Monday, February 22, 2016

How Accurate are the Various Forensic Techniques Used in Criminal Cases?

On the morning of June 25, 1982, two Latino men broke into a woman's apartment in Stock Island, Florida. One of the men sexually assaulted the woman and had deposited semen on her clothes.  After the perpetrators left, the victim called the police and described the situation, noting that the Latino man who was in her bedroom was not wearing a shirt and had no hair.  The officers who were contacted, searched locations near the perimeter of the crime scene.  They found a man who fit the description.  He had no shirt, no hair and was of Latino descent.  He was at a convenient store parking lot near the victim's apartment.  The victim identified this man as the perpetrator who was in her bedroom.  The man who was identified was Orlando Boquete.  Boquete was arrested and charged with burglary and attempted sexual battery.  Although Boquete testified that he was with his family all night and later went to the convenient store with his cousins, the victim had identified him as the perpetrator and a blood test analysis did not eliminate him as a suspect (Type A blood and other spots were found at the crime scene.  Boquete was tested and the results showed that Boquete had Type O blood.  He was not excluded because the spots that did not have markers could have been his).  After Boquete's trial, he was sentenced to 50 years in prison.  After spending almost half the time, Boquete was exonerated based on DNA tests from the victim's clothing.  Orlando Boquete was a free man after 23 years.
Boquete is not the first man to have been exonerated for eyewitness misidentification and/or improper forensic science.  According to the Innocence Project, 337 people in the US have been exonerated since 1989 due to DNA tests that have proved them to be innocent.  The two main causes of false convictions are eyewitness misidentification and 'improper' forensic science analyses .  But recently, another possible cause (not as common as the other two), are the flaws in DNA processes.  In this post, I will address the two main causes, and note ways forensic analysts and anthropologists have used more precise and accurate techniques to exonerate the innocent people convicted. Then, I will discuss a new possible factor of false convictions caused by flaws in DNA processes and what ways the technique can be improved objectively.



Eyewitness Misidentification

Eyewitness misidentification is one main factor that is related to false convictions.  According to the Innocence Project, when looking at 179 cases of people falsely convicted, 75% were due to eyewitness misidentification.  Eyewitness misidentification can be a result of many factors.  One such factor is the concept of false memories.  False memories can occur based on a variety of reasons, but one important reason that I want to discuss further, is how post-event information (such as media coverage on the incident) and/or misleading information/evidence/statements effect the victim when identifying the perpetrator. 

According to Loftus, (2003) Roediger & Gallo, (2004) and Schacter & Scarry (2000),

"False memory refers to the memory distortion in which people sometimes develop vivid and detailed recollections of events that were never experienced; or people confuse events that happened before or after the target event with the event itself."

False memories are not only a factor in eyewitness misidentification, but an occurrence in peoples normal lives. When recalling childhood memories, one might think that they are accurately remembering an event that occurred in their childhood, but, might actually be a 'memory' influenced by interactions and experiences in life.  Researchers conducted a study in 1995 to better understand the concept of false memories.  Their subjects consisted of 20 college students who were interviewed by researchers about specific events that occurred in their childhood.  The researchers sent a questionnaire to the subjects' parents for specific information.  The questionnaire included things such as the age of the subject (range from two to ten) when they experienced any of the six categories (listed below), as well as distinct information about the specific events.
  1. Getting lost
  2. Going to the hospital
  3. An eventful birthday
  4. Loss of a pet
  5. A family vacation
  6. Interaction with a prominent or famous person
In one of the experiments, the researchers interviewed the subjects twice about events that occurred in their childhood and compared the results to their parents questionnaire (which depicted the more accurate result).  The researchers included false information for some events along with true information for others.   In the first interview, subjects recalled 62 out of the 74 true events. (an average of 3.7 events per subject).  In the second interview (which took place a couple days - a week after), 4 out of the 20 subjects incorporated the misleading/false information when they described two events they had 'recalled' after the first interview.  The researchers stated that this was due to connecting the false information to something true that occurred.  For example, one of the subjects knew the event of having a birthday party was real, and accepted the false information (a clown attending the party) as something that could have occurred.  The fact that the subjects recalled false details based on the information the researchers gave them, can be compared to eyewitness memories.  Both can be distorted after the event (first interview in the case of the study), based on post-event information.  In the case of the study, the false information given to the subjects by the researchers, contributed to the subject 'recalling' specific details of a simple event that occurred (birthday party). One thing to note in this study is that it includes a small sample size and the subjects are close to the same age.  This might lead to skewed results.  However, there are other studies that show that regardless of sample size, age and sex, the results are similar.  False/misleading information can alter the memory of the individual when recalling a specific event.

