Actual Innocence Research

Eyewitness Identification is not 20/20

Guest author Holly Moon Poeschl on a case of wrongful conviction due to witness misidentification.

By Allison D. Redlich and Robert J. Norris

Although wrongful convictions have only gained consistent national attention in recent decades, scholars have studied the phenomenon for nearly a century. From Borchard (1932) through the most recent collection of cases from the National Registry of Exonerations (www.exonerationregistry. org), one finding has remained consistent: eyewitnesses often make mistakes, and those mistakes can lead to erroneous convictions. And while psychologists have been examining eyewitness identification error since at least the 1970s (e.g., Wells, 1978), our knowledge of the subject remains incomplete. This month's column calls attention to eyewitness identification, discussing the case of Wisconsin exoneree Steven Avery.

The guest-author this month is Holly Poeschl. Holly earned her Bachelor's degree in Psychology from Carroll University (Waukesha, WI), where she also majored in Criminal Justice and obtained a minor in Spanish. She is currently attending Marymount University in Arlington, VA where she is completing her Master's degree in Forensic Psychology. She is a panelist on the Marymount University Academic Integrity Committee and volunteers at a domestic violence shelter. Additionally, Holly has interned with both the Fairfax and Arlington County Office of the Public Defender to pursue her interests in mitigation, sentencing advocacy, and offender's rights.

Holly Moon Poeschl
Marymount University

According to the Innocence Project, eyewitness misidentification can be singled out as the number one reason for wrongful convictions. Although numerous research endeavors have shown that eyewitness identifications lack reliability (Kebbell, 2000; Lindsay, Wells, & O'Connor, 1989; Wells, Small, Penrod, Malpass, Fulero, & Brimacombe, 1998), this type of evidence is still frequently used to secure convictions. The unreliability of eyewitness identifications is the consequence of both the fallibility of human memory (Clark & Godfrey, 2009), and the type of identification procedure utilized (Wells et al., 1998). Like many exonerated individuals, the aforementioned reasons played a role in the wrongful conviction of a young man named Steven Avery. Our guest author will discuss the case, one of the contributing causes of his conviction, and finally areas of future research.

Case Overview

Information about the case was gathered from the Innocence Project.

On the afternoon of July 29, 1985, near Two Rivers, Wisconsin, Penny Ann Beernsten was jogging along the Lake Michigan shoreline. While she was nearing the completion of her workout, a man ambushed her from behind. Although Beernsten was physically fit, she simply could not overpower her assailant and was dragged into a wooded area where she was raped, choked, and left unconscious.

Upon regaining consciousness, Beernsten immediately reported her victimization and provided the police with a physical description of the perpetrator. From her depiction the police assembled a nine-option photo array. When presented with this identification task, Beernsten selected with confidence the photo of 22-year-old Steven Avery. Avery was arrested the next day.

During the trial prosecutors introduced two key pieces of evidence: Beernsten's identification of Avery, and a hair recovered from Avery's shirt determined by the State's forensic serologist to be consistent with Beernsten's. To aid in his defense, Avery presented 16 alibi witnesses. One of these defense witnesses was a store clerk in Green Bay. The clerk not only remembered Avery, his wife, and their five children shopping in his store, but also had a copy of Avery's receipt printed at 5:13pm on the day of Beernsten's attack. The clerk's testimony provided persuasive evidence that Avery could not have attacked Beernsten. If Avery was the attacker, after the assault (around 4:05pm) he would have had to walk a mile to the nearest parking area, drive home, situate his five children and wife in his vehicle, and then drive 45 miles to Green Bay. A police officer who testified for the state said he completed the trip in only 57 minutes. These 57 minutes, however, did not account for the time necessary to load the family in the car and were calculated by driving 10 miles per hour over the posted speed limit.

While the timeframe engendered doubt, on December 14, 1985, a jury of his peers found Steven Avery guilty of sexual assault, attempted murder, and false imprisonment. He was sentenced to 32 years in prison.

