For decades, forensic analysts have performed examinations of evidence collected in criminal cases to help answer investigative questions and establish connections between suspects, victims, and crime scenes. Chemical analysis is used to determine if a package of suspicious powders or leaves are controlled, illegal substances or if traces of chemicals are present on evidence that could have been used to accelerate a fire. Microscopes have been used to observe and compare characteristics on fired bullets, hairs, and fibers. Other specialists visually compare tire tracks, footwear impressions, and fingerprints. Those trained in medicine help investigators and courts understand how and by what means a victim sustains a fatal injury, or if marks in their flesh could have come from another person’s teeth.
In the late 1980’s and early 1990’s, scientists studying deoxyribonucleic acid, or DNA, which carries the genetic instructions used in the growth, development, functioning, and reproduction of all known living organisms, saw its potential for use in the investigation of crimes. Blood, semen, saliva, and even skin cells left behind at a crime scene have the potential to identify a suspect with a high degree of certainty through DNA testing. When DNA testing was applied to evidence in cases where people had been convicted because of other types of forensic analysis, like hair or bite mark comparisons, many defendants were proven to be actually innocent of the crimes for which they were convicted. These exonerations shed new light on other forensic practices, and as we study these exoneration cases, we see that un-validated or improper forensic evidence is a contributing factor in one out of every five of DNA exoneration cases.
In 2009, after hundreds of innocent men and women had been exonerated, many with new DNA testing, the National Academy of Sciences (NAS) published a landmark report called Strengthening Forensic Science in the United States: A Path Forward. This report clearly outlines that there are serious gaps in the science practiced by biologists, chemists, and physicists in other fields, and their analyst counterparts at crime labs. With the important exceptions of chemical analysis and DNA, the examinations being conducted at crime labs came about because of their usefulness in criminal investigations, not the scientific method. The scientific method requires a specific line of inquiry: the observation, measurement of data, and testing of hypothesis are defining characteristics. According to NAS, many forensic science disciplines “have never been exposed to stringent scientific inquiry.” In other words, just because an analyst is using the tools of a scientist, it does not mean their findings are based on experiments with validated and peer-reviewed conclusions and replicable results. Jurors must be able to trust that the testimony presented as science is valid, and for this reason the National Academy of Sciences has called for rigorous review and reform.
When a defendant applies to MIP for assistance, we track the type of forensic analysis used in their case. MIP, along with many Innocence Network organizations around the country, has identified three disciplines that recent progress in forensic science has shown to be particularly problematic: the microscopic comparison of hairs, forensic odontology and bite mark comparisons, and fire scene or arson investigations.
The microscopic comparison of hairs
In 1981, Kirk Odem, a young man from the District of Columbia, was convicted of rape, sodomy, armed robbery, and first degree burglary. Earlier that year, a woman had been attacked by a stranger, who broke into her home armed with a gun. The woman caught only a glimpse of her attacker before she was blindfolded. Her assailant raped her and stole hundreds of dollars in traveler’s checks. Mr. Odom became involved in the case when a police officer thought he resembled a composite sketch of the suspect.
At his trial, an FBI Agent testified that a hair from the crime scene was similar to samples of hair taken from Mr. Odem, and stated that “the samples were indistinguishable”. The agent testified that he had only seen eight or ten instances of hairs that were indistinguishable in the past ten years, during which time he had analyzed thousands of hairs. But hair microscopy is not an exact science and should not be used as a means to conclusively identify a victim, suspect, or any individual. The analyst’s findings were exaggerated, and as DNA testing would later prove, incorrect. Mr. Odem was exonerated in 2012.
We know intuitively that some hairs look different than others. A person with short blond hair will have hair that is different from a person with long black hair. But hair examiners were looking at hairs microscopically and observing different characteristics and comparing features that are much subtler, such as cuticle thickness and the size of pigment granules. We don’t know how often these characteristics exist in our population, so unlike with DNA testing, we don’t know the rarity of a visual “match”. Further, there is no agreed upon or standard number of characteristics required to determine if one hair “matches” another. This means that two hairs that come from different people can look exactly the same and we have no valid way of knowing how often that happens.
Examining hairs microscopically is useful for elimination. For instance, if a hair found on the shirt of a victim is compared to hairs collected from a suspect and the analyst observes visual differences, it’s reasonable to conclude that the hair did not come from that suspect. Hair can also be subjected to DNA testing, if either nuclear DNA on root tissue is present on the hair, or if mitochondrial DNA testing can be done on the hair’s shaft. Examining the hair microscopically first can be a valuable screening tool for further testing.