The main reason this occurs, is because the brain has not evolved to remember every specific detail of every event that has transpired in one's life.  If this were the case, there wouldn't be as much room for other useful information, and people would have to remember memories that they would rather want to forget.  So, the brain has evolved to remember very impactful and important memories (long-term memories), rather than minor information, such as a shirt someone wore, what you ate a week ago, etc. (as long as none of those details are meaningful, you most likely wont remember them as vividly as your first pet, your first job and so forth). Well, what causes one to remember some events for a longer period of time, compared to minor details/unimportant events for a few seconds to a minute? An answer to that lies within where specific information is stored based on the 'type of memory' it is.  For simplicity's sake, the human memory can be broken down into two parts; short-term and long-term memory.  Once a memory is created, it's stored in the brain.  Like I mentioned previously, because the brain does not keep all the information it receives, it must find a way to filter the stimuli and store it into the proper category it belongs in. Once a sensory stimuli has reached neurotransmitters in the brain, it activates regions of the prefrontal cortex which then processes and temporarily stores the information.  The information starts off as short-term memory (which can only hold up to 7 plus/minus 2 items for about several seconds to a minute), but, depending on the significance of the information, or how much one has repeated it over and over again, the brain will determine whether or not it will transfer the STM information into long-term memory. Long-term memory differs from STM due to the fact that it takes multiple processes to store LTM.  One important process of storing stimuli/information as LTM, is a process called long-term potentiation.  This process consists of strengthening the synapses in the neurons, which ultimately leads to remembering the information more accurately and for a longer period of time.  The hippocampus is also an important factor to LTM storage.  The hippocampus plays an important role in fusing information from STM to LTM, likely strengthening/reconfiguring neural connections for several months.  It's likely that the reason why most eyewitnesses cant remember specific details of the suspects appearance and clothing, are due to the fact that those minor details are not as significant as the event itself.  If one witnesses a car accident, is a victim of robbery or assault, the event itself is impactful enough to be stored as LTM.  The specific details are remembered as STM, but soon disappears after several seconds.  

Eyewitness testimonies can be unreliable when looking at how long STM is stored in the brain. How post-event information based on other peoples interpretations of the incident, false information, misleading evidence and statements can effect the victim when trying to identify their perpetrator.  Although eyewitness testimonies are still used today, there are ways to be more accurate. DNA testing is the most objective route, but if that is not an option, there are guidelines on whether or not a witness' testimony can be considered. 



'Improper' Forensic Science Tests

Improper forensic tests are another component of false convictions.  Improper forensic tests/analyses are defined as tests that can be unreliable and inaccurate.  Pattern-matching tests can be considered undependable and some examples of these analyses are bite mark analysis, blood spatter, and even fingerprint analysis.  Bite mark analysis can be inaccurate due to several factors.  One is that there aren't predetermined scientific guidelines and rules on how an expert goes about identifying the bite mark(s).  The expert identifies the suspect based on patterns from the suspects teeth compared to the pattern of the bite mark (i.e size, depth, gaps, etc.), this is based on subjective analysis of qualitative data rather than objective evidence.  According to the Innocence Project, another reason why bite mark analysis isn't as reliable is because most of the time the forensic odontologist is self-employed and not accredited by a scientific entity, thus receiving no oversight during analysis of the bite mark. There are cases where people have been exonerated for inaccurate convictions based on bite mark analysis.

Another example of a forensic test that has falsely convicted people is the blood spatter analysis.  Blood spatter analysis refers to the interpretation of the pattern, size and shape of the blood found at the crime scene. This assists forensic specialists in coming up with a conclusion about what occurred and how.  According to the National Institute of Justice, when a panel of experienced blood spatter analysts tried to examine and identify specific blood patterns, "13.1% of these classifications did not include the correct pattern type for the rigid surfaces and 23.4% for fabric surfaces. The highest rates of misclassification were 59% for satellite stains (pg. 3) from a drip pattern on fabric surfaces and 19% for impact patterns (pg. 2) on rigid surfaces." These results show similar subjectivity to bite mark analysis (pattern-matching).  Blood can also get contaminated and mixed in with the victims, thus confusing experts when trying to figure out whose blood it is and what the blood stain represents (when forming an opinion about what occurred).  Before DNA tests, blood spatter, hair, bite mark and fingerprint analysis were the main techniques used to determine and convict suspects.  These forensic techniques are mostly based on pattern-matching, identifying suspects based on a similarity between the pattern observed from the data and a specific trait the suspect possesses.  