Avery tried to appeal his conviction but to no avail. In 1995, his petition for DNA testing was awarded. The results concluded the skin under Beernsten's fingernails contained the DNA of an unknown individual; however, Avery could not be excluded. In 2002 the Wisconsin Innocence Project advocated for DNA testing of the pubic hairs discovered at the crime scene. The Wisconsin Crime Laboratory concluded the hairs belonged to Gregory Allen. In light of this finding, Steven Avery was exonerated and released on September 11, 2003, after unjustly serving 17.5 years in prison.

Eyewitness Misidentification

Within the psychological community, results supporting the necessity of identification reform have been repeatedly substantiated. The Court system, although vague, has created some standards to help address the inaccuracies of eyewitness identification. The Court first ruled in Neil v. Biggers (1972) that five criteria must be considered when determining whether or not an identification procedure was likely to result in an uncorrectable misidentification: (1) opportunity to view, (2) amount of attention, (3) detail of description, (4) time before identification, and (5) certainty at time of identification. The Court then reaffirmed their previous ruling in Mason v. Braithwaite (1977), and added a two-pronged test for exclusion. The first prong determines whether the procedure was unnecessarily suggestive, and the second prong refers to whether or not the identification, even though improperly suggestive of a specific individual, was nevertheless reliable.

There are many distinctive identification procedures such as show-ups, line-ups, and photo arrays (Wells et al., 1989). While each identification task differs in suggestibility, each undertaking poses some risk (Wells et al., 1989). In order to address the issue of suggestibility and create a reliable identification procedure there are four accepted rules: (1) the individual conducting the line-up should be blind to who the suspect is, (2) the eyewitness should be told the criminal might not be present, (3) foils should be based on the eyewitness's verbal description of the criminal, and (4) confidence should be reported at the time of identification (Kebbell, 2000). Without knowing the specifics concerning the aforementioned rules, the nine-person photo array presented to Beernsten was inherently suggestive (Clark & Godfrey, 2009). The simultaneous nature of the photo array containing Avery and the eight fillers allowed for relative decision-making. Identification tasks that engender relative decision-making are suggestive because they allow the identifier to compare individuals and select the person who best fits what they recall (Wells et al., 1998). Ideally, an identification procedure should be sequential, this forces the witness to make an absolute judgment: is this the person I saw, yes or no (Wells et al., 1998).

Research Ideas

The remedy for reducing wrongful convictions grounded in eyewitness misidentification lies not only in the construction of an identification procedure, but also in what evidence is presented to the jury, and their perception of it. Research has found that jurors often place undue weight and confidence on eyewitness identifications (Lindsay et al., 1989), much like that of the jury in the case of Steven Avery. Since it is known that jurors tend to be heavily influenced by positive eyewitness identifications, one way to mitigate this over- reliance is to educate the jury. Education in the form of expert testimony is necessary to inform jurors about how identification procedures can increase the risk of false identifications, as well as how aspects associated with witnessing a crime can impact identification reliability (Cutler & Kovera, 2011).

Expert evidence often sways verdicts in a parallel direction to the presented testimony (Cutler & Kovera, 2011). In turn, future research should focus on jurors' perceptions of eyewitness expert testimony. Using surveys, researchers could acquire jurors' options regarding the presented expert testimony, and what qualities made the testimony reliable or unreliable (e.g., Cutler, Penrod, & Dexter, 1989). The information gained from such research endeavors is important to defense attorneys who have to overcome the hurdle of a positive eyewitness identification of their client. The information could help defense attorneys select an expert and ensure he or she is seen as reliable, which is critical. The information they present has the ability to teach jurors the inaccuracies of eyewitness identifications and in turn reduce the number of wrongful convictions.

Another area of future research regarding eyewitness expert testimony concerns attorney versus court-appointed experts. Pezdek (2006) postures court-appointed experts will be perceived as more impartial, however it is not known if and how perceived impartiality will impact juror decision-making. Additionally, it is not known how perceived impartially will impact the evidential weight assigned to the testimony. To study these issues, scholars could first look into the perceived impartiality between court- and attorney-appointed experts by conducting a mock trial using both types of experts. Questionnaires provided to the jury could be constructed to uncover what expert was perceived as more impartial, why, and how this impacted their verdict. Additionally, a question could be added to see if the evidential weight of the testimony differed by who presented it.

