Bite Mark Data

In February, the Innocence Project filed a lawsuit asking for access to the federal archive of bite mark convictions at the National Museum of Health and Medicine. The Innocence Project believes that accessing the cases could “hold the key” to building cases for exoneration for individuals still imprisoned based on bite mark evidence.
In an Innocence Project article discussing the lawsuit, Chris Fabricant is quoted, saying “This archive could be the only way to identify unknown cases in which bite mark evidence was used and, potentially, produced wrongful convictions.” The lawsuit comes after the Innocence Project had been denied permission to conduct research within the archive, despite the museum normally allowing research of the archive. Furthermore, the Innocence Project filed a Freedom of Information Act application for the archive’s records last year, but so far has received no reply.
The PCAST Report noted, in 2016, that of scientific studies of bite mark evidence, “the observed false positive rates were so high that the method is clearly scientifically unreliable at present.” Read the Innocence Project’s take on their lawsuit at

Geneological DNA Analysis

A BuzzFeed News article by Peter Aldhous, published in February, described the newly emerging genealogical DNA analysis industry, centered around two private forensic science services. The piece, “The Golden State Killer Arrest has Spawned a New Forensic Science Industry,” also discusses how genealogical DNA analysis techniques work, and some of their pitfalls, including the possibility of wrongful conviction.
Last year’s arrest of Joseph James DeAngelo, alleged to be the “Golden State Killer”, through genealogical DNA testing sparked a national conversation about the field. In the DeAngelo investigation, police likely uploaded a DNA profile – believed to be of the perpetrator of the murders – to a genealogical website, which would provide a list of individuals (who had submitted their DNA to the site) sharing similar DNA profiles. From that list, police would have been able to note the families that these individuals belong to, and then would have been able to use widely available genealogical records to look for specific individuals. Since the police knew the Golden State Killer’s gender and general age, they would have had a narrow group of individuals to focus their search on, believing one of them to be the suspect. They could then use conventional DNA standards from those individuals to reaffirm their suspicions.
Recently, two companies have begin to provide similar services: Parabon NanoLabs and Bode Technology. These companies are hoping to solve cold cases through genealogical DNA analysis, but as Aldhous points out, “the big question is who is going to pay, because local police and crime labs already have stretched budgets and are struggling to keep up even with the demand for conventional DNA testing.”  Furthermore, genealogical DNA analysis can point law enforcement in the direction of innocent people. Just like conventional DNA testing, genealogical analysis and its application are not perfect. Aldhous discusses two cases where genealogical DNA analysis led police to innocent people and families, but does not note any wrongful convictions that have occurred.

Judge Rakoff on Forensics

Judge Jed Rakoff, a district court judge in the Southern District of New York, recently spoke in a video by Inside Science describing how types of forensic evidence “continue[] to be routinely admitted by the courts, both state and federal, even though considerable doubts have now been raised as to whether forensic science really is science at all, and whether it is reliable and valid.” Rakoff, who had been a member of the National Commission on Forensic Science, also discussed how judges should approach scientific evidence. His comments are available at

Duke Law Forensics Events

Duke Law School will host two forensic evidence programs in March 2019.

Getting Forensics Right: Strengthening the Connection Between Forensics, Statistics, and Law Ten Years After “A Path Forward”

Wednesday, March 6, 2019 • 12:30 PM • Law School 3037

What are the stakes when forensics go wrong? Keith Harward will tell his story: he was exonerated by DNA testing, but spent 34 years in prison in Virginia for a murder he did not commit, based on multiple erroneous bite mark comparisons. Peter Neufeld, co-founder and co-director of the Innocence Project will join in the conversation. M. Chris Fabricant, who directs special litigation for the Innocence Project, will moderate. Prof. Brandon Garrett will introduce the panel. Lunch will be provided. The conference is made possible for the Center for Statistics and Applications in Forensic Science (CSAFE), and it is also supported by the Innocence Project. Sponsored by the Duke Wrongful Convictions Clinic and Criminal Law Society. For more information contact Prof. Brandon Garrett at Registration is not required.

