This post was originally published on July 25, 2019 on Psychology in Action.
The advent of forensic techniques able to analyze DNA evidence resulted in the exoneration of many who had been convicted of a crime they did not commit (Innocence Project, 2018). How did so many innocent people get convicted in the first place? Analysis of cases in which DNA evidence overturned a conviction found that the majority of these convictions were due to incorrect witness identifications (Innocence Project, 2019). This finding prompted research into the validity of eyewitness testimony and the conditions that optimize the quality of witness identifications.
One factor that has been subject to much debate in recent years is whether a simultaneous or a sequential lineup leads to better eyewitness performance. Lineup format is of interest to researchers and policymakers because it is a system variable, meaning that it is something the justice system can control, unlike many other factors (such as whether the witness had poor visibility of the suspect). In the U.S., lineups typically consist of six pictures presented to the eyewitness on a computer screen, one of the suspect and five of known-innocent fillers who look like the suspect (Seale-Carlisle, Wetmore, Flowe, & Mickes, 2019). In a simultaneous lineup, the suspect’s face is presented in a grid along with the fillers. The witness views all of these faces together before deciding to make an identification. In contrast, a sequential lineup presents each face to the witness one at a time. The witness decides whether or not to make an identification as each face is presented and only moves on to the next photo after deciding not to identify the current photo as the guilty person.
Early research on lineup formats indicated that sequential line ups provide better witness discrimination than simultaneous lineups. Better discrimination means that witnesses were less likely to identify a known-innocent filler as the suspect if a sequential lineup was used compared to a simultaneous lineup (Lindsay & Wells, 1985; Steblay, Dysart, Fulero, & Lindsay, 2001). According to the absolute-relative judgment theory (Wells, 1984), sequential lineups result in better witness discrimination because the format encourages people to make a strict judgment against their own memory. The witness must decide if a photo is of the suspect before moving to the next person without knowing what the next face will look like. Alternatively, simultaneous lineups promote relative judgments. By viewing all faces together, witnesses evaluate faces based on which one is closest to their memory, even if it doesn’t entirely fit what they remember. This “close enough” judgment is consequently not as accurate as an absolute judgment. It may not be surprising then that, in addition to a lower rate of false identifications (i.e. picking a known-innocent filler), the number of correct identifications from sequential lineups is less than from simultaneous lineups (Clark, 2012). The fewer overall IDs made from sequential lineups suggests that this format promotes more conservative responding in witnesses than simultaneous lineups (Palmer & Brewer, 2012). This has important implications for the justice system. On one hand, these statistics suggest that witnesses given a sequential lineup are less likely to identify an innocent suspect; on the other, it is less likely that a guilty suspect would be identified. Both are serious consequences, though those in the justice system and policymakers have mostly agreed that the reduced chance of convicting an innocent person outweighs the drawbacks of a guilty person staying free (Wells, Small, Penrod, Malpass, Fulero, & Brimacombe, 1998).
The significant amount of evidence favoring sequential lineups lead to policy changes in how eyewitness lineups were conducted, with some cities going so far as to mandate that only sequential lineups be used in their police departments (Pratt & Dunbar, 2013). However, recent research complicates the simple conclusion that sequential lineups lead to better witness discrimination than simultaneous lineups. Earlier studies computed witness discrimination with a simple ratio: the number of correct identifications divided by the number of known-innocent filler identifications. For example, if a witness made four correct identifications and eight filler identifications (false IDs), the witness’ diagnosticity ratio would be 0.5. Recent studies employing more sophisticated statistical techniques do not show a benefit of sequential lineups (Andersen, Carlson, Carlson, & Gronlund, 2014; Mickes, Flowe, & Wixted, 2012). Instead, results suggest that the earlier results favoring sequential lineups may be due to more conservative responding rather than better discrimination. In essence, sequential lineups disproportionately reduce the number of filler identifications compared to the reduction in correct identifications (Steblay, Dysart, & Wells, 2011). Because known filler IDs are more common than missing a guilty suspect, this conservative responding gave sequential lineups better ratios than simultaneous lineups.