Finally, research should seek to discover how verdicts are impacted based on when expert testimony is presented during trial. Scholars could conduct mock trials, one where eyewitness expert testimony is presented prior to the introduction of any other evidence, and one where it is presented later in the trial. The rationale behind this study would be that jurors would be less inclined to rely on eyewitness identifications if education occurred prior to the presentation of eyewitness evidence itself (Pezdek, 2006). By conducting a survey post-trial, the evidential weight assigned to both the expert testimony and eyewitness identification could be compared. Collectively, the aforementioned research endeavors are important because understanding juries is critical in reducing wrongful convictions based on eyewitness misidentifications. Jurors are often the fact-finders of a case. If scholars gain insight into their decision-making process, changes can be made within the court system to mitigate jurors' over- reliance on eyewitness identifications.


In the case of Steven Avery, like others, Beernsten's positive identification of Avery as her attacker was a pivotal piece of evidence. The lack of reliability pertaining to eyewitness identifications is blatant, yet innocent people continue to be convicted because inherently suggestive procedures continue to be admissible as evidence (Wells & Quinlivan, 2009). In order to achieve due process change is compulsory. By continuing to modify the identification procedures utilized by the police (Clark & Godfrey, 2009; Kebbell, 2000; Lindsay et al., 1989), and further researching the role of eyewitness expert testimony, the number of wrongful convictions will decrease and justice will be achieved.


Borchard, E. M. (1932). Convicting the innocent. New Haven, CT: Yale University Press.

Clark, S. E., & Godfrey, R.D. (2009). Eyewitness identification evidence and innocence risk. Psychodynamic

Bulletin & Review, 16, 22-42. doi:10.3758/PRB.16.1.22

Cutler, B. L., & Kovera, M. B. (2011). Expert psychological testimony. Current Directions in Psychological

Science, 20, (1), 53-57. doi: 10.1177/0963721410388802

Cutler, B. L., Penrod, S., & Dexter, H. R. (1989). The eyewitness, the expert psychologist, and the jury. Law and Human Behavior, 13, 311-332.

Innocence Project. Accessed June 27, 2013, from

Kebbell, M. R. (2000). The law concerning the conduct of lineups in England and Wales: How does is satisfy the recommendations of the American Psychology-Law Society? Law and Human Behaviour, 24, 309-315. doi: 01477307/00/0600/0309$18.00/1

Lindsay, R. C. L., Wells, G. L., and O'Connor, F. J. (1989). Mock-juror belief of accurate and in accurate eyewitnesses: A replication and extension. Law and Human Behavior, 13, 333-339 doi:0147-7307/89/0900-0333$06.00/0

Mason v. Braithwaite (1977). 432 U.S. 98. Neil v. Biggers (1972). 409 U.S. 188.

Pezdek, E. (2006). Expert Testimony on Eyewitness Memory and Identification. In Expert Psychological

Testimony in the Courts (M. Costanzo, D. Krauss, & K. Pezdek, Eds.) (pp. 99-118). Mahwah, NJ: Erlbaum. Wells, G. L. (1978). Applied eyewitness-testimony research: System variables and estimator variables. Journal of Personality and Social Psychology, 36, 1546-1557.

Wells, G. L., & Quinlivan, D. S. (2009). Suggestive eyewitness identification procedures and the Supreme Court's reliability test in light of eyewitness silence: 30 years later. Law and Human Behavior, 33, 1-24. doi:10.1007/s10979-008-9130-3

Wells, G. L., Small, M., Penrod, S., Malpass, R. S., Fulero, S. M., & Brimacombe, C. A.E. (1998). Eyewitness identification procedures: Recommendations for lineups and photospreads. Law and Human Behavior, 22, 603-647. Doi: 0147-7307/98/1200-0603$15.00/1