Whiskey in the Courtroom – Evolving Trends in Forensic Science: Cognitive Bias in Forensic Science and in the Courtroom

Friday, March 8, 2019 • 8:45 AM • Law School

Indigent Defense Services, the Duke Law Center for Criminal Justice and Professional Responsibility, and the Center for Statistics and Applications in Forensic Evidence are co-sponsoring the fifth annual CLE, “Whiskey in the Courtroom: Evolving Trends in Forensic Science.” Forensic science experts and attorneys who have litigated cases involving complex scientific evidence will present on a range of topics designed to help attorneys understand the latest trends in forensic evidence, limitations to this evidence, and legal challenges that can be made. This year the program will have a special focus on Cognitive Bias.

Attendees will pay $25 to cover the cost of food and parking. IDS will cover that cost for IDS employees. The program is expected to carry 6.25 CLE hours. If you want CLE credit, you will need to pay the NC State Bar the CLE fee of $3.50 per credit hour.

Registration is nearly at capacity. Sign up here.

Delaware CODIS Failures

An investigation into the Delaware Division of Forensic Science has yielded no criminal charges, according to an article written by Esteban Parra in the Delaware News Journal.  The investigation sprung out of the Delaware Medical Examiner’s Office failure to enter the data of more than 1,600 convicted individuals into the CODIS system. The investigation began last year after a box had been found four years earlier at the Delaware Medical Examiner’s Office, containing DNA samples from those individuals. These samples, taken between 2001 and 2012, had not been entered in CODIS.

The scandal over the unentered entries occurred while the state’s forensic science programs were administered by the state Medical Examiner’s Office. In 2014, the Division of Forensic Science in Delaware was transferred to the Delaware Department of Safety and Homeland Security. The transfer occurred after the division was the subject of a series of scandals, including an evidence mix-up.

One of the unentered samples contained in the box was DNA taken in 2002 from an individual convicted of a sex offense, who was indicted last year for three rapes that occurred in 2010, 2014, and 2017. The 2010 and 2014 cases were considered cold by investigators before his DNA was entered in CODIS in 2014.

The State’s Department of Safety and Homeland Security addressed concerns about the data input scandal by announcing a set of reforms. According to Parra’s article, the department authorized two more individuals to put data into CODIS, and it also created a spreadsheet to track samples between their entry to the lab and their inventorying.

Parra’s write-up of the conclusion of this investigation is available at

LA Times Op-ed

In January, Edward Humes wrote an Op-Ed in the Los Angeles Times, titled “Bad forensic science is putting people in prison.” In his article, Humes describes the human consequences of forensic science errors, discussing cases like those of Jo Ann Parks and Bill Richards.

Parks is still in prison, convicted based on now-discredited arson investigation methods. Humes states an expert witness testified flashover did not occur in a housefire that Parks survived (flashover, according to Humes, can “make an accident look like arson.”) Many forensic scientists now believe flashover did occur at this fire. Despite the new confidence of some forensic scientists that Parks was convicted using erroneous expert conclusions, Parks’s conviction has not been overturned.

Richards was convicted of murdering his wife, based on bite-mark testimony. Richards was incarcerated for 22 years, and remained in prison even after the bite-mark expert recanted and DNA evidence implicated another person. According to the Op-Ed: “Richards’ release was delayed for eight years after prosecutors argued successfully that only factual testimony, not expert opinions, can be false evidence under California law.” It took an act of the California legislature to change this definition and allow Richards to finally be freed.

Humes uses these cases to emphasize the need for forensic science reform. He states, “the stakes are too high to maintain the status quo.” Read Humes’s Op-Ed at

Breath Tests in NJ

Breath test results in New Jersey, presented as evidence supporting more than 20,000 convictions, have been ruled inadmissible by the New Jersey Supreme Court. The Alcotest-brand breath tests, which according to the court were generally admissible if properly used, require calibration every six months. The police officer formerly responsible for calibrating the tests across five New Jersey counties was charged two years ago with falsifying records because he wrongfully certified he followed proper calibration procedure.

