APPLIED COGNITIVE PSYCHOLOGY

Appl. Cognit. Psychol. 20: 859869 (2006) Published online 19 June 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/acp.1237

Memory Distortion in Eyewitnesses: A Meta-Analysis of the Post-identication Feedback Effect

SUMMARY Feedback administered to eyewitnesses after they make a line-up identication dramatically distorts a wide range of retrospective judgements (e.g. G. L. Wells & A. L. Bradeld, 1998 Journal of Applied Psychology, 83(3), 360376.). This paper presents a meta-analysis of extant research on post-identication feedback, including 20 experimental tests with over 2400 participant-witnesses. The effect of conrming feedback (i.e. Good, you identied the suspect) was robust. Large effect sizes were obtained for most dependent measures, including the key measures of retrospective certainty, view and attention. Smaller effect sizes were obtained for so-called objective measures (e.g. length of time the culprit was in view) and comparisons between disconrming feedback and control conditions. This meta-analysis demonstrates the reliability and robustness of the post-identication feedback effect. It reinforces recommendations for double-blind testing, recording of eyewitness reports immediately after an identication is made, and reconsideration by court systems of variables currently recommended for consideration in eyewitness evaluations. Copyright # 2006 John Wiley & Sons, Ltd.

Media coverage of DNA exonerations has highlighted the fact that mistaken eyewitness identications can result in wrongful convictions of innocent suspects (e.g. Doyle, 2005, www.innocenceproject.org). Long before the problem of eyewitness misidentication reached public consciousness, however, psychological researchers explored the memory and social inuence processes underlying identication errors. Recently, this research has generated procedures designed to minimize the likelihood of a false identication (Davies & Valentine, 1999; Technical Working Group for Eyewitness Evidence, 1999; Wells, Small, Penrod, Malpass, Fulero, & Brimacombe, 1998). Current recommendations for police lineups include ve core components: effective use of llers (e.g. Wells, Rydell, & Seelau, 1993); blind administration of the line-up (e.g. Douglass, Smith, & Fraser-Thill, 2005; Phillips, McAuliff, Cutler, & Kovera, 1999); a cautionary instruction to the witness that the culprit may or may not be present in the set of photos (Malpass & Devine, 1981; Steblay, 1997), sequential rather than simultaneous presentation of photos (e.g. Lindsay & Wells, 1985; Steblay, Dysart, Fulero, & Lindsay, 2001), and obtaining a statement of certainty from the witness at the time of the identication decision (e.g. Luus & Wells, 1994). As researchers continue to advance knowledge of best line-up practices, a number of
*Correspondence to: A. B. Douglass, Bates College, 4 Andrews Road, Lewiston, ME 04240, USA. E-mail: [email protected]

Copyright # 2006 John Wiley & Sons, Ltd.

