VIEWS: 13 PAGES: 31 POSTED ON: 5/4/2011
Steering in the eyewitness identification procedure Wendy Alberts Jason Duncan Brian Wallace Prof. Steven Penrod Content Example Eyewitness misidentification Present study Illinois study Major hypotheses Procedure and design Results Conclusion 17 year old Ruben Cantu was convicted of murder and executed in 1993 . The prosecutor accused him of shooting his victim nine times. Now it is thought that he was likely innocent. The only eyewitness in this case was shown the picture of Ruben Cantu three times at which point he “recognised” Cantu. He has recanted and after the execution, a 15 year old accomplice testified that Ruben Cantu was never there…. Source: http://www.ad.nl/buitenland/article65986.ece Eyewitness Misidentifications In June 2008, the Cardozo Law School Innocence Project’s web site reports that 217 individuals convicted of crimes have been exonerated with DNA testing since 1992 (www.innocenceproject.org). 75% eyewitness misidentifications Most crimes do not include DNA-rich biological traces. Wells (1978) argues that two types of variables can be distinguished in eyewitness research: (a) estimator variables and (b) system variables Present Study The influence police officers can have on the eyewitness in the lineup procedure lineup presentation the presence of the perpetrator in the lineup giving cautionary instructions before the lineup procedure begins Lineup presentation Simultaneous display of six photographs Sequential lineup in which the photo’s of the suspect and foils are displayed one at a time (Wogalter et al., 2004). Sequential procedure reduces the number of false identifications with no loss in the number of accurate identifications (Lindsay & Wells, 1985; Cutler & Penrod, 1988; Lindsay et al., 1991; Keller et al., 2001; Steblay & Dysart, 2008). Illinois Study (Mecklenburg, 2006) No field studies until 2006 First field study comparing the simultaneous line-ups and sequential line-ups First field study to examine both photo arrays and live line-ups. However… Confounds blindness and procedure in its design Lower overall rate of false identifications in the single-blind simultaneous condition than predicted by the research experiments police officers may not have been reporting foil identifications when they knew those identifications were erroneous or steered the eyewitness to the identification of the suspect UK Archival Studies –9-person arrays [row percentages] Witnesses/ Suspect Foil No Arrays ID ID ID Slater 843/302 36% 22% 42% Wright & McDaid 1561/616 39% 20% 41% Valentine+(VIPER) 584/295 41% 21% 39% Pike+ 8,800/xx 49% - - W Yorkshire-trad 1,635/xx 35% - - W Yorkshire-VIPER 940/xx 39% - - Wright & (VIPER) 134/134 58% 21% 21% Average [wtd] 14494/xx 45% 21% 40% -- suspects—mix of guilty and innocent unclear -- 9-person arrays -- nearly 1/3??? of IDs are clearly “bad guesses” of foils] Witness Performance in Behrman and Davey (2001) and Behrman & Richards (2005)-Suspect IDs in Sacramento, CA area Presen- Suspect Foil No tations ID ID ID Showups BD 258 76% xxx 24% Photoarrays BD 284 48% ??% ??% Lineups* BD 58 50% 24% 26% Photo (424)+ 461 52% 15% 33% Live (37) BR Lineup Ave [wtd] 519 51% 16% 33% • 6-person arrays • About 1 in 4 IDs is a clear error Witness Performance from US Police Files-Stranger Cases Presen- Suspect Foil No tations ID ID ID California – Simul BD+BR 519 51% 16% 33% Minnesota Photo Seq 178 35% 11% 53% Illinois-Photo Simul 78 53% 1.3% 46% Illinois-Photo Seq+blind 64 44% 9.4% 46% Illinois-Live Simul 55 82% 0% 18% Illinois-Live Seq+blind 29 46% 5.4% 49% Queens, NY Live-Simul 2677 54% 3.0% 43% Possible reasons for the discrepancy are buried in the Illinois report: ◊The low number of filler IDs emerged from a non-blind simultaneous procedure in which it appears that the police could ignore IDs that did not meet their "Probable cause" standards (Appendix, pp. iii-iv) ◊This practice seems akin to the Queens practice in which "an identification [was recorded] only if it was based upon a high level of confidence, so that all tentative identifications were recorded as "no identification." (p. 43). Influence of lineup-administrator Haw & Fisher ( 2004): witnesses are more likely to make decisions consistent with the lineup administrator’s expectation when the level of contact between the administrator and the witness is high rather than low Garrioch and Brimacombe (2001): the confidence of the eyewitness rises when the witness chooses the lineup member who their lineup administrator thinks is guilty. Major Hypotheses 1. Police-officers can influence eyewitnesses to select a suspect and avoid foil identifications. 2. Police-officers can do this without the eyewitnesses realizing they are being led to the suspect or avoiding foils. 3. Police-officers can influence eyewitnesses in a non-blind simultaneous procedure more than in a non-blind sequential procedure. 