Innocence projects identify people who, despite being found guilty in court and having their convictions affirmed on appeal, may be actually innocent of the crimes for which they were convicted. Innocence projects try to exonerate these actually innocent people. Innocence projects have also examined the trials of people who have been exonerated to understand how the conviction of actually innocent people resulted. Armed with that knowledge, innocence projects seek to prevent further wrongful convictions through education about the factors which contribute to the conviction of actually innocent persons and through proposals to improve systemic safeguards against wrongful conviction.

This article attempts to open—or hopefully expand—a dialogue between innocence projects and the forensic science community; an important, though often neglected, participant in the criminal justice system. Two members of the Board of Directors of the Innocence Project of Minnesota, a lawyer and a forensic anthropologist, will provide an overview of the background of innocence projects, general factors identified as contributing most significantly to the conviction of innocent persons, and the initiatives for systemic change—focusing on the role science played in wrongful conviction. It is hoped that through this article, forensic scientists and others will be encouraged to become involved in the effort to prevent wrongful convictions by contributing their ideas to the discussion and assisting in developing and implementing the necessary changes.

Almost everyone would agree that it is wrong to put a person to death or in prison for something they actually didn’t do, something for which they are “actually innocent.” That has been an accepted tenet of our criminal justice system since time immemorial, exemplified by the familiar adage “it is better to let ten guilty men go free than to convict one innocent man.” Early in the twentieth century, however, scholars and authors began to question whether innocent persons were being convicted in American courts and published a number of landmark books arguing that they were: Convicting the Innocent (1932), edited by Yale Law School Professor Edwin M. Borchard; The Court of Last Resort (1952), by Erle Stanley Gardner; The Death Penalty in America (1964), by Hugo Adam Bedau; and, In Spite of Innocence: Erroneous Convictions in Capital Cases (1992), by Michael L. Radelet, Hugo Adam Bedau, and Constance E. Putnam.

Despite such scholarly studies, there were few cases in which everyone agreed that the person convicted was actually innocent of the crime. Traditional investigation techniques had occasionally resulted in exoneration of a wrongfully convicted person, but still left questions about whether the person’s innocence was really established. Cases of agreement, where the convicted person had been in a prison or a jail when the murder was committed, or where the “murder victim” turned up alive after conviction and sentencing, were written off by some as isolated mistakes that were bound to happen occasionally. On the broader question, there remained sharp disagreement and little definitive evidence to convince skeptics that actually innocent people had been convicted in courts in the United States on a larger scale.

In the 1980s, science delivered the prosecution a powerful new tool, DNA testing, which could link a specific person to a crime by analyzing bodily materials or fluids from a victim or a suspect. By the late 1980s, DNA testing began to be used by defense attorneys as a means of proving that a person convicted of a rape or murder did not commit that crime. This post-conviction use of DNA analysis altered the wrongful conviction debate forever by providing proof positive that actually innocent people had been convicted in American criminal courts. In the years 1989, 1990, and 1991, there were a total of five exonerations.

In the early 1990s, Barry Scheck and Peter Neufeld, two lawyers already familiar with DNA analysis, believed that there were many actually innocent persons who had been wrongly convicted and realized that a coordinated effort, an “innocence project,” was necessary to address the depth and national scope of the problem they perceived. They founded the first “innocence project” in conjunction with Cardozo Law School in New York, and began reviewing cases and assisting inmates from across the country who claimed they were actually innocent. The story of the “Innocence Project,” and the stories of some of the innocent people whose lives had been tragically affected by wrongful conviction, are recounted in Actual Innocence: Five Days to Execution and Other Dispatches from the Wrongly Convicted (2000) by Jim Dwyer, Peter Neufeld, and Barry Scheck and in an updated edition, entitled When Justice Goes Wrong and How To Make It Right (2003). As of September 17, 2005, the Innocence Project has documented that 162 actually innocent people, convicted and sentenced to death or long prison terms for crimes they did not commit, had been exonerated since 1989 (

Factors Responsible for Wrongful Conviction
Through a “post mortem” of individual cases in the first 74 exonerations, Dwyer, Neufeld, and Scheck (2000) identified many factors that contributed to the wrongful convictions of actually innocent persons (Table 1). Mistaken eyewitness identifications, police misconduct, prosecutorial misconduct, bad lawyering by defense counsel, false testimony of jailhouse snitches and informants, false confessions, and bad science are some of the factors that have been identified as central to wrongful conviction. Because the same techniques and types of evidence that had resulted in the conviction of the innocent people were being used to obtain convictions every day in courtrooms across the country, the implications of their findings were clear: actually innocent people were beingwrongfully convicted on a large scale.