860

A. B. Douglass and N. Steblay

jurisdictions in the United States are bringing these science-based recommendations to effective eld practice (see Klobuchar, 2005). Line-up research recently has produced an ancillary line of investigation focusing on the integrity of an eyewitnesss recollections after the line-up decision is made. This growing body of literature has revealed the astonishing power of a casual comment from a line-up administrator to affect eyewitness memory. The rst study to examine this effect (Wells & Bradeld, 1998) demonstrated that conrming post-identication feedback received by the witness immediately after the identication (i.e. Good. You identied the actual suspect.) signicantly inated retrospective condence reports when compared with a control group told nothing about the accuracy of the identication (participants indicated how certain they were at the time of their identication). Perhaps more alarming is that an extensive range of variables was inated in conjunction with retrospective certainty, including witness reports of the quality of their view of the perpetrator, how much attention was paid, ease of the identication, and basis for the identication. Participants who received conrming feedback were also more willing to testify about their identication and reported a greater ability to remember strangers. The post-identication feedback effect bears a resemblance to Fischhoffs hindsight bias (1977) in which participants given the correct answer to a decision indicated how they would have responded, had they not known the correct answer. Participants estimates of their own accuracy were routinely higher than the actual accuracy of participants who made the same decision without knowing the correct choice. There are two important differences between Fischhoffs paradigm and the post-identication feedback paradigm. First, in the feedback paradigm participants cannot misremember their prior decision because feedback is administered immediately after the identication is made. Second, participants are asked to recall judgements surrounding a decision, rather than a decision itself (see Bradeld & Wells, 2005, for a fuller discussion of these issues). Therefore, the post-identication feedback effect demonstrates that outcome information can distort memories beyond the boundaries rst outlined by Fischhoff. Subsequent research has replicated the post-identication feedback ndings with variations in experimental design designed to explore their theoretical underpinnings. One explanation for the effect hypothesized that participants do not consider their judgements before being queried in the dependent measures questionnaire. At that time, the only way to consider judgements about the witnessed event and the identication procedure is through the lens of the feedback received. One set of experiments explicitly tested this possibility. Wells and Bradeld (1999) manipulated whether participants were instructed to think about testimony-relevant judgements before receiving feedback. Participants who answered questions about their certainty before hearing feedback were inoculated against the effect of feedback on the certainty dependent measuretheir judgements did not show the typical post-identication feedback ination on retrospective certainty. Research in a related paradigm demonstrated a similar ability of prior thought to protect participants against the memory distorting effects of feedback (Bradeld & Wells, 2005). The post-identication feedback effect is noteworthy for multiple reasons. First, eyewitnesses in the feedback paradigm typically have made identications from targetabsent photospreadsall of their identications are inaccurate. Consequently, their distorted reports correspond to mistaken identications of innocent suspects, a forensically relevant scenario of critical importance given the eyewitness errors exposed by DNA exoneration cases (e.g. Davies, 1996; Rattner, 1988). Second, this powerful effect is
Copyright # 2006 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 20: 859869 (2006)

Post-identication feedback meta-analysis

861

produced by a simple and seemingly casual comment from the line-up administratora system variable (Wells, 1978) that potentially could be controlled in police practice. Third, the aspects of eyewitness experience distorted by post-identication feedback (e.g. certainty, witness perception of his/her view of the perpetrator, attention given to the witnessed event, ease of identication) are the very attributes that are likely to bolster eyewitness credibility in the eyes of investigators, prosecutors, and juries. Research has established that people who evaluate eyewitness identications routinely and naturally assume that condence (certainty) is correlated with accuracy (e.g. Leippe, 1994) and continue to use condence to assess accuracy even after being told that the two are not reliably linked (Fox & Walters, 1986). Finally, court systems have explicitly recommended using some of the very criteria distorted by post-identication feedback in evaluations of eyewitnesses. The US Supreme Court recommends using certainty, view, and attention reports (e.g. Neil v. Biggers, 1972); courts in England and Wales recommend using view and attention (R v Turnbull, 1977); the Australian Law Reform Commission (2005) is currently reviewing jury instructions regarding eyewitness identication evidence. Considering the research ndings reported above, these recommendations demand scrutiny. Since the post-identication feedback effect entered the published eyewitness literature in 1998, many researchers have explored this phenomenon. However, in spite of strong academic interest in the topic, there has not been a systematic organization and evaluation of the research. The current research aims to provide such structure using the tool of metaanalysis. Meta-analysis already has been useful in the psycho-legal realm, as it provides objective quantitative indicators of the status of a hypothesized effect, detailed analysis of effect moderators and direction for advances in research design, theory and practice (see recent meta-analyses on topics of line-up instruction, Steblay, 1997; sequential presentation, Steblay, et al., 2001; and showups, Steblay, Dysart, Fulero, & Lindsay, 2003.). Systematic evaluation of extant research on the post-identication feedback effect will assist future researchers by guiding the selection of variables and experimental paradigms that can target the causes and parameters of the effect. Equally important, this metaanalysis will provide a summary of knowledge for a broader audience that includes line-up administrators and court personnel. Line-up administrators are often interested in learning about strategies for obtaining eyewitness evidence that are immune to challenges from the defence. Similarly, court personnel (including defence attorneys) are interested in hearing from experts about procedures that might have compromised the integrity of the eyewitnesss memory. With a meta-analysis as a foundation for their recommendations, experts involved in conversations with these constituencies will be able to inform both more comprehensively.