4. Cautionary instructions have no effect on the identification decision of the eyewitness. Procedure and Design (2x3x2x2) 240 Participants (eighteen years or older) Two videos, one distracter video and one of a staged indoor theft Twenty minutes filler task Identification procedure Cautionary instructions: “The suspect may not be in the lineup, you don’t need to make a choice and you should not assume that the administrator knows who the suspect is.” Lineup where the thief was present or a lineup with a substitute for the thief (innocent suspect). Match to description procedure. ‘innocent suspect’: a foil who looked very similar to the perpetrator. Identification procedure (2) The line-up was administered either - simultaneous - sequential show-the-pictures-twice (S2) - sequential-with-stopping-rule (SS) sequential procedure: stack was back loaded to twelve cards (six blank cards), so the participants did not knew that the stack contained only six photographs. Identification procedure (3) An administrator who was trying to steer them to choose the (innocent) suspect/away from a foil or a blind condition Steering: The administrator signalled in the direction of the suspect with non-obvious, non- verbal cues and made comments like “good” and “that is helpful” and gave no reinforcements for the foils. Blind: screen between the administrator and the participant, so that the administrator couldn’t see the lineup and couldn’t influence the participant. Lineup was also randomised. Percentages of choices participants made in the Target Absent (TA) condition as a function of lineup procedure Lineup Procedure TA Lineup decision N SS(%) N S2 (%) N Sim(%) False Identification 5 11.9 10 25.0 10 26.3 all n.s. Correct Rejection 33 78.6 19 47.5 8 21.1 SS-S2** SS-SM ** S2-SM* Foil Identification 4 9.5 11 27.5 20 52.5 SS-S2* SS-SM** S2-SM* Total (N=120) 42 100.0 40 100.0 38 100.0 *p<.05 SS = Sequential-with stopping-rule procedure **p<.01 S2 = Sequential-show-the-photos-again procedure Sim = Simultaneous procedure Results First, procedure mattered: - Results show superiority of the sequential lineup procedure when it is properly conducted Table 2: Percentages of choices participants made in the Target Present (TP) condition as a function of lineup procedure Lineup Procedure TP Lineup decision N SS(%) N S2 (%) N Sim(%) Hit 8 20.5 18 42.9 19 48.7 SS-S2** SS-SM* S2-SM Lineup Rejection 23 59.00 11 26.2 7 17.9 SS-S2** SS-SM** S2-SM Foil Identification 8 20.5 13 31.0 13 33.3 all n.s. Total (N=120) 39 100.0 42 100.0 39 100.0 *p<.05 SS = Sequential-with stopping-rule procedure **p<.01 S2 = Sequential-show-the-photos-again procedure Sim = Simultaneous procedure Results The first hypothesis, that police officers can influence eyewitnesses to select a suspect and avoid foil identifications was confirmed. steering administrator: more suspect identifications. They could be steered in every procedure Percentages of choices in the Target Absent (TA) and the Target Present (TP) condition, as a function of administrator steering Administrator Steering Blind Lineup and decision N (%) N (%) Target Absent lineup False Identification 19 31.7 6 10.0** Correct Rejection 28 46.7 32 53.3 Foil Identification 13 21.7 22 36.7 Total (N=120) 60 100.0 60 100.0 Target Present Lineup Hit 34 56.7 11 18.3** Rejection Lineup 16 26.7 25 41.7 Foil Identification 10 16.7 24 40.0** Total (N=120) 60 100.0 60 100.0 Results The second hypothesis, that police officers can influence eyewitnesses to select a suspect and avoid foil identifications without the eyewitnesses realizing they were being led was not entirely confirmed. Witnesses in the steering condition mentioned things the administrator did or said that influenced their decision more than witnesses in the blind condition (χ2=30.84, p<.01). Witnesses also indicated that the administrator influenced their decision significantly more when the administrator was steering than when he was blind [t(233)=2.171, p<.01]. Results Influence of the administrator on their decision: - steering (M=2.73) vs. blind (M=2.10). low end of a 9-point scale Witnesses thought lineup photos and their memory were much more influential than the administrator: -photos steering (M=4.98) vs photos blind (M=5.77) (sg) -memory steering (M=7.32) vs memory blind (M=7.25) Results The third hypotheses, that police officers can influence eyewitnesses to select a suspect in a non-blind simultaneous procedure more than in a non-blind correct conducted sequential procedure (SS) was confirmed. Simultaneous procedure (55.3%) vs SStop procedure (26.8%) . Results No significant difference was found between witnesses with a S2 procedure (51.2%) and the simultaneous procedure (55.3%) when they were in the steering condition. That is, when the sequential procedure was not conducted properly (showing them the pictures twice), it was as easy to steer the witness to the (innocent) suspect as in the simultaneous procedure. Results The fourth hypothesis, that the cautionary instructions have no effect on the identification decision of the eyewitness was also confirmed. Conclusion If we compare the results of the present study with those of Illinois and Queens, we see that the steering administrator is a plausible explanation for the substantially lower overall rate of false identifications in the non-blind simultaneous conditions used in those studies. In the present study the witnesses made fewer foil identifications in the steering condition than the blind condition and did so without dramatically triggering witness suspicions. Questions? email@example.com 0(031)6-45092028
"Steering in the eyewitness identification procedure"