Conviction: Unvalidated Science in the Courtroom

The movie Conviction is a compelling hit. It is both a condemnation of the criminal justice system and a celebration of familial love that leads a sister to find the evidence to exonerate her brother who was wrongfully convicted of murder. In that movie -- which was based on a real case -- science, specifically DNA testing of long concealed evidence, was also a heralded hero. But DNA evidence in the courtroom, does not exemplify forensic science in the ordinary criminal case.

Last year the National Academy of Sciences (NAS) released its report Strengthening Forensic Science in the United States: A Path Forward to impel the much-needed changes in how forensic sciences are used in our criminal justice system. "It is clear" the report concludes, "that change and advancements, both systemic and scientific, are needed in a number of the forensic science disciplines..." Some of our sacred cows that have caused the conviction and long term incarceration of innocent people: hair matching, bite mark comparison, handwriting analysis, tool mark comparison, fingerprint analysis, shoe prints and fabric imprints. The report details how bad science is pushing good science out of the criminal justice system. Some of the forensic sciences have been shown to be inherently unreliable. They are unvalidated science, that is disciplines that have not been validated scientifically by independent research.

The NAS report criticizes the lack of a proper scientific basis for the acceptance by courts of many of the forensic sciences especially the sciences that rely on subjective visual interpretation by individual examiners. "Match" is often the key word in such testimony, which implies to a jury that there is a 100 percent ability to connect the suspect to the victim. Yet, science in the courtroom can never be 100 percent because there are too many uncontrollable variables outside of the laboratory.

Our most hallowed example of comparison forensic evidence is fingerprinting. Even if we assume that everyone's fingerprints are different without having actually proved it -- just as we assume that every snowflake is different -- mistakes can be made when comparing a small partial poorly formed fingerprint to a full set of prints carefully made on a special card. The NAS report refers to Oregon attorney Brandon Mayfield's 2004 arrest for the terrorist bombing of a train in Spain when the FBI examiners wrongly matched a partial fingerprint to him and his own defense expert agreed. Mayfield's wife was a Muslim and he had converted to Islam. Once the FBI examiners learned about his background they became more entrenched in their previously formed erroneous opinions.

There are many other examples of flawed science testimony permitted in the courtroom, not addressed in this report such as incorrect autopsy findings, bullet lead analysis, refreshed memory recall, and most significantly, eyewitness testimony.

Often, after the police investigation has identified the perpetrator, the prosecutor seeks scientific evidence not as an investigative tool but as a means of adding the mantle of science in the courtroom. The crime lab scientists are considered part of the prosecution team. But scientists should not be team players. Scientific findings are what they are; they must be independent of the desires of the prosecution or of the defense. Scientists, however, may be swayed by the information provided by whomever they are working for, which information is often one-sided. Right now, 260 innocent persons who have been wrongfully convicted and released from prison, some from death row, on the basis of DNA evidence. DNA, which was born in research laboratories, where all variables are controlled, not in a courtroom, is a reliable science because it is not controlled by subjective bias of an individual examiner unlike many of the other forensic sciences criticized in the NAS report.

The adversarial system has not prevented bad science from bearing witness. Rarely has there been any punishment for an expert who has testified to false information, for a prosecutor who has been overzealous, for a defense attorney who was inadequate or for a judge who allowed the bad evidence to be presented to the jury.

Twenty-four hour cable news and the Internet often downplay the presumption of innocence and have added to potential jury bias by disseminating premature and inaccurate scientific information to the public.

There are other shortcomings in the justice system. The solving of murders is falling greatly in this country; from a "solved" rate of over 80 percent during the 1960s to about 50 percent today. As the pressure to solve these crimes and protect the public increases, the pressure of bad science and bad scientists in the courtroom will be greatly increased.

Some years ago the Supreme Court tried to address the issue of the admission of valid science in the courtroom. In a decision called Daubert, criteria were established for the admission of scientific testimony in the federal courtroom. But that decision did not contemplate the issue of biased forensic science in the criminal courtroom. And it may have unintentionally exacerbated the problem of invalid science that has now leeched into both federal and state criminal proceedings. It also spawned certification mills for experts, and gave birth to a deceptive procedure to validate biased science. "Peer review," one of the five benchmarks suggested in Daubert to ensure that another scientist independently confirmed scientific methods and opinions, has been misused. Some criminalists now testify that their supervisor, who is administratively responsible to sign off on their work, agrees with their opinion. This type of hearsay second opinion, cloaked in the aura of experience from an allegedly more knowledgeable scientist under the guise of "peer review," doubly misleads a jury into believing invalid evidence is based on a valid science.

Our president, Congress, and our courts should pay attention to the NAS report and devise ways to improve our forensic science infrastructure. Bad science will still enter the courtroom unless and until there are strict national standards formulated by independent scientists who are not subject to prosecutorial, legal or judicial manipulation. The NAS report recommends that Congress should establish and appropriate funds for an independent federal entity, the National Institute of Forensic Science. Scientists need to be properly trained and board certified by appropriate agencies not diploma mills. But most important, they need to be independent.

Unfortunately, a disturbing trend is arising. Some, instead of reforming the problems, are protecting existing law enforcement crime labs sections to continue as is, rather than ensuring the rigorous independent scientific validation of forensic testing. Some governmental bodies are seeking to give their imprimatur to certain past crime lab practices merely because some other third party accredits lab procedures. For example, recently the New York State Forensic Science Commission was asked to accredit a unit of a police lab in duct tape comparison. Over the objection of some of its members, the Commission voted to give the New York State accreditation based upon the an accreditation by a group known as the American Society of Crime Lab Directors. This was done even there are no such validation studies of this 'discipline', the exact type of lack of scientific validity the NAS report highlighted.

Until long-range solutions can be implemented, interim stopgap remedies must be employed. Judges in the many fragmented jurisdictions across the United States are not equipped to be the gatekeeper of forensic science or scientists during the many years required for this transition to take place. Select caches of judges in all jurisdictions need to be immediately trained about unvalidated science. And just as is done now in the mass tort cases, which only involve money, these specialized judges should be assigned criminal cases concerning the deprivation of one's life or freedom based on forensic science.

The solutions are massive and extreme, but then again, so is the problem.