With this knowledge, the Innocence Project, and other similar organizations like the Innocence Project of Minnesota, has been pursuing initiatives to bring about systemic changes to the criminal justice process to decrease the chance of wrongful conviction of an actually innocent person. Two non-forensic initiatives undertaken to change criminal justice processes, which are already in place in some jurisdictions, involve police interrogations and eyewitness identification procedures.

Despite the difficulty in understanding why an innocent person would confess to something they didn’t do, especially something as serious as rape or murder, the post mortem showed it happened in 22% (Scheck et al. 2000; 26.92% in the first 130 exonerations in Scheck et al. 2003) of exoneration cases. An innocence project initiative to require recording police interrogations and “confessions” is being pursued in an effort to decrease false confessions.

Eyewitness identification, often thought of as the gold standard of a criminal prosecution requiring proof beyond a reasonable doubt, played a role in convicting the innocent in 81% of the first 72 exonerations (Scheck et al. 2000; 77.69% in the first 130 exonerations in Scheck et al. 2003), with five cases involving three or more mistaken eyewitnesses! (For more information on psychological studies of eyewitness identification, see, e.g., Gary Wells, Ph.D., Iowa State University, Wells and his associates have established that an innocence project initiative to make relatively minor changes to the common police procedure of displaying photographs to witnesses—to use a double blind, sequential display process and to control the information given to the witness—could dramatically reduce the number of mistaken identifications, without a reduction in accurate identifications.

table 1
Table 1: Factors Leading to Wrongful Convictions in 74 U.S. Cases. (Schecket al. 2000)

table 2
Table 2: Forensic Science Factors Contributing to Wrongful Convictions in 25 of the first 82 cases of exoneration (modified from; September 2005)

Wrongful Conviction and Forensic Science
Documenting and understanding the role scientific evidence and forensic scientists have played in cases of wrongful conviction is currently ongoing. To date, a systematic review by individuals with scientific expertise of all cases of wrongful conviction and exoneration to clarify and identify the forensic science-based factors has not been done. General overviews have been published, however, that provide insight into the role forensic science evidence and testimony have played in smaller subsets of cases of wrongful conviction (Conners et al. 1996; Gross et al. 2005;, and Saks and Koehler 2005). Conners et al. provide the most detail when summarizing common attributes of evidence presented during and after trial for 28 cases. Several case examples are presented when discussing the use of forensic evidence and alleged misconduct and malfeasance by the government (1996:15–18). Behaviors significant to the wrongful conviction include perjured testimony (i.e., about qualifications), exaggerating results, misrepresenting results as conclusive when they were inconclusive, changing laboratory records, and manufacturing evidence. To be sure, the focus of these studies is not the role of forensic science in wrongful conviction. Criteria used to classify types of forensic science error or misconduct are not defined in most of these studies, however, and it is clear that a more systematic and thorough review to better understand and ultimately remedy undeniable problems inforensic science as it is practiced in the United States is needed.

Personnel from the Innocence Project (based in New York) identified 25 cases in which forensic science and scientists played a contributing role in the wrongful conviction of the first 82 exonerees. Frequency data for each of the forensic science-based factors identified and categorized for the 25 cases (of the first 82 exonerations) are presented in Table 2. Key forensic science factors in wrongful conviction include, in order of frequency of occurrence, Misinterpretation (11/25 cases, 44%), Statistical Exaggeration (10/25 cases, 40%), Suppression of Evidence and/or Exculpatory Results (7/ 25 cases, 28%), Falsified Results (7/25 cases, 28%), Falsified Credentials (5/25 cases, 20 %), Contamination (5/25 cases, 20 %), Testified to Tests Never Conducted (4/25 cases, 16 %), and Other (2/25 cases, 8 %) (

Misinterpretation and Statistical Exaggeration are the two most significant forensic science-based factors contributing to wrongful conviction in the 25 cases and are both frequently present in identified cases. Examples of misinterpretation include analyst misinterpretation of test results, misrepresentation or lack of understanding of the limitations and scope of a particular forensic science discipline, and confusion over the validity of the result, i.e., indicating contamination or accuracy. Statistical Exaggeration includes and reflects a lack of understanding of statistics, probability theory, and calculation of error rates. A review of the case summaries of exonerees featured on the Innocence Project, New York website (, identifies several cases that illustrate misinterpretation of results and presentation of inaccurate statistics.

