Karen Kafadar is Commonwealth Professor and chair of the Department of Statistics at the University of Virginia and a member of the Forensic Science Standards Board. Anne-Marie Mazza is the director of the Committee on Science, Technology and Law of the National Academy of Sciences. The authors contributed this article to Live Science's Expert Voices: Op-Ed & Insights.
Historically, forensic science has had a huge impact on identifying and confirming suspects in the courtroom, and on the judicial system more generally. And yet, a 2009 report from the U.S. National Academy of Sciences (NAS) identified numerous shortcomings in the field, including an absence of a scientific basis for most forms of forensic evidence, a lack of uniform standards and the need for independence from law enforcement. In short, the report called for nothing less than major reform.
Forensic scientists have been an integral part of the judicial process for more than a century. The most well-known and widely used forensic evidence involves fingerprints left at a crime scene, which Edmond Locard and Francis Galton in the 19th century asserted as "unique" and reliably capable of identifying a single individual ("The History of Statistics: The Measurement of Uncertainty Before 1900 by Stephen M. Stigler," Harvard University Press, 1986). Other types of material then followed, such as other forms of pattern evidence (e.g., shoe prints, tool marks, tire tracks, bite marks and handwriting analysis) and chemical signatures, such as compositional analysis of bullet lead (CABL) and the presence or absence of 13 specific alleles found in human DNA.
Of those methods, only DNA analysis arose not out of a forensic need but from academic research (in biology); not until later was it used as a method of forensic identification. The technique's ability to establish a numerical probability for comparing two single-source DNA signatures (one from biological evidence found at the crime scene and one from a suspect) and the formal documentation of objective laboratory procedures for conducting DNA analysis (e.g., 13 specific alleles known to have high sensitivity and specificity for identification), contributed to its reputation as a "gold standard" for criminal investigations.
Where forensic science is failing
Unfortunately, most forms of forensic evidence other than DNA have lacked similar scientific foundations. Instead, they have been characterized by much subjectivity, human observer bias, error and variability in processing and interpreting the evidence, lack of standardized procedures and accreditation programs in crime laboratories, inconsistent validation and unknown error rates, and, most worrisome, little incentive for conducting research into better, more reliable methods. [How Digital Forensics Can Help Reveal Online Fraud ]
Consequently, over the years "experts" have provided unjustified testimony in the courtroom. For example, for decades, a "bullet lead expert" could testify that bullets found at a crime scene were "consistent with their having come from the same box of ammunition" as that found in a suspect's possession, even though such a statement could not be statistically justified. (The FBI discontinued Compositional Analysis of Bullet Lead in September 2005.)
The 2009 report from the National Academy of Sciences, "Strengthening Forensic Science in the United States: A Path Forward," documented similar shortcomings associated with other forms of non-DNA forensic evidence, as well as the need for increased training and education among forensic practitioners and judicial personnel. Five years later, the National Institute of Standards and Technology (NIST), jointly with the U.S. Department of Justice, responded to this call for reform by establishing the National Commission on Forensic Science (NCFS) and by coordinating leading forensic practitioners, researchers, law enforcement officers and judicial personnel through NIST's Organization of Scientific Area Committees (OSAC).
But why is reform needed, and what can we expect reform efforts to accomplish?
Protecting the innocent
It is important to recognize that errors arising from overreaching interpretations of forensic evidence can have serious consequences. If forensic evidence or eyewitness identification incorrectly implicates an innocent person in a crime, the innocent person is unfairly (and immorally) convicted, and possibly incarcerated for years. DNA evidence has revealed erroneous findings from other types of forensic evidence and led to the reversal of more than 300 cases, freeing innocent individuals who were unfairly found guilty and spent decades in prison, says the Innocence Project. Worse, with false evidence, the guilty person is free to conduct additional crimes; in roughly half of the overturned cases, the new evidence led to the conviction of the person who actually committed the crime. [Innocent Suspects Confess Under Pressure ]
Mistaken eyewitnesses were present in more than 70 percent of these false accusations. For examples of mistaken eyewitness identification, see "Convicting the Innocent: Where Criminal Prosecutions Go Wrong" (Harvard University Press, 2011). For an assessment of eyewitness identification research, see the National Research Council report titled "Identifying the Culprit: Assessing of Eyewitness Identification" (National Academies Press, 2014).
When those falsely convicted individuals are exonerated by DNA evidence, news reports of the exonerations lead the general public to question the validity of procedures used by the criminal justice system, overshadowing the system's many other positive impacts. The 2009 report also emphasized that results of tests on forensic evidence (e.g., fingerprint comparisons, source of bite marks) need to be clear and transparent, and independent of law enforcement, to limit the effects of unintentional bias.
A path toward better forensics
The NCFS and OSAC are beginning to address the underlying causes of such errors by recommending more scientific research into the validity and reliability of specific forms of forensic evidence, such as:
- Quantifying the uncertainty in forensic analyses
- Investigating effects of human bias, variability and errors in forensic examinations, and the establishment of well-validated procedures for reducing such effects in practice
- Increasing transparency and standardization of operations in crime laboratories
- Developing certification standards for forensic practitioners and accreditation programs for crime laboratories
- Developing standard terminology for reporting forensic evidence and presenting testimony
- Expanding training and oversight of forensics programs
- And most importantly, conducting scientific research on validation, reliability and reproducibility of forensic procedures
These efforts need well-documented and validated research that has been subjected to rigorous review by the scientific community. To that end, a statement from the NCFS Scientific Inquiry and Research Subcommittee presents criteria for assessing scientific literature for its consistency with scientific principles.
Without such reform, what will happen? As Judge H.T. Edwards of the U.S. Court of Appeals for the District of Columbia Circuit noted in his comments at NCFS's inaugural meeting on Feb. 3, 2014:
"Absent meaningful action by scientists and forensic analysts, the courts will continue to admit forensic evidence in criminal trials, without regard to its scientific validity and reliability. Why? Because precedent supports this practice. Yes, there have been a few trial court decisions that have limited the admission of some forensic evidence, but, to date, there has not been a single federal court of appeals decision that has curbed its admissibility."
Until such reform is implemented, errors will occur, innocent victims will be falsely accused while the true perpetrators continue to commit crimes, and the public ultimately will lose confidence in the criminal justice system.
Historically, new problems have stimulated scientific research. The proposed reforms to the forensic science field, which the legal and scientific communities are addressing together, can have lasting benefits, not only for increased accuracy in criminal investigations but also for advancing science.
The opinions expressed in this article are those of the authors and do not necessarily reflect the views of the Forensic Science Standards Board; the Committee on Science, Technology and Law; or the National Academy of Sciences. Follow all of the Expert Voices issues and debates — and become part of the discussion — on Facebook, Twitter and Google+. The views expressed are those of the author and do not necessarily reflect the views of the publisher. This version of the article was originally published on Live Science.