METHOD Sample The sample included 20 experimental tests from 14 studies. Studies were obtained from a search of PsycInfo and additional conversation with researchers within this area of expertise. The nal sample included 10 published and 4 unpublished studies, representing 2477 participant-witnesses. The majority of the studies were conducted in the United
Copyright # 2006 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 20: 859869 (2006)

862

A. B. Douglass and N. Steblay

States (n 11) with others conducted in the United Kingdom (n 1), and Australia (n 2). In order to be included in the meta-analysis, the study must have included a laboratory test of the conrming feedback effect, retrospective certainty as a dependent variable, and data that could provide calculation of an effect size for the comparison of a group that received conrming feedback to a control group. The studies included in this analysis were conducted between 1998 and 2005 using participants who ranged in age from 11 to 97; most were college students. Participants of both genders were included in 100% of the studies. All studies used videotaped stimuli as the witnessed event with a range of length from 60 seconds to 180 seconds and required participants to make an identication from a photospread containing colour photographs. One exception was Bradeld, Wells, & Olson, 2002 in which participants made an identication from a videotaped lineup. Sample sizes ranged from 62 to 320 (M 176.93). Dependent measures A total of 13 dependent measures were recorded from the 14 studies analysed, not all of which were included in each study analysed. The measures fell into three broad categories. First were retrospective judgements regarding the witnessed event. Measures in this category include: view, attention paid, ability to make outfacial features, basisfor an identication, quality of the culprits image in memory, distance of the camera from the perpetrator, and length of time the perpetrator wasin view. A second set of measures concerned aspects of participants identication experience: retrospective certainty, easeof identication and timeneeded to make the identication. Finally, measures concerning summative judgements were analysed: general ability to remember strangers, reports of trust in an eyewitness who had similar viewing conditions, and willingness to testify.

Statistics Cohens d, the standardized mean difference between two groups, was calculated as the effect size indicator for each comparison (Cohen, 1988). In the following results, d is used to indicate a mean effect size across tests. A meta-analytic Z (Zma) was calculated using Rosenthals (1991) method of combining t-values. This Zma provides an overall probability level associated with the observed pattern of results. A fail-safe N (Nfs) was calculated as a means to determine the number of unretrieved studies averaging null results necessary to bring the overall p-value to a specic level of signicance (in this case, p 0.05). This number of studies, or tolerance for future null results, allows us to evaluate the resistance of the review conclusion to a le drawer threat (Rosenthal, 1991).

Comparisons Eleven of the tests compared a conrming feedback (CF) condition to a no feedback (NF) control group. Six compared CF to disconrming feedback (DF) condition. Three compared DF and NF groups. The focus of our study is the rst comparison (CF vs. NF) as that is the forensically relevant contrast because of the inationary power of conrming feedback for a witness who has identied a suspect.
Copyright # 2006 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 20: 859869 (2006)

