Clark, S. E., & Davey, S. L. (2005). The target-to-foils shift in simultaneous and sequential lineups. Law and Human Behavior, 29, 151-172.
These researchers were seeking to examine relative responding with the target-removed lineup, which was first employed by Wells (1993). They reasoned that since sequential lineups are not as influenced by relative decision processes, then the target-to-foil shift should be smaller in sequential lineups. The target-to-foil shift involves the idea that if participants picked the perpetrator in the target-present lineup, then they should reject the lineup when the target is removed from the lineup. However, participants tend to pick foils in the lineup instead of correctly rejecting the lineup, creating a target-to-foil shift.
They conducted two experiments to examine this question. Participants viewed a videotape of a crime, and then they viewed either a simultaneous or sequential lineup after a short delay. In the first experiment, they found that simultaneous and sequential lineups had equal target-to-foil shifts in lineups with the target removed. They also had interesting findings with order effects and similarity of foils to the target in the sequential lineup but not the simultaneous lineup, depending on the position in the lineup of one of the foils. When the next-best alternative was in the second position, many participants chose him, but the perpetrator had actually not been seen yet. So, the sequential advantage occurred only when the next-best alternative was shown after the perpetrator because participants had already identified the perpetrator.
The authors were concerned about the similarity of the next-best alternative to the perpetrator in the first experiment, so they conducted a second experiment with a less similar foil. All the other procedures from the first experiment were the same in the second experiment. A target-to-foil shift occurred equally for both types of lineups again, which indicates that similar decision processes may underlie both types of lineups. Similar order effects occurred in the second experiment, even when the foil was less similar to the perpetrator. An interesting finding from the second experiment was a criterion shift occurred for the next-best alternative shown later in the lineup, indicating that some participants lowered their decision criterion to make an identification.
The authors discussed the findings about the similarity of the foils and made the argument that similarity of the foils may affect sequential lineup administration. Since the target-to-foil shift occurred for both lineup types, they argued that there is more to be discovered about the decision processes for these types of lineups.
By HC
Though I have some experience with lineups in general, various forms of administration, and issues surrounding all lineups and their administration, I learn something new every time I read an article. This time around, I learned some interesting things about the specifics behind positioning and order effects, and also about the target-to-foil shift. I think Clark and Davey (2005) do a nice job presenting the argument that the seminal work of Lindsay and Wells circa 1985 may have jumped the gun a bit on claiming that simply due to a lack order or positioning effects found in there studies, that the sequential lineup is superior. Yes, much research has found that with TR/TA lineups, you are better off using sequential. Yet, according to research, this superiority drops off in TP lineups.
Additionally, the results of Clark and Davey (2005) suggest that a premature decision may have been made by Lindsay and Wells (1985), especially given that sequential lineups still produce error in terms of order effects. Thus, if a TP lineup, the target is shown in p-2 and the next-best alternative is shown in p-4, then you are more likely to have a correct id, b/c they will probably pick the target when it is shown early. If, however, you reverse the order in which the target and the NB appears in the lineup, you can really screw things up in terms of the witness making a decision that the NB was the person they saw, simply because the NB was presented before the target in the sequence, and thus the target never had a chance at being correctly identified.
Furthermore, the criterion shift worries me as well. If you have high-similarity foils in a TR lineup then the witness doesn’t have to lower their criterion to make a decision, but when you have a bunch of faces in a TR lineup that don’t really look like the target, instead of the witness rejecting the lineup (correct decision), they lower their criterion and choose the low-similarity, but NB alternative? This indicates that the field may be a bit overzealous about the superiority of sequential lineups. In TR lineups, you are still going to get errors with the sequential lineup, apparently regardless of similarity. I would argue as well that lowering the criterion (making them more liberal and likely to choose) is not something that is a good thing, and that we as researchers should not push without further evidence, because it could very well be the case that by saying sequential are better, they may be used more, and in turn, even if the witnesses are given proper admonishments (i.e., may or may not be present), they are still privy to errors associated with order effects and criterion shifts in cases in which low-similarity foils comprise the lineup.