DNA tests have been the main factor in exonerating the falsely accused and convicted.  In the example of analyzing a bite mark, genetic material can be recovered through the saliva where the bite mark occurred on the skin.  This is the more objective route because although DNA testing isn't perfect, the accuracy is a lot higher than that of an expert's qualitative assumption based on certain markings.  If done properly and carefully, DNA testing can also help authorities identify a suspect in a crime scene with blood spatter.  A major drawback to that is whether or not the blood has been contaminated. 



Flaws in DNA Processing/Software

DNA testing at crime scenes seems to be the most accurate way to convict a suspect.  Taking into account human errors and possibility of contamination, the odds of a coincidental match is 1 in 1 trillion (this varies based on where you look).  Sounds pretty accurate right? Well, recently three scientists who worked for the New York State Police (NYSP) crime lab have sued the agency, and alleged that administrators retaliated against the scientists who found flaws with the NYSP's way of processing DNA.  The scientists supported and recommended another program called TrueAllele that the NYSP implemented for a little bit, and then got rid of.  TrueAllele is a computerized system that "removes an element of subjectivity from the analysis called Combined Probability of Inclusion that the agency has been using for mixed DNA crime scenes, which is inaccurate especially concerning relatives with similar genetics," the scientists mention.  Although TrueAllele has been successful thus far convicting the people who are 'guilty,' some suggest that the software's secrecy is a little suspicious.  Cybergenetics, the company that runs TrueAllele claims that their source code is a trade secret and that revealing their code would ruin the company financially (since their competitors would know the code).  They claim that objective peer-reviewed studies have been done on TrueAllele and that this software has been used for over 200 cases. The fact that Cybergenetics has denied to share the code, raises suspicion on what methods they use to validate conviction of the suspect. According to the company, they objectively use a two step system to separate the contaminated sample. Based on mathematical models, they show which amount of DNA is more dominant in the sample.  With that, they can prosecute anyone who was at the crime scene (this may include accomplices and/or bystanders).  Like noted previously, contamination can occur during transportation processes, examination processes or from the crime scene itself and this can cause error when determining suspects.

In conclusion, although there are multiple techniques to solve a crime, all are inaccurate in their own way.  Eye witness testimony and bite mark analysis being the most inaccurate followed by blood spatter analysis.  DNA testing is still the most accurate out of the techniques, but it can be inaccurate if there is contamination or mixing.  Software such as TrueAllele, is said to perform with high accuracy, thus precisely extracting DNA from contaminated species.  With that, prosecutors can convict perpetrators and/or accomplices with greater accuracy.  However, the fact that Cybergenetics hides their TrueAllele software code and refused to share it in court, has given people the impression that this software may be flawed or biased. One of their 'peer reviewed' and 'objective' studies are done by their own employees at Cybergenetics.  Based on previous readings of companies/governments doing their own research, I am also skeptical about TrueAllele's code design and accuracy. As of now though, it is important to try to use forensic science techniques (especially DNA since it is the most accurate) as objectively as possible.  It's easier said than done when it comes to the US judicial system. DNA samples should be treated carefully and acquired immediately at the crime scene to lower the risks of contamination. If DNA testing isn't useful or a reliable option at the crime scene, third parties such as forensic anthropologists, chemists, physicists, etc. should be hired for specific situations to confirm or disprove a certain analysis. 

Currently, DNA testing and analysis have assisted in exonerating innocent people who have been sent to prison.  Once technology gets better and companies are willing to share codes to keep improving existing software, there will be a chance of less false convictions. So, people like Orlando Boquete don't sit in a prison cell for 23 years for a crime they didn't commit.



References
Innocence Project + Exonerations: 








Eyewitness Misidentification and STM vs LTM: 



















Bite Mark Analysis:








Blood Spatter Analysis:






DNA & TrueAllele Software:









Posted by at

No comments:

Post a Comment

Subscribe to: Post Comments (Atom)