The officer was supposed to measure the temperature of a “simulator solution” with a thermometer to ensure that the solution matched body temperature. The testing of the simulator solution quantifies the level of measurement uncertainty present in the device, meaning, the calibration gives a numerical range for how far away the results of the breath test might be from the real Breath Alcohol Concentration of the tested individual.  Since the state calibration is the only non-manufacturer produced check on the devices, an expert witness testified about the necessity of calibration stating that, without proper calibration, “if [the manufacturer] accidently [sic] used the wrong temperature in calibrating the calibration units and the probes, then the temperature variance would go undetected and the Alcotest’s readings would be factually inaccurate.”

Nick Corasaniti and Sharon Otterman of the New York Times wrote an article further explaining the ramifications of this decision, available at The decision itself is available at


Firearm “Fingerprint”?

This piece describes how in response to gun crime, “A new machine, on loan from the Bureau of Alcohol, Tobacco and Firearms … scans spent bullet casings from crime scenes” and enters them into a national database.  “Spent shells are loaded into the machine, and up comes 3-D scans of the unique scuffs and divots that every gun leaves behind… The scans are sent to technicians in Alabama who can match them up instantly to identical ones from another shooting—across county or state lines, to show they came from the same gun.”

The article inaccurately describes a claim that “every gun leaves” marks that are “unique” and that this is “just like your fingerprint.”  And the article discusses a “unique ballistics imprint.”

No surprise, perhaps, that the article does not mention the presence or lack of research to support such claims about the technology and its reliability.  The PCAST Report does speak to such firearms comparisons and the unfortunately lack of foundational research.

Roberts Exoneration

Earlier this month, Horace Roberts was exonerated after two decades of wrongful imprisonment. Roberts had been convicted of a 1998 murder. A watch had been found near the crime scene, and when police showed Roberts a picture of it, he stated that he thought it was his. Later, however, a similar watch was found at Roberts’s apartment. DNA testing was not done on the watch before Roberts’s initial trial.  DNA testing would eventually support Roberts’s innocence claim. When testing was eventually done on the watch, it was consistent with the DNA profile of Googie Harris, Jr. Harris Jr.’s father. Testing was also done on the victim’s clothing and fingernails, which produced a CODIS result consistent with the profile of Joaquin Leal, who is Harris Sr.’s nephew.  After the California Innocence Project announced the DNA results, the Riverside County District Attorney’s office reopened their investigation into the death. The Riverside DA dismissed the charges against Roberts and instead charged Harris Sr. and Leal with her murder.

The California Innocence Project’s writeup of Horace Roberts’s wrongful conviction is available at Sandra E. Garcia of the New York Times also an article detailing his case, conviction, and exoneration, available at


Siggers’ Firearms Exoneration

Darrell Siggers was released from prison in August after his conviction for a 1984 murder was vacated, but just last week prosecutors decided not to seek a new trial. He has maintained his innocence since he was accused of the murder three decades ago. In Siggers’s 1984 trial, an expert in the Detroit Police Crime Lab – testified that bullets he found from the victim’s body, bullets at the crime scene, and bullets in Siggers’s home (a vacant apartment next door, really) were fired from the same gun. Investigators never discovering the murder weapon. The firearms work in this case was reviewed in 2015 by another expert, David Townsend, who stated that Houseworth’s findings were “unbelievable,” “erroneous,” and “highly improbable”.
The Detroit Police Crime Lab – which had conducted the testing of firearms evidence in Siggers’s case – was closed in 2008 after an audit found a host of serious problems in the lab. Auditors had found that “[n]otes, photographs, sketches, and/or other examination documentation are almost nonexistent in the case file records to support the conclusions of the examiners.” They had also discovered an almost ten percent error rate in its firearms work, which according to the audit, created “a strong likelihood of wrongful convictions”.
Read Sarah Cwiek’s report on Darrell Siggers’s wrongful conviction and his exoneration at
%d bloggers like this: