Simultaneous, Sequential and Sequential Presentation Only line-ups in mediating Hit & False alarm rates
Exculpating individuals due to mistaken identifications in simultaneous line-ups have stirred questions of trust to the procedure. Researchers proposed a sequential presentation line-up to reduce the false identifications more than they reduce hits as a superior procedure. This study investigated if a novel sequential-presentation-only (SPO) line-up could draw the benefits of both procedure by hypothesizing (a) a sequential line-up will yield fewer hits and false alarms than a simultaneous line-up and (b) a sequential presentation only line-up will yield fewer false alarms than a simultaneous line-up without a significant reduction of hits. 713 participants assessed a set of images to an image of the perpetrator in different line-up conditions. The first hypothesis was partially supported while the second hypothesis was not. This study suggests that simple manipulations have potential to make the simultaneous line-up more reliable and a novel SPO line-up does not redeem the effects.
Eyewitness identifications are among the most persuasive, and sometimes only, juncture in the apprehension of criminals. This typically involves a simultaneous line-up(SIM) where the suspect(target) is placed among known innocents(foils) who resemble the witness’s description of the perpetrator (Wells & Olson, 2003). The selection or lack of selection from the witness is given significant legal weighting. However, 75% of convictions involved exculpation through DNA testing where eyewitness misidentification was at fault. Furthermore, in 38% of these cases, multiple witnesses have misidentified the same innocent person (Project, 2009), which brings to question the accuracy of the procedure.
Lindsay and Wells (1985) devised the sequential line-up (SEQ) procedure as a better alternative. Each line-up member is presented one at a time and witnesses must decide if the line-up member matches the perpetrator before moving on to the next. Witnesses are unaware of the number of members shown, similar to real world cases, where each member is shown once. The “superiority effect” stems from the enhanced overall accuracy as SEQ reduces false identifications(false alarms) when the target is absent(target-absent conditions), more than it reduces correct identifications (hits) when the target is present (target-present conditions) (N. Steblay, Dysart, Fulero, & Lindsay, 2001).
The differences in line-up performance can be attributed to witness’s decision strategies (Gronlund, 2004). In SIMs, witnesses employ a relative judgement strategy where they evaluate the similarity of line-up members to their recollection of the culprit relative to one another, even when the resemblance Is vague (McQuiston-Surrett, Malpass, & Tredoux, 2006; N. K. Steblay & Phillips, 2011). In target-present conditions, the perpetrator is the closest match, resulting in more hits (N. Steblay et al., 2001). In comparison, target-absent conditions risk foils with the closest resemblance to be picked, in effect producing more false alarms. This flaw is further enhanced when the dissimilarity of appearance in line-up members increases (Charman, Wells, & Joy, 2011). Accordingly, presenting a line-up sequentially is said to eliminate relative judgements and to allow absolute comparisons to each line-up member exclusively to memory (Lindsay & Wells, 1985; N. K. Steblay & Phillips, 2011). However, in target-present conditions, sequential line-ups produce lower hit rates compared to simultaneous line-ups (McQuiston-Surrett et al., 2006).
The reduction in hit rate is better understood with signal detection theory (SDT) (Meisser, Parker, Parker, & MacLin, 2005). SDT posits that our ability to recognise and differentiate between familiar and novel stimuli rests on our response criterion and discrimination accuracy. Discrimination accuracy is the ability of an individual to correctly detect a signal (hits) vs. correctly reject its absence (correct rejections), while response criterion is the strength of evidence required before a signal (hit) is registered. In line-ups, the response criterion is familiarity-based and if a line-up member exceeds the familiarity threshold and corresponds to the witness’s memory of the perpetrator, it produces a hit, or otherwise it is rejected (Gronlund, 2004).
It is important to note that witnesses lack awareness of the number of line-up members they will be shown in sequential line-ups. This raises the criterion threshold, which means more hits will be unlikely and more misses are produced. Furthermore, since witnesses cannot revise their previous decisions on a line-up member, they are subjected to a conservative response bias (McQuiston-Surrett et al., 2006). In effect, this reduces the hit and false alarm rates.
Ideally, a line-up procedure that employed absolute judgement without a criterion shift would confer the best of both simultaneous and sequential procedures. This maximises hit rates while minimises false alarms, optimising the discrimination accuracy. This study examines a novel type of sequential procedure, named sequential presentation only (SPO), which theoretically can do so. The SPO involves line-up members being shown one at a time, while leaving the decision making until after all line-up members have been shown. This retains absolute judgement in the decision making process while decreasing response bias. As a result, the hit rates should be comparable to those of a simultaneous line-up.
It follows that this study hypothesises in target-present manipulations (a) a sequential line-up will yield fewer hits and false alarms than the simultaneous line-up and (b) a sequential presentation only line-up will yield fewer false alarms than the simultaneous line-up without a significant reduction in hits.
