Flawed forensic science — sometimes called junk science — has contributed to wrongful convictions across the United States for decades. Bite mark analysis, microscopic hair comparison, bullet lead analysis, and certain forms of soil comparison testimony have all been implicated in cases where people were convicted on evidence that lacked scientific validity. A landmark 2009 National Academy of Sciences report found that many forensic disciplines used routinely in criminal trials had never been rigorously tested, lacked standardized protocols, and overstated the certainty of their conclusions to juries. The problem is not confined to the past. Cases built on discredited techniques remain in the system, and the legal mechanisms designed to filter out unreliable evidence have repeatedly failed to do so.
Bite Mark Analysis: A Technique Without Scientific Foundation
Bite mark analysis rests on two core assumptions: that each person’s dental pattern is unique, and that human skin can reliably preserve and record those patterns for forensic comparison. The National Institute of Standards and Technology found no scientific data supporting either assumption. Human skin distorts, decomposes, and changes between the time of injury and the time of examination. Dental patterns, while variable, have not been shown through population studies to be unique in the way fingerprints have. And the matching process itself is inherently subjective — different examiners reviewing the same bite mark have reached different and contradictory conclusions.
Despite this, bite mark evidence has been admitted in criminal trials for decades and has contributed to wrongful convictions. The technique gained courtroom legitimacy in part through the 1979 Florida conviction of Ted Bundy, in which bite mark evidence played a role. The Bundy case made the methodology appear validated when it had not been scientifically tested — and because Bundy was guilty by overwhelming independent evidence, the technique’s role in his conviction was never interrogated. Subsequent cases applied the same methodology to defendants whose guilt was far less independently established, with more consequential results. Supreme Court Justice Sotomayor has publicly acknowledged the problems with bite mark evidence, a recognition that reflects the growing scientific and legal consensus that the technique should not be admitted in criminal proceedings.
Microscopic Hair Comparison: What the FBI Admitted
Microscopic hair comparison involves visually examining hair samples under a microscope and comparing characteristics between a sample from a crime scene and a sample from a suspect. The technique was presented in courts for decades as a scientifically reliable method of identification. It is not.
Following a review conducted in cooperation with the Innocence Project and the National Association of Criminal Defense Lawyers, the FBI acknowledged that in over 90% of the cases reviewed, examiners had provided testimony containing erroneous statements about hair evidence. The errors were not minor. They included claiming that hair evidence could identify a specific individual — a claim the methodology cannot support — and asserting statistical significance for matches that had no validated statistical basis. Prosecutors and defense attorneys in hundreds of affected cases were notified. The scale of the problem, once formally acknowledged, was significant enough to require a systematic remediation effort affecting capital and non-capital cases across multiple decades of prosecutions.
A consistent feature of discredited forensic testimony is the gap between what the underlying methodology can actually support and what examiners have said in court. Bite mark analysts have testified that a bite mark was “made by this defendant to the exclusion of all others.” Hair analysts have testified to matches with a certainty that has no statistical basis. This overstating is not incidental to how these disciplines have operated — it is a structural feature of how forensic evidence has been packaged for jury consumption. Juries hear “scientific expert” and reasonably expect the certainty that science implies. The problem is that the certainty being expressed has not been earned by the underlying methodology.
Firearms and Toolmark Analysis: Contested Validity
Firearms analysis attempts to match bullets or cartridge cases to specific weapons by comparing the toolmarks left during firing — marks that examiners argue are unique to individual firearms. The technique has been used in criminal cases for decades and is still widely admitted. Its scientific foundation is contested.
Studies published in peer-reviewed literature have documented high rates of inconclusive results and significant examiner error rates in controlled proficiency tests. The foundational assumption — that each firearm leaves marks that are both unique and reproducible enough to support reliable identification — has not been validated through the kind of population-level empirical research that other identification disciplines require. Scientific American and researchers at Johns Hopkins have examined the evidentiary basis of firearms forensics and found it wanting relative to the certainty with which conclusions are presented in court. The discipline is not without value, but the gap between what the methodology supports and what examiners have testified to in criminal proceedings is significant.
Soil Evidence and the Dennis Salerno Case
The Dennis Salerno case is included here because it illustrates a documented pattern of soil evidence overstatement in Michigan courtrooms. The central claim about the soil evidence — that subsequent investigation challenged the uniqueness of the soil attributed to the crime scene location — is drawn from prior reporting on the case. The expert witness, Michigan State University Professor Thomas Vogel, is a named and verifiable figure whose academic status and MSU misconduct finding are matters of public record. Claims about the soil evidence itself are characterized as contested rather than established, consistent with source discipline standards.
Soil can appear to be the ideal silent witness. It can be location-specific, it is physically recoverable from shoes, vehicles, and clothing, and forensic geologists have testified that chemical and mineral analysis can establish that soil from a suspect came from a specific location. That testimony, however, rests on a methodology that the 2009 National Academy of Sciences report specifically identified as lacking rigorous validation and standardization.
