Selection of lineup foils by similarity to the suspect is likely to misfire

When suspects are arrested or screened by their fit with the initial description provided by the eyewitness but foils are not selected by that criterion, lineups are naturally biased against the suspect. Using probability theory, it is shown that constructing lineups by selecting foils that are similar to the suspect does not eliminate that bias. Any reduction in the amount of bias, if it occurs, is likely to be small. It is argued that for this reason and others discussed by Luus and Wells (1991), foils should rather be selected by their fit with the initial description. The implication is that there is little basis for the hope that if a lineup is constructed properly, a lineup identification is likely to indicate a high match in appearance between the suspect and the offender, regardless of the response criterion. Rather, present lineup procedures can indicate only whether or not the resemblance of the suspect to the offender exceeds, even slightly, the suspect's fit with the initial description.