The participants were 713 PSYC20007 Cognitive Psychology students who completed the task in groups as part of a laboratory experiment. Students were randomly assigned to conditions with 240 in the Simultaneous presentation condition, 229 in the Sequential presentation condition and 218 in the Sequential Presentation Only condition. 26 Participants were removed for having incomplete data files.
Stimuli and Apparatus
On each trial, the words “Get Ready” were presented for 1000 ms followed by the presentation of a target face (the perpetrator), which was presented for 500 ms and was immediately backward masked by a scrambled image of that face presented for 1000 ms. The line-up was then presented, and the participants response recorded.
In the simultaneous presentation condition, all five faces were presented simultaneously in a row across the centre of the screen with a small gap between each face along with identifying numbers 1 to 5. The number of remaining trials was displayed on the screen at this stage. Participants were instructed to respond with 1 to 5 indicating the line-up member that they thought was the target or to respond 6 if the target was not present in the line-up. The experiment then advanced to the next trial.
In the sequential condition, the five line-up members were presented one at a time for until a response was made. For each line-up member, the participant made a response (yes or no). There was a 1000 ms blank interval between each face. Once all of the five line-up members were completed, the participants were informed of the number of remaining trials for 1500 ms, and the experiment advanced to the next trial.
In the Sequential Presentation Only condition, the five line-up members were presented one at a time for 1000 ms each. There was a blank interval of 1000 ms between each line-up member. After the final line-up member, the response scale was presented until a response was made. The number of remaining trials was displayed on the screen at this stage. Participants were instructed to respond with 1 to 5 indicating the line-up 7 member that they thought was the target or to respond 6 if the target was not present in the line-up. The experiment then advanced to the next trial.
In each condition, the line-up was constructed from a set of five faces drawn from a set of 54 possible faces. On target present trials, the target was drawn randomly from the set of line-up faces. On target absent trials, the target was drawn from the remaining 49 faces. There were 50 trials in the experiment.
The mean ratings for the Simultaneous, Sequential and SPO condition as a function of Hit rates and False alarms are shown in Figure 1.
Figure 1. Mean hit rates and false alarms as a function of line-up condition
A one-way ANOVA showed a significant difference in the hit rate, F(2,684) = 12.62, p < .001, ?2 = .04. A post-hoc analysis using the Bonferroni correction for multiple comparisons showed significant difference in the hit rate between the SIM condition and the SEQ condition (Mean Difference = 0.08, p < .001) and between the SIM and SPO condition (Mean Difference = 0.05, p = .02), but not between the SEQ and the SPO condition (Mean Difference = 0.04, p = .10).
A one-way ANOVA showed a significant difference in the false alarms, F(2, 684) = 9.28, p < .001, ?2 =.03. A post-hoc analysis with the Bonferroni correction for multiple comparisons showed significant difference in the false alarm rate between the SIM and SPO condition (Mean Difference=0.69, p < 0.01), but not between the SIM condition and the SEQ condition (Mean Difference= 0.01, p=1.00) as well as the SEQ and the SPO condition (Mean Difference= 0.06, p= .004).
According to Cohen (1988), both hit rate and false alarms demonstrated a small to medium effect of line-up procedure.
This study investigated if a novel SPO line-up could attain optimal discrimination accuracy. In target-present conditions, it was hypothesized that (a) sequential line-up will yield fewer hits and false alarms than the simultaneous line-up and (b) the SPO line-up will yield fewer false alarms than the simultaneous line-up without a significant reduction in hits. The first hypothesis was partially supported as there were fewer hit rates but not false alarms in the SEQ condition. Our second hypothesis was not supported. Results showed that the sequential condition produced less hits compared to the simultaneous condition, this was consistent with previous research (Lindsay & Wells, 1985). However, false alarm rates were comparable. As for the SPO condition, a significantly lower mean hit rate was produced, as well as a higher false alarm rate compared to both the other conditions.
Simultaneous line-ups induce a relative judgement decision approach (Gronlund, 2004). In a target-present condition such as ours, the line-up member that most resembles witness’s memory of the culprit will induce a higher mean hit rate (N. Steblay et al., 2001). This was supported in our results. In a simultaneous target-absent line-up, the foil with most resemblance should be picked with similar reasoning, producing more false alarms. However, our results showed that false alarm rates were higher in the sequential line-up instead.
