The Michigan Murders Files: Forensic Evidence Collapse Timeline
How each piece of science used in the 1970 Collins conviction was subsequently discredited, abandoned, or contaminated. Click any entry to expand the full record. Filter by evidence category.
Click any entry to expand the full record and sourcing
MSP Corporal David Leik’s basement becomes the crime scene. Officers collect hair samples without gloves. No chain of custody documentation for biological material meets modern standards.
Hair clippings vacuumed from the basement of Collins’ uncle, MSP Corporal David Leik, while the Leik family was on vacation. The prosecution would later present these clippings as the physical link between Collins and Beineman. The collection process did not meet standards that would later be required. Multiple officers handled the evidence. No controlled environment was maintained.
Source: Ann Arbor News trial archives (1970); Officer testimony confirmed at Leiterman trial (2005)
Prosecution calls a graduate chemist who testifies NAA proves the basement hair chemically matches hair on Beineman’s clothing. Judge Conlin admits it over defense objection.
Walter Holz, a graduate chemist from the Washtenaw County Health Department, testifies that the elemental composition of hair samples from the Leik basement matches hair found on Karen Sue Beineman’s clothing. The testimony implies unique identification. NAA examines elements including gold, copper, and arsenic in hair strands and compares them across samples. The technique was treated in 1970 as producing near-certain identification. Defense counsel Neil Fink objected. Judge Conlin admitted the testimony. Once Conlin’s imprimatur was on the method, the jury processed it as established science. The defense objections were on the record. They did not change what the jury heard.
Source: Ann Arbor News, 1970 trial coverage; Gregory Fournier, FOIA research documentation
Second forensic method: prosecution expert testifies that physical characteristics of hair strands under microscope are consistent with a common source. Jury hears this as confirmation of the NAA result.
Microscopic hair comparison involves examining the physical characteristics of hair strands, including cross-sectional shape, pigment distribution, medullary structure, and scale pattern, and testifying that these characteristics are consistent between two samples. In 1970, experts commonly testified that hairs “matched” or were “consistent with a common origin.” The technique had no validated statistical framework and no population database against which to evaluate frequency. The testimony implied identification without being able to establish it. Combined with the NAA result, the jury heard two independent scientific methods both pointing to Collins. The reinforcing effect of two flawed methods presented in sequence was significant.
Source: DOJ Microscopic Hair Review (2015); FBI Forensic Science Communications, archival
Jury convicts on August 19, 1970 after 27 hours of deliberation. Conlin sentences Collins to life. Original trial transcripts are subsequently purged from Washtenaw County records.
The jury deliberated for 27 hours across four days before returning a unanimous first-degree murder verdict. Judge Conlin sentenced Collins to life in prison without possibility of parole. The case was considered closed. Six other murders attributed to Collins were never prosecuted. The original trial transcripts, which would have preserved Conlin’s precise admissibility reasoning and the full expert testimony, were later purged from Washtenaw County court records. Researcher Gregory Fournier documented this through FOIA requests. What survives is approximately 800 pages of newspaper clippings and archived news summaries.
Source: Fournier, Gregory A. FOIA documentation, 2013. Goodreads author blog.
Within years of the Collins conviction, the FBI stops using NAA to identify individual hairs. The elemental composition of hair varies too widely within a single person to function as an identifier.
The core problem with NAA as applied to hair is that the elemental composition of a single person’s hair changes over time and varies across different hairs from the same person. There is no stable chemical “fingerprint.” The technique can establish whether two samples are inconsistent, but cannot establish that they came from the same source. The FBI’s own scientists determined this in the years following the widespread adoption of NAA in criminal trials during the 1960s and early 1970s. By the mid-1970s, the technique had effectively been retired for individual hair identification, though it continued to be used in other forensic contexts.
Source: FBI Forensic Science Communications, archival policy documentation; DOJ review background (2015)
NAS report “Strengthening Forensic Science in the United States” concludes that microscopic hair analysis lacks a sufficient scientific foundation and that examiners routinely overstated its significance in court.
The 2009 National Academy of Sciences report is the definitive institutional repudiation of the forensic methods that dominated American criminal courts for decades. The report examined microscopic hair comparison specifically and found that the technique had no validated statistical foundation, no population database, and no established error rate. Examiners who testified that hairs “matched” or were “consistent with a common origin” were testifying beyond what the science could support. The report called for mandatory reform across multiple forensic disciplines and for the establishment of scientific standards for methods used in criminal proceedings.
Source: National Research Council. “Strengthening Forensic Science in the United States: A Path Forward.” 2009. National Academies Press.
Joint DOJ/FBI review of 268 criminal trials finds that FBI examiners provided scientifically unsupportable hair comparison testimony in 257 of them. 33 defendants had been sentenced to death.
The 2015 Department of Justice and FBI review examined cases in which FBI hair comparison testimony had been given. Of 268 trials reviewed, examiners gave flawed testimony in 257. The flawed testimony included claims of a “match” or “consistent with a common source” without statistical grounding, claims that the probability of coincidental match was vanishingly small, and failure to acknowledge the limitations of the technique. At the time of the review, 33 of the defendants in these cases had been sentenced to death. The review covered only federal cases. State-level cases, including Collins, were not formally reviewed under this program.
Source: U.S. Department of Justice. “FBI/DOJ Microscopic Hair Comparison Analysis Review.” 2015.
Retired nurse Gary Leiterman is arrested for forging a prescription for pain medication. His DNA reference sample is collected and entered into the Michigan CODIS database. This sample will later appear on 1969 evidence.
Gary Earl Leiterman, 59, is arrested in late 2001 for forging a prescription. As a result of the arrest and subsequent conviction, Michigan law requires a DNA reference sample. His saliva sample is collected, analyzed, and entered into the state’s CODIS database. At this point, Leiterman has no known connection to Jane Mixer or to the Michigan Murders. The MSP cold case unit is conducting separate analysis of Mixer evidence in approximately the same laboratory during this period.