Post-identication feedback meta-analysis

863

RESULTS Primary analysis: Comparison between conrming feedback (CF) and no feedback (control) groups, on each dependent measure Certainty Certainty is arguably the most important dependent measure in the post-identication feedback paradigm. Participants receiving conrming feedback expressed signicantly more retrospective condence in their decision compared with participants who received no feedback (d 0.79, Zma 13.42, p < 0.0001). An effect size of 0.79 is considered large, based on Cohens rule of thumb (Cohen, 1988) (see Table 1). Biggers criteria Eyewitness certainty (noted above), opportunity to view the perpetrator, and attention paid to the event are qualities of the eyewitness viewing experience that, according to the US Supreme Court (Neil v. Biggers, 1972), are criteria relevant to juror decision-making (recommendations in England and Wales focus on view and attention). Participants retrospective reports of view and attention (ds 0.50 and 0.46, respectively) were signicantly affected by conrming feedback, producing medium effect sizes. Related subjective measures CF participants demonstrated consistently inated perceptions on subjective measures related to their line-up performance compared to the control group. Participants who received conrming feedback reported that they possessed a signicantly better basis for making the identication (d 0.77), greater clarity of the perpetrators image in mind (d 0.68), greater ease of identication (d 0.80), and needing less time to make their ID (d 0.45). They also reported a better memory for strangers faces (d 0.45) and greater trust in the memory of another witness with a similar experience (d 0.52). Not surprisingly, then, they are also more willing to testify about their identication decision (d 0.82).
Table 1. Conrming feedback vs. No feedback comparison: Retrospective reports Dependent measure Certainty at time of ID How good a view? Opportunity to view face Attention paid Good basis to make an ID Ease of making an ID Speed of ID Willing to testify My memory for strangers Clarity of image in my mind Trust in eyewitness with similar experience How far away? How long in view? Condence right now Tests 11 9 9 9 9 9 9 9 8 7 3 2 4 2 d 0.79 0.50 0.55 0.46 0.77 0.80 0.45 0.82 0.45 0.68 0.52 0.12 0.29 0.53 Range 0.20 0.02 0.04 0.27 0.56 0.35 0.12 0.43 0.19 0.30 0.41 0.10 0.09 0.31 (min, max) 1.27 0.90 1.04 0.67 1.10 1.02 0.67 1.13 0.84 1.17 0.71 0.13 0.69 0.75 Nfs 590 132 165 145 386 587 104 437 75 150 <1 <1 <1 19

Comparisons for all measures produced statistically signicant Zma values (p < 0.05), except for how far away and how long in view. Copyright # 2006 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 20: 859869 (2006)

864

A. B. Douglass and N. Steblay

Objective measures Smaller effect sizes and no statistically signicant differences were found for attributes of the participants experience that are ostensibly objective: time that the perpetrator was in view and distance from the camera to the perpetrator (ds 0.29, and 0.12, respectively). These smaller effect sizes are noteworthy for at least two reasons: They suggest some limits to the inuences of conrming feedback; they also indicate discernment and seriousness on the part of the research participants (i.e., participants were not simply employing a thoughtless response set across measures). However, they also tap information to which a subject could surmise the experimenter has accessknowable factsunlike an investigator in a real crime situation. Moderators The small number of studies available did not allow for extensive moderator analysis. However, it may be noted that post-identication feedback effects are quite robust. Overall, the studies involved a reasonably diverse sample of participants (undergraduates, children, adults) and stimulus materials. The effects were achieved for witnesses who made accurate identications in target-present lineups as well as false IDs in target absent arrays, although the effect is stronger for inaccurate witnesses (Bradeld et al., 2002). Semmler, Brewer, & Wells (2004) found post-identication effects for witnesses who rejected the lineup also (Hes not there). Semmler et al., also found the effects when a cautionary instruction (may or may not be in the lineup) was provided; Douglass and McQuiston-Surrett (in press) found the effects with both sequential and simultaneous lineups. Secondary analysis: Comparison between disconrming feedback (DF) and no feedback (Control) groups, on each dependent measure. Only three tests (in three separate studies) explored the impact of disconrming feedback on participants retrospective reports. These reports produce small average effect sizes and some inconsistencies. For dependent measures of view and ease of ID, participants receiving disconrming feedback indicate less positive retrospective reports in all three studies, (d 0.14 and 0.31, respectively). On measures of retrospective condence (d 0.21), ability to make out details of the face (d 0.04), attention (d 0.08), basis for ID (d 0.10), time to make an ID (d 0.01), and willingness to testify (d 0.10), the three tests show mixed resultstwo tests with less positive reports from the DF condition, one test with more positive reports (negative effect sizes indicate higher scores from the NF control condition).1