Sequential line-ups cause a criterion shift, influencing a conservative “no” or “unsure” response in witnesses which reduces hits and false alarm rates (McQuiston-Surrett et al., 2006). A possible explanation our results did not replicate this is the use of instructions which states explicitly the target may be absent or present. Instructions may prompt witnesses in realising that a criminal’s absence was a genuine possibility (N. Steblay et al., 2001), and thus will consider each person in succession (absolute judgement strategy) (Dysart & Lindsay, 2001). Since simultaneous line-ups do not induce a criterion shift, witnesses have less of a conservative bias and will refrain from guessing more (Palmer & Brewer, 2011). This accounts for the high rate in the simultaneous line-up and low false alarm rates compared to the sequential condition. However, since the same instructions were given to both conditions, it does not explain the unexpected results in the sequential condition. Perhaps showing all the line-up members in one sitting works at ease to witnesses’ memory and hence decisions (Smith et al., 2014). Further investigations need to verify this result.
Participants in the sequential condition refrain from making an identification as a result of the criterion shift (Palmer & Brewer, 2011). This reduces the overall hit rate. To negate this effect, all decision making is reserved to the end of the line-up procedure in the SPO condition. However, it still produced a significantly lower mean hit rate compared to the simultaneous condition, suggesting a criterion shift is unaffected by when participants report their decisions.
Furthermore, the results indicate that the difference between the simultaneous and SPO line-ups were not significant in regards to false alarms. The SPO condition was created to retain the low false alarm rates by appealing an absolute judgement framework. It appeals to memory in the sense of match-making, instead of a relative judgement among line-up members (Gronlund, 2004). By theory, this would discourage false alarms from occurring. Nonetheless, our results were comparable between the simultaneous and SPO procedures. This could also be explained in terms of the effect of imposing instructions. The caution that it gives participants makes it more likely in minimizing false alarm rates in the simultaneous condition, but may not be as dominant in a sequential set-up.
This study included presenting photographs of suspects as opposed to actual human entities in line-up conditions. The photographs only contained the suspect’s neck to facial features, excluding their physical build-up. This means that our line-ups may not include ecological validity (McQuiston-Surrett et al., 2006) and should be validated in similar trials and psychological concepts before being used in practice.
In conclusion, this study found that the simultaneous condition have potential in possessing optimal discrimination accuracy in regards to more hits and fewer false alarms. This is a contradiction to previous research as the superiority effect of the sequential condition may not hold even under target-absent settings. This finding followed when validating a novel SPO procedure was ineffective. A lower hit rate and higher false alarm rate resulted, becoming the worst performer of all conditions. However, it should be taken into consideration photographs cannot replicate real world ecology and may affect response criterion. Further research should focus on binding the findings to psychological concepts related to memory and decision making in line-up procedures.
Charman, S., Wells, G., & Joy, S. (2011). The Dud Effect: Adding Highly Dissimilar Fillers Increases Confidence in Lineup Identifications. Law & Human Behavior (Springer Science & Business Media B.V.). 35(6), 479-500.
Cohen, J. (1988). Statistical Power Analysis for the Behavioural Sciences (2nd ed.). New Jersey: Lawrence Erlbaum Associates.
Dysart, J. E., & Lindsay, R. C. L. (2001). A Preidentification Questioning Effect: Serendipitously Increaseing Correct Rejections Law and Human Behaviour, 25(2).
Gronlund, S. D. (2004). Sequential line-ups: Shift in criterion or decision strategy? Journal of Applied Psychology, 89(2), 362-368.
Lindsay, R. C., & Wells, G. L. (1985). Improving eyewitness identifications from lineups: Simultaneous versus sequential lineup presentation. Journal of Applied Psychology, 70(3), 556-564.
McQuiston-Surrett, D., Malpass, S. R., & Tredoux, C. G. (2006). Sequential vs. Simultaneous lineups: A review of methods, data, and theory. Psychology, Public Policy, and Law, 12(2), 137-169.
Meisser, C. A., Parker, C. G., Parker, J. F., & MacLin, O. H. (2005). Eyewitness decisions in simultaneous and sequential lineups: A dual-process signal detection theory analysis. Memory & Cognition, 33(5), 783-792.
Palmer, M. A., & Brewer, N. (2011). Sequential lineup presentation promotes less-biased criterion setting but does not improve discriminability. Law & Human Behavior, 36(3), 247-255.
Project, T. I. (2009). Reevaluating Lineups: Why witnesses make mistakes and how to reduce the chance of a misidentification. .
Smith, A. M., Bertrand, M., Lindsay, R. C. L., Kalmet, N., Grossman, D., & Provenzano, D. (2014). The Impact of Multiple Show-Ups on Eyewitness Decision-Making and Innocence Risk. Journal of Experimental Psychology, 20(3), 247-259.
Steblay, N., 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 Behaviour, 25(5), 459-473.
Steblay, N. K., & Phillips, J. D. (2011). The not-sure response option in sequential lineup practice. Applied Cognitive Psychology, 25(768-774).
Wells, G. L., & Olson, E. A. (2003). Eyewitness testimony. Annual Review of Psychology, 54(1), 277-295.