These new analyses rely on ROC curves, a method commonly used to evaluate the accuracy of recognition data, because it is not biased by how liberal or conservative a witness’ responses tend to be (Gronlund, Wixted, & Mickes, 2014). Once the base rate of making any type of identification is accounted for, there is no difference between simultaneous and sequential lineups, if not a slight advantage for simultaneous lineups (Carlson & Carlson, 2014; Gronlund et al., 2012). The diagnostic feature detection theory (Wixted & Mickes, 2014) seeks to explain why a simultaneous lineup may lead to better discrimination than a sequential lineup. According to this theory, simultaneous lineups may lead to better discriminability because viewing all lineup faces (suspect and known innocent fillers) together allows witnesses to compare and contrast specific features in each face. Witnesses can observe common features that are shared by many or all of the faces that might be non-diagnostic and base their decision on unique features of a face that would be diagnostic.
Furthermore, sequential lineups may have position effects. When evaluating faces in a sequential lineup, eyewitnesses may be more conservative early on in the lineup compared to later in the lineup. This position effect arises because witnesses may be more concerned with making a misidentification early on in the lineup, promoting conservative judgments, but towards the end of the lineup may become worried that they will make no identification and subsequently use more liberal judgments (Carlson, Gronlund, & Clark, 2008). Position effects are concerning because they suggest that a suspect’s order in the lineup affects their probability of being picked, rather than if they are guilty. Yet, it should be noted that not every study has found these position effects (Carlson & Carlson, 2014; Wells, Steblay, & Dysart, 2011). Some have also suggested that there could be position effects in simultaneous lineups, with identifications influenced by whether the suspect’s photo is in the center of the photo array or towards the edge (Wells, Steblay, & Dysart, 2015).
Research on eyewitness identifications is complicated by the difficulty of relating laboratory findings to real world cases. For instance, I discussed earlier about how sequential lineups generally lead to more conservative responding than simultaneous lineups. Those who view a sequential lineup are less likely to make an ID compared to a witness viewing a sequential lineup. However, a large study of real-world cases in police departments that randomly assigned cases to either use a simultaneous or a sequential lineup found that there was no difference in response rates: Witnesses viewing a simultaneous lineup were just as conservative as witnesses viewing a sequential lineup (Amendola & Wixted, 2015). Given the high stakes decision-making of witness identification, it is plausible that witnesses in real world cases are more conservative than laboratory participant witnesses regardless of lineup format.
Moreover, conditions of lineups can differ significantly from the conditions of witness lineups in laboratory studies. Researchers are particularly concerned with “pristine conditions.” Pristine conditions are a combination of factors believed to improve the accuracy of witness identifications (Wells et al., 1998). Though the optimal lineup format is still hotly debated, these conditions are generally accepted as improving eyewitness identifications. Factors needed for pristine conditions include only one suspect per lineup, ensuring the suspect does not stand out in the lineup, cautionary instructions to the witness that the offender might not be in the lineup, double-blind administration so that the lineup administer cannot unconsciously bias the witness, and taking an indication of witness confidence at this initial identification (Wixted & Wells, 2017). At times, the advantage of simultaneous over sequential lineups, or vice versa, on witness discrimination has been found when there are pristine conditions. It is not clear if these findings would always hold if those conditions are not present, as is often true of actual cases.
There is still considerable debate as to whether simultaneous or sequential lineups produce better witness identifications. Those who support the absolute-relative judgment theory argue that sequential lineups outperform simultaneous lineups because it encourages absolute judgments against memory as opposed to comparisons of which face in the lineup seems the closest match. Researchers who support the diagnostic feature detection theory believe that simultaneous lineups lead to more accurate identifications because they allow witnesses to assess which features are diagnostic and discount nondiagnostic features, improving judgments. Research on these theories is ongoing and there is not yet overwhelming evidence for one or the other. In fact, there is mixed evidence for many of the findings discussed in this post, with some papers finding an effect and others failing to detect that same effect. Ongoing obstacles for researchers in this area include divergent findings between laboratory studies and “real-world” cases, as well as if and when these results hold when pristine conditions are not met.
References
Amendola, K. L., & Wixted, J. T. (2015). Comparing the diagnostic accuracy of suspect identifications made by actual eyewitnesses from simultaneous and sequential lineups in a randomized field trial. J Exp Criminal, 11, 263-284. DOI: 10.1007/s11292-014-9219-2
Andersen, S. M., Carlson, C. A., Carlson, M. A., & Gronlund, S. D. (2014). Individual differences predict eyewitness identification performance. Personality and Individual Differences, 60, 36–40. DOI: 10.1016/j.paid.2013.12.011
Carlson, C. A. & Carlson, M. A. (2014). An evaluation of perpetrator distinctiveness, weapon presence, and lineup presentation using ROC analysis. Journal of Applied Research in Memory and Cognition, 3, 45–53.
