrated with adults for all three identification procedures and with children for lineups.
With mug-shot searches, innocent people may be chosen simply because they happen to be wearing clothing that matches what the perpetrator was wearing. This is problematic because people identified from mug shots are often treated as suspects in the absence of any definitive proof of their innocence (e.g., a strong alibi). Mug-shot searches are hard to protect from clothing bias because the pictures already exist. To control the clothing for future mug shots, the police could take mug shots of people dressed in standard clothing (e.g., large, loose coveralls) or take pictures from the neck up to hide the clothing worn. To control the clothing in extant mug shots, the photographs could be altered (edited) to cover up (mask) clothing or reveal only the head.
Clothing bias is of great concern in a lineup. If the suspect is the only lineup member wearing clothing similar to the perpetrator’s, the suspect will stand out in the lineup—a clear source of lineup bias. Additionally, if the witness selects the suspect, the police and prosecutors may treat the identification and the match between the witness’s description of the clothing and the suspect’s attire as corroborating evidence of the suspect’s guilt. The logic of corroboration is flawed in such cases because the identification and the clothing are not independent sources of evidence if clothing bias exists in the identification procedure.
To protect a lineup from clothing bias, the clothing of all lineup members, including the suspect, should not match the description of the perpetrator’s clothing given by the witness. Ideally, the lineup would consist of only head shots, or all lineup members would be dressed alike. Corroborating evidence can be obtained by creating clothing lineups and asking witnesses to attempt to identify the clothing independently of the person. Sequential lineup presentation has been shown to reduce the size of the clothing bias effect.
Showups generally occur shortly after the crime occurs. Police investigators often will use the description provided by a witness to search the immediate area for potential suspects. Since the descriptions provided by witnesses rarely are detailed enough to ensure that only the perpetrator matches the description and because clothing information often forms a substantial and distinctive portion of the information provided in descriptions, clothing cues are likely to be an important factor in apprehending suspects who appear in showups. As a result, many suspects are likely to have been apprehended near the scene of the crime, shortly after it occurred, and because their clothing was at least a reasonable match to the witness’s description of the perpetrator’s clothing. This can result in witnesses viewing suspects wearing clothes that closely match the description they provided, which can in turn lead to false identifications of innocent suspects. Even when the witness’s description of the clothing is incorrect, innocent suspects wearing clothing that matches the inaccurate description are at heightened risk of both apprehension and false identification. Suspects in showups wearing distinctive clothing (e.g., shirts or jackets with logos) are at greater risk of false identification due to clothing bias than those wearing common clothing (e.g., plain white T-shirts).
To protect a showup from clothing bias, the suspect should not wear clothes that match the description of the perpetrator’s clothing. Sometimes it is not possible to change a suspect’s clothing for showups, (e.g., when they are conducted live at the scene of a crime). In this case, the clothing of suspects could be covered in some way, such as having a blanket covering their body, so as to prevent their clothing from being a cue to the witness. However, this method has not been tested, so its effects on identification decisions are currently not known.
Conviction of innocent people for crimes can be the result of clothing bias during identification procedures. Identifications should be based on recognition of a person, not the clothing they are wearing. Clothing bias is a concern for the three commonly used methods of identification: mug shot, lineup, and showup.
References:
- Dysart, J. E., Lindsay, R. C. L., & Dupuis, P. R. (2006). Show-ups: The critical issue of clothing bias. Applied Cognitive Psychology, 20, 1009-1023.
- Lindsay, R. C. L., Nosworthy, G. J., Martin, R., & Martynuck, C. (1994). Using mug shots to find suspects. Journal of Applied Psychology, 79, 121-130.
- Lindsay, R. C. L., Wallbridge, H., & Drennan, D. (1987). Do the clothes make the man? An exploration of the effect of lineup attire on eyewitness identification accuracy. Canadian Journal of Behavioural Science, 19, 463-178.
Return to the overview of Eyewitness Memory in Forensic Psychology.