Eyewitness Identification Field Studies

Project Purpose/Goal

Mistaken eyewitness identification is the leading cause of wrongful conviction of innocent people in the United States. The significant role that mistaken eyewitness identifications (EWID) have played in convictions of the innocent has led to a strong interest in finding ways to reduce eyewitness identification errors.


Approach, Results and Implications

In 2008, the American Judicature Society’s Center for Forensic Science and Public Policy, collaborated with the Innocence Project and the Police Foundation to examine eyewitness identification procedures in the field, namely the reliability of simultaneous versus sequential lineups administered under double-blind conditions using laptop computers.

The primary variable tested in the EWID Field Studies was sequential versus simultaneous presentation of photo lineups under double-blind administration (in which the lineup administrator is not aware of the identity of the suspect). A specially-designed proprietary software program was used to enable scientists to examine the primary research question, improve the accuracy of data collection, and allow witnesses to self-administer the photo lineup using laptop computer.

During the first phase of the EWID Field Studies, administered by the American Judicature Society, scientists gathered and analyzed data from approximately 1,200 actual eyewitness identifications. The analysis of data collected across four sites–the Austin (TX) Police Department, the Charlotte-Mecklenburg (NC) Police Department, the Tucson (AZ) Police Department, and the San Diego (CA) Police Department–has been completed and a report was issued in September 2011.

The Police Foundation led the second phase, which found that there was no significant main effect for the type of lineup (sequential or simultaneous). Subsequent evaluation of the data from Austin, particularly the evidentiary strength ratings from police, prosecutors, defense attorneys, and judges demonstrated a small advantage for simultaneous lineups (see Amendola & Wixted, 2015a; Amendola & Wixted, 2015b; and Amendola & Wixted, 2015c). Other recent research has also suggested the possibility of a simultaneous advantage, and some policies have been changed in response to the 2014 National Academy of Sciences Report, entitled “Identifying the Culprit:  Assessing Eyewitness Identification.” That report did not favor one alternative over the other, but suggested that the previous method designed to assess accuracy (“diagnosticity ratios”) did not include consideration of another important variable, the response bias or degree of evidence that the observer finds acceptable to make an identification. As such, former assertions that the sequential method led to greater accuracy could not be supported due to the exclusion of “response bias” considerations.

Evidentiary strength ratings were captured using the Strength of Evidence Scale, see:

Amendola, K. L. and Slipka, M. G. (2009). Strength of Evidence Scale.  Washington, DC:  Police Foundation.


Point of Contact

Karen Amendola, Ph.D.
Chief Behavioral Scientist
(202) 833-1460



Eyewitness, identification, wrongful conviction, ID



Amendola, K. L., & Wixted, J. T. (2017).  The Role of Site Variance in the American Judicature Society Field Study Comparing Simultaneous and Sequential Lineups. Journal of Quantitative Criminology, 1-19.  doi:  10.1007/s10940-015-9273-6. https://link.springer.com/article/10.1007/s10940-015-9273-6

Amendola, Karen L. and Wixted, John (2015).  Comparing the Diagnostic Accuracy of Suspect Identifications made by Actual Eyewitnesses from Simultaneous and Sequential Lineups in a Randomized Field Trial. Journal of Experimental Criminology, 11(2), 263-284. doi:  10.1007/s11292-014-9219-2  https://link.springer.com/article/10.1007/s11292-014-9219-2

Amendola, Karen L. and Wixted, John (2015).  No Possibility of a Selection Bias, but Direct Evidence of a Simultaneous Superiority Effect: A Reply to Wells et al.  Journal of Experimental Criminology, 11(2), 291-294.  doi:  10.1007/s11292-015-9227-x  https://link.springer.com/article/10.1007/s11292-015-9227-x