DISCUSSION Through this review, the reliability and robustness of the post-identication feedback effect are well documented. Over 2400 participant-witnesses have been tested, with remarkably consistent outcomes. Compared to control participants, those who receive a simple postidentication conrmation regarding the accuracy of their identication signicantly
1 Not surprisingly, a comparison of CF and DF conditions indicates substantial differences between the groups in retrospective certainty (d 1.07), view (0.69), memory for the face (0.77), attention (0.67), basis for judgment (0.79), ease of ID (1.01), time to make an ID (0.60), willingness to testify (1.01), memory for strangers (0.58), clarity of image in memory (0.99), and trust in an eyewitness with similar experience (0.56).

Copyright # 2006 John Wiley & Sons, Ltd.

Appl. Cognit. Psychol. 20: 859869 (2006)

Post-identication feedback meta-analysis

865

inate their reports to suggest better witnessing conditions at the time of the crime, stronger memory at the time of the lineup, and sharper memory abilities in general. Participants apparently make what would otherwise seem to be reasonable post hoc inferences about their witnessing experience and behaviour during the identication. However, these inferences are based on erroneous information external to their actual memory of these events. This creeping determinism (Fischhoff, 1975) produces a memory distortion that is by no means benign. The implications of these results are quite profound. Both memory for a crime and condence in ones memory are fragile and potentially slippery evidentiary elements. Indeed, one of the startling lessons of line-up research is just how powerful seemingly subtle aspects of line-up construction and investigator behaviour can be. The simple addition of a cautionary instruction (the perpetrator may or may not be in the lineup) produces a signicant (25%) drop in false identications (Steblay, 1997); and use of a sequential lineup cuts the false identication rate almost in half (by 23%) in target absent lineups (Steblay et al., 2001). Similarly, subtle changes in investigator behaviour derived from the knowledge an investigator has about the identity of the suspect can inuence witnesses identication decisions (e.g. Douglass et al., 2005; Phillips et al., 1999) and condence (Garrioch & Brimacombe, 2001). Although the present data do not allow for a precise calculation of the number of errors that could be avoided with a change in practice, they do provide dramatic evidence that post-identication feedback can compromise the integrity of a witnesss memory. Wells and Bradeld (1998) made this point clearly when reporting participant-witnesses disproportionate use of the extreme end of the certainty scale: 50% of CF participants in that study indicated certainty of six or seven on a 7-point scale (compared with 15% using the extreme end of the scale in the DF condition). The relevance for real cases is clear. First, as noted by Wells and Bradeld (1998), it is reasonable to assume that an eyewitness must exceed some threshold of credibility in order for investigators and prosecutors to move ahead in their case against a suspect. Witnesses who reconstruct and enhance their report of both witnessing and identication procedures may well increase the likelihood that a case against that suspect will be pursued. Frighteningly, this enhancement is not due to increased accuracy, but to extra-memory factors. Second, witnesses with feedbackenhanced memories will likely be more compelling witnesses at trial, increasing the chances of a convictionan unwelcome outcome if an innocent suspect was identied. Clear understanding of the impact of post-identication conrmation can facilitate the goal of many eyewitness researchersprevent mistaken identications from resulting in wrongful convictions. This meta-analysis can help accomplish this goal in several ways. First, this research should provide police with a strong rationale as to why it is critically important to administer double-blind photospreads and to immediately record eyewitness condence. These procedures could decrease the likelihood that juries will be erroneously impressed by a falsely condent eyewitness. This is especially critical because at least one study demonstrates that participant-jurors are not sensitive to eyewitnesses who display condence that has inated over time (Bradeld & McQuiston, 2004). Additionally, this meta-analysis should inuence the treatment of information regarding post-identication feedback effects in court by providing attorneys and experts with a stronger foundation from which to argue that a witnesss memory could be distorted if double-blind procedures were not followed and immediate condence reports not recorded. For experts who testify in court, this meta-analysis will facilitate admittance of testimony on this topic. Most American courts now use the Daubert standard for admitting expert testimony, one element
Copyright # 2006 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 20: 859869 (2006)