Josiah Sutton was convicted in 1999 for a rape that occurred in Texas. Among other errors, an analyst from the Houston Police Department Crime Laboratory testified that Sutton’s DNA was an exact match with the DNA found on the victim and that only one person in 694,000 could have been the donor of the semen. Re-examination of the evidence concluded that Sutton could not have deposited the sperm and, statistically, 1 in 16 black men share the reported DNA profile. Mr. Sutton, a teenager at the time of his conviction and incarceration, was released in 2004 after serving 4.5 years in prison. An investigation of the Houston crime lab identified serious problems, including insufficiently trained and poorly educated staff, and laboratory conditions that resulted in contamination of evidence.

Suppression of Evidence and/or Exculpatory Results occurred in 7 of the 25 cases and refers to instances where examiners failed to report or make available to legal counsel, test results that are inconclusive or exclude the defendant. The re-investigation of the Gilbert Alejandro case that led to his release in 1999, illustrates this factor. Alejandro was convicted in 1990 of aggravated sexual assault in Ulvade County, Texas. During the trial, the analyst who conducted the DNA testing testified that Alejandro, and no one else, was identified as the source of the semen on the victim’s clothing. Upon review, a DNA analyst from Bexar County reported that the test the original analyst, the notorious Fred Zain, testified about was actually inconclusive and that two previous DNA tests had yielded results that excluded Alejandro. It was reported that Zain knew of these results and failed to inform the defense that Mr. Alejandro had been excluded as the donor.

There are numerous other examples that clearly illustrate the significant and devastating role forensic science and forensic scientists have played in wrongfully convicting innocent people (see Connors et al. 1996, The question becomes, given what we know about what went wrong in cases of wrongful conviction and recognizing that there is more to learn, how do we move forward? How do we as members of the forensic sciences community remedy these errors and reduce/prevent wrongful conviction from occurring in the future? Within the forensic science community, areas highlighted for reform are identified on three levels: the individual scientist, the crime laboratory, and the various forensic science disciplines themselves.

Recommended Remedies
On the individual level, attention must be directed to how students are prepared for a career in the forensic sciences. What should the educational and experiential requirements be for future forensic scientists? How many courses in the natural sciences, the social sciences, statistics, and law, for example should be required? In response to a National Institute of Justice sponsored assessment of current needs and conditions of forensic science in the United States, a Technical Working Group on Forensic Science Education (TWGED) was formed and presented recommended standards for undergraduate and graduate forensic science education programs, including curriculum, faculty, graduation requirements, as well as a mechanism by which existing and new programs could be evaluated. The Forensic Science Education Program Accreditation Commission (FEPAC) was formed to evaluate the quality of forensic science education programs and develop criteria and procedures for accreditation; FEPAC is into their third year as an accrediting commission. Accreditation cannot ensure that individuals will not make mistakes in the future but it goes a long way to building consensus about what is the best course of study to prepare for a career in the forensic sciences. Also important, on the individual scientist level, is the role of board certification in development and maintenance of professional credentials. Most forensic science disciplines offer the opportunity for certification. The certification process for each, however, is different. As a community we must consider what the significance of board certification is. Should it be a pre-requisite to expert testimony, the end of a training period? Again, achieving board certification does not ensure that mistakes won’t be made in the future, but it does require the applicant to be current inthe literature and methods and techniques practiced.

On the agency level, Scheck et al. (2003:351-362) present several recommendations to reduce the frequency of wrongful convictions. Many of the recommendations center on reducing misconduct by the agency and individual examiners. Required laboratory accreditation is high on the list of recommendations. Achieving and maintaining accreditation addresses such issues as the existence of standard operating procedures, protocols, and laboratory procedures to ensure high quality performance of instrumentation and personnel and accuracy of results and conclusions through proficiency testing (including blind), validation studies, hiring practices, and other quality control and quality assurance processes. Additionally, it is recommended that crime laboratories be subjected to regulatory oversight via independent panels that are empowered with actual regulatory authority.

At the discipline level, forensic scientists must work toward establishing the scientific validity of each discipline and the techniques utilized under the standards defined by Supreme Court rulings to reduce the admissibility of junk or fraudulent science. Recent challenges to the validity of, for example, fingerprint examination, document examination, and other forensic sciences have encouraged and resulted in a flurry of research directed at assessing the evidentiary potential of various categories of physical evidence, determining error rates, and evaluating reliability. Such research can only strengthen and enhance the role of science in the courtroom. Technical working groups (TWGs) are engaged in the development of professional standards and guidelines to ensure the best practice of various disciplines in the areas of analysis, reporting, and expert testimony.