Carlson, C. A., Gronlund, S. D., & Clark, S. E. (2008). Lineup composition, suspect position, and the sequential lineup advantage. Journal of Experimental Psychology: Applied, 14, 118–128. DOI:10.1037/1076898X.14.2.118
Clark, S. (2012). Costs and benefits of eyewitness identification reform: Psychological science and public policy. Perspectives on Psychological Science, 7, 238–259.
Gronlund, S. D., Carlson, C. A., Neuschatz, J. S, Goodsell, C. A., Wetmore, S. A., Wooten, A., & Graham, M. (2012). Showups versus lineups: An evaluation using ROC analysis. Journal of Applied Research in Memory and Cognition, 1, 221-228.
Gronlund, S. D., Wixted, J. T., & Mickes, L. (2014). Evaluating eyewitness identification procedures using receiver operating characteristic analysis. Current Directions in Psychological Science, 23(1), 3-10. DOI: 10.1177/09633721413498891
Innocence Project. (2018, April 18). DNA’s revolutionary role in freeing the innocent. https://www.innocenceproject.org/dna-revolutionary-role-freedom/
Innocence Project. (2019). DNA Exonerations in the United States. Retrieved July 23, 2019. http://www.innocenceproject.org/dna-exonerations-in-the-united-states/
Lindsay, R. C. L., & Wells, G. L. (1985). Improving eyewitness identifications from lineups: Simultaneous versus sequential lineup presentation. Journal of Applied Psychology, 70, 556-564.
Mickes, L., Flowe, H. D., & Wixted, J. T. (2012). Receiver Operating Characteristic Analysis of Eyewitness Memory: Comparing the Diagnostic Accuracy of Simultaneous vs. Sequential Lineups. Journal of Experimental Psychology: Applied, 18, 361-376.
Palmer, M. A., & Brewer, N. (2012). Sequential lineup presentation promotes less-biased criterion setting but does not improve discriminability. Law and Human Behavior, 36(3), 247-255.
Pratt, T., & Dunbar, W. (2013, January 7). Austin police: Change to lineup procedure can cut down on false arrests, convictions. KUT. https://www.kut.org/post/austin-police-change-lineup-procedure-can-cut-down-false-arrests-convictions
Seale-Carlisle, T. M., Wetmore, S. A., Flowe, H., & Mickes, L. (2019). Designing police lineups to maximize memory performance. Journal of Experimental Psychology Applied. DOI: 10.1037/xap0000222
Steblay, N. K., Dysart, J., Fulero, S., & Lindsay, R. C. L. (2001). Eyewitness accuracy rates in sequential and simultaneous lineup presentations: A meta-analytic comparison. Law and Human Behavior, 25, 459-473.
Steblay, N., Dysart, J., & Wells, G. L. (2011). Seventy-two tests of the sequential lineup superiority effect: A meta-analysis and policy discussion. Psychology, Public Policy, and Law, 17, 99–139. DOI:10.1037/ a0021650
Wells, G. L. (1984). The psychology of lineup identifications. Journal of Applied Social Psychology, 14(2), 89-179. DOI: 10.1111/j.1559-1816.tb02223.x
Wells, G. L., Small, M., Penrod, S., Malpass, R., Fulero, S. M., & Brimacombe, C. A. E. (1998). Eyewitness identification procedures: Recommendations for lineups and photospreads. Law and Human Behavior, 22(6), 603-647.
Wells, G. L., Steblay, N. K., & Dysart, J. E. (2011). A test of the simultaneous vs. sequential lineup methods. Des Moines, IA: American Judicature Society. Retrieved from http://216.36.221.170/wc/pdfs/ EWID_PrintFriendly.pdf
Wells, G. L., Steblay, N. K., & Dysart, J. E. (2015). Double-blind photo lineups using actual eyewitnesses: An experimental test of a sequential versus simultaneous lineup procedure. Law and Human Behavior, 39(1), 1-14.
Wixted, J. T., & Mickes, L. (2014). A signal-detection-based diagnostic-feature-detection model of eyewitness identification. Psychological Review, 121(2), 262-276. DOI: 10.1037/a0035940
Wixted, J. T., & Wells, G. L. (2017). The relationship between eyewitness confidence and identification accuracy: A new synthesis. Psychological Science in the Public Interest, 18(1), 10-65. DOI: 10.1177/1529100616686966