866

A. B. Douglass and N. Steblay

of which is that the information presented by an expert must have achieved general acceptance in the relevant scientic community (Daubert v. Merrell Dow Pharmaceuticals, 1993). The consistency in outcomes demonstrated in this review lends credence to the argument that post-identication feedback effects should be generally accepted. Directions for future research The effects reported here are remarkably consistent, suggesting that future research will target explanations for the post-identication feedback effect rather than resolution of inconsistencies. One direction for future research is to identify ways in which to anchor the witnesss memory in the witnessing experience itself rather than in post-event information. Wells and Bradeld (1999) have found some success in moderating the post-identication effect with instructions to the witness to privately think about his or her condence and attributes of the witnessing experience prior to receiving feedback. The videotaping of lineup procedures may also provide the means to later remind a witness (as well as a jury) of his or her condence and perceptions at the time of the lineup (e.g. Kassin, 1998; Sporer, 1993). However, because even a 48-hour delay did not diminish distorted retrospective reports (Wells, Olson, & Charman, 2003), nding other ways to anchor witnesses memory is critical. Although the small number of studies did not allow for comprehensive analyses of moderator variables, there was enough evidence to suggest that objective measures are less susceptible to memory distortion than are subjective measures. Therefore, it might be worthwhile to examine this variable more systematically. Perhaps the difference is due to the fact that participants realize those questions can be evaluated for accuracy by the experimenters. Would the same difference appear in a paradigm where participants knew that experimenters did not have access to accurate answers (i.e. in a more ecologically valid paradigm)? Other directions for future research include pursuing explanations for conditions under which the feedback effect is diminished such as when disconrming feedback is administered. Although witnesses who receive disconrming feedback probably have minimal impact on the criminal justice system (i.e. because they do not testify in court), an explanation for the smaller effects of disconrming feedback could provide information about the nature of eyewitness memory and how it interacts with social inuence cues. Finally, researchers might pursue feedback analogues. Would learning about accuracy from sources outside the immediate identication experiencee.g. a news report, a prosecutor, another witness (cf. Luus & Wells, 1994)have the same distorting effects on retrospective condence and perceptions of the witnessing conditions? The current research reveals an increased willingness of witnesses to testify in court. Does this eagerness translate to differences in subsequent interview and/or courtroom behaviour? Recommendations The primary recommendation to be made from this meta-analysis is straightforward feedback to the witness should not be part of the identication procedure. There is also a straightforward strategic solution: use a blind line-up administrator, thoroughly record the line-up process, and obtain eyewitness reports (particularly condence) immediately after the identication. Currently, blind administrators are recommended in order to guard against memory errors during the line-up decision (e.g. Wells et al., 1998). In England and Wales, the Police and Criminal Evidence Act (PACE) dictates that an ofcer who is not
Copyright # 2006 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 20: 859869 (2006)

Post-identication feedback meta-analysis

867

involved in the investigation conduct the line-up procedure (although he or she does know who the suspect is, Davies & Valentine, 1999). Collateral benets of a blind administrator are afforded in that no feedback could be provided to the witness until after completion of the line-up procedure and documentation of testimony-relevant judgements (Technical Working Group for Eyewitness Evidence, 1999). A nal recommendation is directed at courts considering recommendations for evaluations of eyewitnesses. In the United States, the Supreme Court should reconsider its current recommendation for evaluations of eyewitness testimony. Of the ve criteria outlined by the US Supreme Court in Neil v. Biggers (1972), three are dramatically distorted by post-identication feedback: condence, attention and view. Similarly, the Turnbull Rules in England and Wales also include two variables distorted by postidentication feedback: attention and view (R v Turnbull, 1977). Rulings from courts suggesting that these variables only be taken into account if post-identication feedback has not been administered would likely do much to decrease the incidence of feedback in real world cases. Barring reconsideration of these criteria in court recommendations, researchers should continue to press for immediate witness reports and blind line-up administration. These practices are best suited to prevent the memory distorting effects of post-identication feedback.