Unlike DNA analysis — which has defined protocols, validated error rates, and peer-reviewed population databases — soil comparison has no universally accepted standard for how comparisons should be conducted or how conclusions should be qualified. One expert might emphasize particle size and mineral composition; another might focus on color or organic content. The absence of standardized methodology opens the door to confirmation bias and to the overstating of conclusions that characterizes other discredited forensic disciplines.
Dennis Salerno
The Salerno case in Michigan provides a documented example of how soil evidence can be presented in ways that exceed what the underlying science supports. Michigan State University Professor Thomas Vogel, a geologist, testified that soil found on Salerno’s shoes was consistent with and unique to the alleged crime scene location. Subsequent investigation challenged the claimed uniqueness of that soil evidence, raising questions about the foundation of the expert’s conclusions and the degree of certainty with which they were presented to the jury.
That evidentiary dispute is compounded by a separate matter of documented record: Vogel was stripped of his MSU emeritus title in 2018 following a finding that he had violated the university’s sexual misconduct policies, as reported by The State News. The misconduct finding is a separate matter from the soil testimony. Both are part of the documented record relevant to assessing the credibility and institutional accountability of a forensic expert whose testimony contributed to a criminal conviction.
The National Academy of Sciences 2009 report on forensic science in the United States was among the most significant institutional assessments of the discipline ever conducted. It found that with the exception of nuclear DNA analysis, no forensic discipline had been rigorously shown to have the reliability that courtroom presentation implies. It called for standardization, independent oversight, mandatory proficiency testing, and significant investment in validating the scientific basis of techniques already in widespread use. Many of its recommendations have not been implemented. The gap between what that report identified and what has changed in the years since represents an ongoing institutional failure with ongoing human consequences.
What Reform Actually Requires
Forensic methods should be required to demonstrate scientific validity — through peer-reviewed research, known error rates, and population-level empirical testing — before being admitted in criminal proceedings. The current system allows techniques to gain courtroom legitimacy through repetition and precedent rather than through scientific validation. Gatekeeping standards like Daubert were designed to prevent this but have in practice failed to exclude techniques that lack foundational scientific support. Legislative requirements for pre-admission validation, with independent scientific review rather than judicial assessment alone, would address the root mechanism through which junk science enters the courtroom.
Forensic laboratories and the disciplines they practice require standardized protocols developed by independent scientific bodies rather than by law enforcement agencies with an institutional interest in maximizing the usefulness of forensic evidence. Independent accreditation, mandatory proficiency testing with disclosed results, and external audit of laboratory practices are prerequisites for a forensic science system that can credibly claim scientific validity. The 2009 NAS report called for a National Institute of Forensic Science to provide this function. That recommendation has not been implemented at federal scale.
Courts should require that forensic evidence meet established scientific standards at admission — not merely that a qualified expert endorses the technique. And for cases where convictions were obtained in part on evidence subsequently discredited by scientific review, a systematic post-conviction review mechanism is needed. The FBI’s hair analysis notification process demonstrated that such review is possible at scale. It should not require a specific FBI-initiated review to trigger it: discrediting of a forensic technique by a recognized scientific body should automatically initiate review of convictions in which that technique was used.
Prosecutors, defense attorneys, and judges who evaluate forensic evidence need systematic education in the scientific principles and documented limitations of the disciplines they encounter in court. The current system places lawyers and judges in the position of assessing scientific validity without scientific training, and relies on adversarial cross-examination to surface problems that require scientific expertise to identify. Continuing education requirements for legal professionals that include forensic science methodology and its documented failures would reduce the degree to which junk science can be presented to juries without effective challenge.
The cases built on discredited forensic techniques do not disappear when the science is revised. They remain in the system — in prisons, in appellate records, in the lives of people who were convicted on evidence that cannot support the conclusions drawn from it. Addressing the flaws in forensic science is not a retrospective concern. It is an ongoing obligation to the integrity of a system that claims to convict people on proof, not on the appearance of proof.
Sources
Rita Williams, Flawed Forensics: The Dangerous Legacy of Junk Science in Criminal Justice, Clutch Justice (Apr. 23, 2025), https://clutchjustice.com/2025/04/23/flawed-forensics-how-bite-marks-hair-bullet-analysis-and-soil-have-failed-justice-and-what-we-can-do-about-it/.
Williams, R. (2025, April 23). Flawed forensics: The dangerous legacy of junk science in criminal justice. Clutch Justice. https://clutchjustice.com/2025/04/23/flawed-forensics-how-bite-marks-hair-bullet-analysis-and-soil-have-failed-justice-and-what-we-can-do-about-it/
Williams, Rita. “Flawed Forensics: The Dangerous Legacy of Junk Science in Criminal Justice.” Clutch Justice, 23 Apr. 2025, clutchjustice.com/2025/04/23/flawed-forensics-how-bite-marks-hair-bullet-analysis-and-soil-have-failed-justice-and-what-we-can-do-about-it/.
Williams, Rita. “Flawed Forensics: The Dangerous Legacy of Junk Science in Criminal Justice.” Clutch Justice, April 23, 2025. https://clutchjustice.com/2025/04/23/flawed-forensics-how-bite-marks-hair-bullet-analysis-and-soil-have-failed-justice-and-what-we-can-do-about-it/.