As we know from the large scale review of cases from the FBI’s Hair and Fiber Analysis Unit, brought about in part by Mr. Odem’s exoneration, there are serious concerns with the way that analysts were testifying to the results of hair comparisons; that testimony could have been a significant contributing factor in convicting innocent people. The American Society of Crime Lab Directors Laboratory Accreditation Board recently recommended reviews of microscopic hair comparison cases because their ethical obligation to “take appropriate action if there is potential for, or there has been, a miscarriage of justice”. In February of 2015, FBI Director James Comey sent letters to U.S. Governors, notifying them of their review and encouraging reviews of cases from state-level Crime Labs where analysts were trained by the FBI. The Midwest Innocence Project has launched a review of microscopic hair comparison cases in its five-state area. Read more about that review here.
In 1992, Ray Krone was convicted and sentenced to death for the murder of a young woman in Phoenix, Arizona. The victim’s body was discovered at the bar where she worked. Investigators noted that there were apparent bite marks on her body. A friend of the victim’s told police that Mr. Krone had planned to help the victim close the bar on the night she was killed. A Styrofoam impression of Mr. Krone’s teeth was compared to the marks on the victim’s skin and experts testified at trial that they matched. Eventually, DNA evidence proved Krone’s innocence and he was released from prison in 2002.
Forensic odontology is the process by which odonatologists, who study the structure and diseases of teeth, attempt to match marks found at crime scenes with the dental impressions of suspects. If a victim is bitten by a perpetrator during a crime and police have a suspect, odonatologists can attempt to “match” the bite mark to the suspect’s teeth.
Although bite mark evidence has been used across the country in many criminal prosecutions, there is no real scientific support or research into the accuracy or reliability of bite mark evidence. We do not know how many people have similar-looking teeth impressions. Further, injuries that appear to be a bite can actually be an unrelated injury, or a mark on the skin that is mistakenly identified as a bite. Human skin is elastic; it swells, heals, and can deform or warp a bite, making it difficult to affirm the accuracy of measurements. Perhaps most importantly, there has been no scientific validation for the notion that a person’s dentition is unique to him or her.
As with other impression comparisons such as fingerprints, tire tracks, footwear impressions, and ammunition components, these comparisons, by their nature, involve a level of subjectivity. Thus, according to the NAS Report, different experts have found different results when looking at the same bite mark evidence.
Fire scene investigation
Kristine Bunch, a young mother from Decatur County, Indiana, was convicted of the murder of her three-year-old son; she was also convicted of arson, for allegedly setting the fire that took his life. Prior to her arrest, a state arson investigator determined that the fire had been started in two locations and that accelerants, such as kerosene or charcoal lighter fluid, had been used to start both fires. At trial, the investigator testified that there was a trail of accelerant from the bedroom where the little boy died, through the home, and out the front door of the residence. An Agent from the Bureau of Alcohol, Tobacco, and Firearms also testified that he found accelerant in both locations. Ms. Bunch was exonerated in 2012 when the defense became aware that the ATF agent had not found accelerant in two locations, only one – in the living room where a kerosene heater was used to keep the home warm in the winter. After further review of the case by three separate fire scene experts, Ms. Bunch was set free.
Fire scene investigation seeks to determine the cause, origin, and development of a fire or explosion. To determine where a fire originated is key to understanding the cause of the fire. Fire scene investigation, sometimes referred to as arson investigation, is a combination of inquiry into the means, motive, and opportunity of a suspect to have started the fire, as well as an analysis of physical evidence. The combination of the dual lines of inquiry is problematic because of the potential for the contaminating effects of cognitive bias. Cognitive bias can take many forms in criminal investigations and forensic analysis, with fire scene investigation being particularly vulnerable to confirmation bias and role bias. Confirmation bias occurs when we look for information that confirms what we already believe to be true and ignore information that contradicts that belief. Role bias can occur when a forensic scientist or an investigator identifies as an “arson specialist” and may be naturally more inclined to interpret what they observe as signs of arson, which can affect their objectivity. This is particularly dangerous when the investigator’s duty is to determine if a fire was an accident, or was intentionally set.
The 2009 NAS report makes clear the need for forensic science to be independent from law enforcement. The report states “[F]orensic investigations should be independent of law enforcement efforts either to prosecute criminal suspects or even to determine whether a criminal act has indeed been committed.”
In addition to the potential for cognitive bias, the science behind the determination of the origins of fires lacks sufficient validation. According to Paul Bieber, Director of the Arson Research Project, while “fire behavior has been the subject of a good deal of scientific research, studies regarding the behavior of the fire investigator tasked with determining how and where the fire began have been largely absent. As a result, the underlying reliability and validity of forensic fire scene examination remains in question.”