ACKNOWLEDGEMENTS We thank Margaret Mandeville and Emily Parker for help with data entry.

REFERENCES

Articles included in the meta-analysis

Australian Law Reform Commission (2005). Discussion Paper 69, Review of the Uniform Evidence Acts. Retrieved January 4, 2006 from http://www.austlii.edu.au/au/other /alrc/publications/dp/69/ Bradeld, A., & McQuiston, D. E. (2004). When does evidence of eyewitness condence ination affect judgments in a criminal trial? Law and Human Behavior, 28(4), 369387. Bradeld, A., & Wells, G. L. (2005). Not the same old hindsight bias: Outcome information distorts a broad range of recollections. Memory and Cognition, 33(1), 120130. Bradeld, A. L., Wells, G. L., & Olson, E. A. (2002). The damaging effect of conrming feedback on the relation between eyewitness certainty and identication accuracy. Journal of Applied Psychology, 87, 112120. Coddington, K. A., & Brigham, J. C. (2000, March). The malleability of eyewitness metamemory judgments: The effect of question difculty. Poster presented at the biennial meeting of the American Psychology-Law Society meeting. New Orleans, LA. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Daubert v. Merrell Dow Pharmaceuticals, Inc.,113 S.Ct. 2786 (1993). Davies, G. M. (1996). Mistaken identications: Where law meets psychology head on. The Howard Journal, 35, 232241. Davies, G. M., & Valentine, T. (1999). Codes of practice for identication. Expert Evidence: International Journal of Behavioural Sciences in Legal Context, 7(1), 5965. Douglass, A. B. (see Bradeld), & McQuiston-Surrett, D. (in press). Post-identication feedback: Exploring the effects of sequential photospreads and eyewitnesses awareness of the identication task. Applied Cognitive Psychology. DOI: 10.1002/acp1253
Copyright # 2006 John Wiley & Sons, Ltd. Appl. Cognit. Psychol. 20: 859869 (2006)