Source: Michigan v. Leiterman (2005); Wixted et al., Bayesian Statistical Analysis, pre-publication
33 years after Jane Mixer’s murder, the MSP cold case unit submits surviving evidence to the forensic laboratory. DNA profiles are developed from perspiration stains on a nylon stocking found at the scene.
The MSP cold case unit, conducting routine review of unsolved homicides, submits Mixer evidence to the MSP forensic laboratory. The evidence includes a nylon stocking used as a ligature, items of clothing, and a blood sample scraped from Mixer’s left hand during her 1969 autopsy. The laboratory develops a DNA profile from perspiration stains on the stocking. The precise date of testing for Leiterman’s CODIS reference sample remains unclear in the trial record; testimony established it was processed between July 17 and July 23, 2002. The Mixer cold case analysis was ongoing during this same window.
Source: Michigan v. Leiterman trial record; dna-criminal-justice.blogspot.com documented testimony transcript
December 2003: database analysis returns two matches. Leiterman matches stocking DNA. John Ruelas matches blood from Mixer’s hand. Ruelas was four years old in 1969 and living in Detroit.
The CODIS database search returns two matches. The DNA profile from perspiration stains on the stocking matches Gary Leiterman, then 61. A separate match is found in a blood sample scraped from Mixer’s left hand: the profile belongs to John Ruelas, who had been convicted of murdering his mother in 2002 and whose DNA had entered CODIS as a result. In 1969, John Ruelas was four years old and living in Detroit. He could not have been present at the Mixer murder scene. His DNA on 1969 evidence is explicable only by laboratory contamination. The laboratory supervisor was notified. An internal review was conducted. The laboratory’s position was that the presence of two different DNA contributors on 1969 evidence did not compromise the analysis of Leiterman’s profile.
Source: Michigan v. Leiterman; Wixted et al., Bayesian Statistical Analysis; Newsweek (2016) citing laboratory testimony
Gary Leiterman is convicted after fewer than three hours of deliberation. MSP deputy director states that releasing raw electronic data for independent analysis would be “tantamount to evidence tampering.”
Leiterman is convicted of first-degree murder and sentenced to life without parole. The jury deliberates for under three hours despite hearing testimony that a four-year-old’s DNA appeared on the evidence. The defense contamination argument does not prevail. Post-conviction, the MSP Deputy Director of Forensic Science issues a statement opposing the release of raw electronic data for independent analysis with non-validated software. Professors and outside scientists request the data to perform independent statistical review of the contamination question. The MSP declines. Three scientists, including UCSD Distinguished Professor John Wixted, subsequently publish a Bayesian statistical analysis concluding the contamination explanation is statistically overwhelming. Leiterman dies in prison in 2019.
Source: View from Wilmington blog citing Professor William Thompson’s analysis; Wixted et al.; Moms & Mysteries podcast (2025)
When the Detroit Police Department crime lab closes in 2008, the MSP inherits a warehouse of untested rape kits — over 11,000 — discovered in storage, most never submitted for analysis.
The Detroit Police Department crime lab closes in 2008 following a crisis of inadequate staffing and processing failures. The MSP assumes jurisdiction. In the process, the scope of the backlog becomes clear: over 11,000 rape kits had been stored in Detroit Police facilities, never submitted for forensic testing. The kits represent cases in which biological evidence was collected and then warehoused rather than analyzed. The inheritance expands the MSP backlog substantially and makes the existing accreditation issues more acute.
Source: Imran Syed, University of Michigan Innocence Clinic, quoted in Spartan Newsroom (2021); Michigan State Police documentation
The MSP forensic laboratory backlog peaks above 19,000 cases. The lab’s accreditation from the national board expires and requires multiple extensions, with the board warning officials not to request another.
The American Society of Crime Laboratory Directors/Laboratory Accreditation Board grants the MSP a third accreditation extension but warns that no further extensions will be granted. The backlog, at its peak of over 19,000 cases, represents the cumulative effect of understaffing, inadequate budget allocation, and the Detroit rape kit inheritance. Governor Rick Snyder proposes new resources to address the backlog. The lab director testifies before the House Appropriations Committee. Rep. Ronnie Peterson notes that the backlog may leave innocent people incarcerated. The lab director reframes the definition of “backlog” to minimize the apparent severity of the delay.
Source: Detroit News (2012), archived via Illinois DUI Attorney blog; Spartan Newsroom (2021)
Cathy Broad’s documented advocacy: MSP will not digitize the OCCK case files, will not upgrade DNA sequencing on OCCK evidence, and will not use third-party laboratories for reanalysis — despite many jurisdictions with older cold cases having done all three.
Catherine Broad, whose brother Tim King was the fourth OCCK victim, has documented through years of FOIA requests and public advocacy that the MSP has refused to digitize the OCCK case files, declined to pay for an upgraded Y-STR DNA sequence from Kristine Mihelich’s autopsy, and refused to submit OCCK evidence to third-party laboratories such as Othram Labs, which has resolved cold cases of similar vintage in other jurisdictions. The Boulder, Colorado Police Department, frequently cited as a mishandled investigation, has digitized all of the JonBenet Ramsey case documents, inventoried 2,500 pieces of evidence, and reviewed 40,000 reports. The MSP has not digitized the OCCK files. In a 2024 interview, an MSP Bridgeport crime lab analyst described pulling blue jeans from a cold case and watching them crumble when touched.
Source: Broad, Catherine. catherinebroad.blog, 2022-2025. Multiple documented posts on digitization refusal and third-party lab requests. ClickOnDetroit (2026) quoting Broad.