National and Local Reform Initiatives
In addition to the initiatives mentioned above, the Innocence Project formerly associated with Cardozo Law School, is spearheading and pushing several initiatives they would like to see occur on a national level, including state-by-state formation of Innocence Commissions (IC), Audit Oversight Committees (AOC), and the severance of crime labs from law enforcement agencies. Innocence Commissions would be comprised of stakeholders from many different agencies and organizations including defense attorneys, prosecutors, victim advocates, judges, law enforcement personnel, forensic scientists, and legislators. The ICs could evaluate wrongful convictions and assist in determining what went wrong and propose ways of improving the practice of forensic science. The formation of AOCs is explored in a position paper written by Peter Neufeld and available for viewing on These committees would ideally function to investigate individual scientists that are implicated or accused of serious misconduct. More information on ICs and AOCs is availableat

On the local level, the Innocence Project of Minnesota has undertaken a number of initiatives in order to address the role of forensic sciences in wrongful conviction. First, as an organization, we feel very strongly that a diverse board of directors and advisory board comprised of professionals in the areas of criminal defense, criminal prosecution, forensic science, and law enforcement will facilitate a more holistic multidisciplinary approach to understanding the factors involved in wrongful conviction and developing remedies to decrease the occurrence of wrongful conviction. Our organization currently consists of forensic scientists from crime labs (state, city, county), medical examiner/coroner’s offices, private consultants, law enforcement personnel, educators, and attorneys. Together we have organized educational workshops and seminars designed to bring law enforcement, forensic science, and attorneys together to discuss issues of wrongful conviction, disseminate information about various forensic sciences, and discuss and begin to implement possible remedies. We have also entered into discussions with representatives of local law schools in regard to teaching courses about forensic science and eventually making such courses a permanent component of the law school curriculum.
In conclusion, in-depth, detailed analyses of known exonerations, both those accomplished through DNA analysis and by other means, are needed to better understand the role of forensic science in wrongful conviction and to formulate strategies to develop and initiate reform in order to decrease, and perhaps one day, eliminate our role in wrongful conviction.

• Bedau, Hugo Adam (editor). The Death Penalty in America: An Anthology. Chicago:Aldine Publishing Company, 1964.
• Borchard, Edwin M. Convicting the Innocent: Sixty-Five Actual Errors of Criminal Justice. Garden City, New York: Garden City Publishing Company, 1932.
• Conners, Edward, Thomas Lundregan, Neal Miller, and Tom McEwen. Convicted by Juries, Exonerated by Science: Case Studies in the Use of DNA Evidence to Establish Innocence After Trial. Research Report, National Institute of Justice, 1996.
• Gardner, Erle Stanley. The Court of Last Resort. New York: William Sloane Associates, 1952.
• Gross, Samuel R., Kristen Jacoby, Daniel J. Matheson, Nicholas Montgomery, and Sujata Patil. Exonerations in the United States, 1989 Through 2003. The Journal of Criminal Law and Criminology 95(2):523-560, 2005.
• Radelet, Michael L., Hugo Adam Bedau, and Constance E. Putnam. In Spite of Innocence: Erroneous Convictions in Capital Cases. Boston: Northeastern University Press, 1992.
• Saks, Michael J. and Jonathan J. Koehler. “The Coming Paradigm Shift in Forensic Identification Science.” Science 309:892-895, 2005.
• Scheck, Barry, Peter Neufeld, and Jim Dwyer, 1st Edition. Actual Innocence and Other Dispatches from the Wrongly Convicted. New York: Doubleday, 2000.
• Scheck, Barry, Peter Neufeld, and Jim Dwyer, 3rd Edition. Actual Innocence: When Justice Goes Wrong and How to make It Right. New York: New American Library, 2003.

We would like to thank everybody at the Innocence Project in New York, especially Maddy deLone, Peter Neufeld, Huy D. Dao, and Sarah Tofte for graciously providingus with unpublished information in such a timely fashion.

Michael Cromett, J.D., is Assistant State Public Defender with the Minnesota State Public Defender’s Office in Minneapolis, MN. Michael can be reachedat

Susan M. Thurston Myster, Ph.D. is a forensic anthropologist and Associate Professor and Director of the Forensic Sciences Program at Hamline University in St. Paul, MN. Susan can be reached at