868

A. B. Douglass and N. Steblay

Douglass, A. B. (see Bradeld), Smith, C., & Fraser-Thill, R. (2005). A problem with double-blind photospread procedures: Photospread administrators use one eyewitnesss condence to inuence the identication of another eyewitness. Law and Human Behavior, 29(5), 543562. Doyle, J. M. (2005). True witness: Cops, courts, science, and the battle against misidentication. Palgrave Macmillin: New York. Fischhoff, B. (1975). Hindsight 6 foresight. Journal of Experimental Psychology: Human Perception & Performance, 1, 288299. Fischhoff, B. (1977). Perceived informativeness of facts. Journal of Experimental Psychology: Human Perception & Performance, 3, 349358. Fox, S. G., & Walters, H. A. (1986). The impact of general versus specic expert testimony and eyewitness condence upon mock juror judgment. Law and Human Behavior, 10, 215228. Garrioch, L., & Brimacombe, C. A. E. (2001). Lineup administrators expectations: Their impact on eyewitness condence. Law and Human Behavior, 25(3), 299315. Hafstad, G. S., Memon, A., & Logie, R. (2004). Post-identication feedback, condence and recollections of witnessing conditions in child witnesses. Applied Cognitive Psychology, 18, 901 912. Kassin, S. (1998). Eyewitness identication procedures: The fth rule. Law and Human Behavior, 22, 649654. Klobuchar, A., Steblay, N., & Caligiuri, H. (in press). Improving eyewitness identications: Hennepin Countys blind sequential lineup pilot project. Cardozo Public Law, Policy & Ethics Journal. Leippe, M. R. (1994). The appraisal of eyewitness testimony. In D. F. Ross, J. D. Read, & H. P. Toglia (Eds.), Adult eyewitness testimony: Current trends and developments (pp. 385418). New York: Cambridge University Press. Lindsay, R. C. L., & Wells, G. L. (1985). Improving eyewitness identications from lineups: Simultaneous versus sequential lineup presentation. Journal of Applied Psychology, 66, 7989. Luus, C. A. E., & Wells, G. L. (1994). The malleability of eyewitness condence: Co-witness and perseverance effects. Journal of Applied Psychology, 79, 714723. Malpass, R. S., & Devine, P. G. (1981). Eyewitness identication: Lineup instructions and the absence of the offender. Journal of Applied Psychology, 79, 714734. Neil v. Biggers. 409 U.S. 188 (1972). Neuschatz, J. S., Preston, E. L., Burkett, A. D., Toglia, M. P., Lampinen, J. M., Neuschatz, J. S., Fairless, A. H., Lawson, D. S., Powers, R. A., Goodsell, C. A. (2005). The effects of postidentication feedback and age on retrospective eyewitness memory. Applied Cognitive Psychology, 19, 435453. Phillips, M. R., McAuliff, B. D., Kovera, M. B., & Cutler, B. L. (1999). Double-blind photoarray administration as a safeguard against investigator bias. Journal of Applied Psychology, 84, 940 951. R v. Turnbull. Queens Bench 224 (1977). Rattner, A. (1988). Convicted but innocent: Wrongful conviction and the criminal justice system. Law and Human Behavior, 12(3), 283293. Rosenthal, R. (1991). Meta-analytic procedures for social research. Newbury Park, CA: Sage. Semmler, C., Brewer, N., & Wells, G. L. (2004), Effects of postidentication feedback on eyewitness identication and nonidentication condence. Journal of Applied Psychology, 89(2), 334346. Sporer, S. L. (1993). Eyewitness identication accuracy, condence and decision times in simultaneous and sequential lineups. Journal of Applied Psychology, 78(1), 2233. Steblay, N. (1997). Social inuence in eyewitness recall: A meta-analytic review of lineup instruction effects. Law and Human Behavior, 21(3), 283298. 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 Behavior, 25(5), 459. Steblay, N., Dysart, J., Fulero, S., & Lindsay, R. C. L. (2003). Eyewitness accuracy rates in showup and lineup presentations: A meta-analytic comparison. Law and Human Behavior, 27(5), 523540. Technical Working Group for Eyewitness Evidence. (1999). Eyewitness evidence: A guide for law enforcement. Washington, DC: National Institute of Justice; NCJ 178240. Wells, G. L. (1978). Applied eyewitness-testimony research: System variables and estimator variables. Journal of Personality & Social Psychology, 36(12), 15461557.

Copyright # 2006 John Wiley & Sons, Ltd.

Appl. Cognit. Psychol. 20: 859869 (2006)

Post-identication feedback meta-analysis

869

Wells, G. L., & Bradeld, A. L. (1998). Good, you identied the suspect: Feedback to eyewitnesses distorts their reports of the witnessing experience. Journal of Applied Psychology, 83(3), 360376. Wells, G. L., & Bradeld, A. L. (1999). Distortions in eyewitness recollections: Can the postidentication-feedback effect be moderated? Psychological Science, 10, 138144. Wells, G. L., Olson, E. A., & Charman, S. D. (2003). Distorted retrospective eyewitness reports as functions of feedback and delay. Journal of Experimental Psychology: Applied, 9, 4252. Wells, G. L., Rydell, S. M., & Seelau, E. P. (1993). The selection of distractors for eyewitness lineups. Journal of Applied Psychology, 78(5), 835844. Wells, G. L., Small, M., Penrod, S., Malpass, R. S., Fulero, S. M., & Brimacombe, C. A. E. (1998). Eyewitness identication procedures: Recommendations for lineups and photospreads. Law and Human Behavior, 22, 603647.

Copyright # 2006 John Wiley & Sons, Ltd.

Appl. Cognit. Psychol. 20: 859869 (2006)