The purpose of an eyewitness identification procedure is to determine whether the police suspect is guilty or innocent of the crime in question. Rather than presenting the lone suspect to the witness for an identification attempt, it is generally advised that the police investigator first surround the suspect with known-innocent fillers. The rationale is that if the suspect is guilty, the eyewitness ought to be able to pick this person out of a lineup. If the suspect is innocent, these fillers offer that individual protection from mistaken identification as many witnesses who would have otherwise identified the innocent suspect identify a filler instead (e.g., Smith et al., 2018; Smith et al., 2020). Hence, the idea of surrounding the suspect with known-innocent fillers is that we can strike a better trade-off between correct perpetrator identifications and incorrect innocent-suspect identifications. But producing that improved trade-off requires that fillers be selected so that they are, on average, as similar to the witness’ memory for the perpetrator as is the innocent suspect. Lineups in which fillers prevent the innocent suspect from standing out as a better match to the witness’ memory for the perpetrator are called fair and lineups in which the fillers do not prevent the innocent suspect from standing out as a better match to the witness’ memory for the perpetrator are called biased. More than 40 years of research demonstrates that fair lineups lead to a better trade-off between guilty-suspect identifications and innocent-suspect identifications than do biased lineups and that a suspect identification from a fair lineup is far more reliable than a suspect identification from a biased lineup (e.g., Clark, 2012; Lindsay & Wells, 1980; Smith et al., 2017). In the present line of research, we examine how laypeople perceive the reliability of identifications made from fair versus biased lineups. In Experiment 1, we presented laypeople with (1) a brief description of the crime, (2) the witness’ brief description of the perpetrator, (3) a six-person photographic lineup, and (4) the position of the identified lineup member. Each lay participant viewed one fair lineup, where all lineup members match the witness’ description of the perpetrator, and one biased lineup, where only the lineup member who was identified matches the witness’ description of the perpetrator. On each lineup, participants were asked to the likelihood the witness had made an accurate identification. We posited that, counterintuitively, laypeople would judge identifications from biased lineups to be more reliable than identifications from fair lineups. The rationale was that the non-selected lineup members provide a reference group of alternative outcomes that can be contrasted with the identified lineup member. Compared to when only the identified lineup member fits the witness’ description (biased lineup; high perceptual contrast), when all lineup members fit the witness’ description (fair lineup; low perceptual contrast), laypersons might look to the non-identified alternatives, question why not one of these other persons, and lower their reliability judgment accordingly. As we expected, laypeople rated likely accuracy higher for biased lineup than they did for fair lineups (d = .80). In the follow-up experiments, our goal was to determine what information laypeople needed to know before aligning their perceptions of fair and biased lineups to be equivalent to what the research shows to be the truth. To do this, we incrementally added additional instructions to the lineup task in a series of three separate experiments. In Experiment 2, we added to the instructions that the witness identified the suspect. We expected that this inclusion would key participants into the fact that there is only a single suspect present in the lineup (and that the other five are known-innocent, or not suspected of the current crime). In Experiment 3, we instructed participants to keep in mind that the actual culprit might or might not be present in the lineup, which we expected to help guide them to the conclusion that the “correct” answer might be to reject the lineup (or say that the culprit is not present). In Experiment 4, we informed participants that the police designed the lineup by selecting known-innocent fillers from a database of images. We expected that this instruction would help participants to appreciate the intent with which many biased lineups are constructed—to increase the likelihood that the witness would identify the suspect. However, adding on these additional pieces of information had little impact on participants ratings. Throughout the next several experiments, we were interested in determining how lay participants responded when presented with a lineup fairness question, rather than a question about suspect identification accuracy. In Experiment 5 and Experiment 7, we found that participants rated fair lineups as “fairer” than biased lineups, indicating that they were sensitive to the lineup fairness manupulation. Furthermore, in Experiment 5, we found that priming participants with this fairness question attenuated participants’ preferences for the biased lineup over the fair lineup. In fact, in Experiment 8, we found that providing participants with the full set of instructions (tested in Experiments 2 through 4) and priming them with a lineup fairness question (tested in Experiment 5 and 7), led to laypersons rating fair and biased lineups as almost identical in terms of suspect identification accuracy. Save for priming participants with lineup fairness considerations, none of the subtle instructions we provided to participants had much impact on their accuracy ratings for fair and biased lineups. Hence, in Experiment 9, we take a different approach. We will present laypersons with a short passage describing the purpose of lineups to see whether giving them more of a conceptual explanation leads them to appreciate why suspect identification accuracy is higher for fair than for biased lineups. This passage makes four key points: (1) suspects are generated by using the witness’ description, (2) suspects are sometimes innocent, (3) eyewitnesses frequently make false identifications of innocent persons, (4) the purpose of a lineup is to see if the witness recognizes the suspect as the culprit, and (5) lineups can be a good measure of guilt when all fillers match the witness’ description. The passage reads as follows: When a crime occurs there is sometimes an eyewitness who got a view of the culprit and can give a verbal description of that culprit. Police will generate suspects matching the witness’ description. The suspect might be innocent, of course, and so police sometimes put together a lineup (either live or photographic) that includes that suspect and then show that lineup to the eyewitness. It is not uncommon for the suspect in a lineup to in fact be innocent (not the culprit). Because eyewitnesses often mistakenly identify innocent people, police embed the suspect among known-innocent lineup members. These known-innocent lineup members are called “lineup fillers.” In a typical lineup, there are five of these fillers plus the one suspect, thereby yielding a lineup composed of six people. The goal of the lineup is to make sure that an eyewitness actually recognizes the suspect as being the culprit. But if the fillers do not fit the witness’s description of the culprit, then maybe the witness will pick the suspect only because he best fits the description rather than because the witness actually recognizes the suspect as being the culprit. Therefore, the fillers should fit the verbal description that the eyewitness gave of the culprit. If the fillers all fit the description that the eyewitness gave of the culprit and the eyewitness still manages to pick the suspect, the identification is considered good evidence of guilt. We will also use five manipulation check questions to assess participant comprehension of the short passage. Participants will answer these manipulation check questions at the end of the study, after having rated the likely suspect identification accuracy of both a fair and a biased lineup. If participants do not rate suspect identification accuracy as higher for fair than for biased lineups, these manipulation check questions will help us to determine the extent to which that might have been attributable to comprehension issues rather than to the inefficacy of the passage. The questions are listed below, with the correct answers marked by an asterisk. Which of the following is typically true? • *Police develop suspects using the witness' description of the culprit. • Police do not use witness' description of the culprit when developing suspects. • Police develop suspects who look different from the witness' description of the culprit. • None of the above. Suspects in the lineup are _____ innocent. • Always • Never • *Sometimes • Normally Witnesses... • Are always correct. • Are always false. • Rarely make false identifications • *Frequently make false identifications The purpose of a lineup is: • To see if one of the fillers might be the actual culprit in the case • *To see if the eyewitness actually recognizes the suspect as being the culprit rather than simply being the lineup member who best fits the description • To make the task as easy and obvious as possible so that the witness will pick the person the police want them to pick • None of the above Identifications from a lineup are good evidence of guilt if: • *Fillers match the witness' description of the culprit • Fillers do not match the witness' description of the culprit • Regardless of whether or not fillers match the witness' description of the culprit • The suspect looks very different than the other lineup members