DNA evidence is often reported as being infallible, concrete proof, or a 20 billion to 1 chance of being wrong.
Because of the nature of DNA, the men in white coats, and the sheer power of science this seems to be true; and in part it is. The DNA for every person in the world (who is not an identical twin) is unique, this is due to the sheer volume of combinations of the chemicals that make up a DNA strand. From this it would appear that any sample of DNA can be directly attributed to, with 100%, certainty, to a given individual. In the same that fingerprints can.
However, surety of DNA profiles is not quite what it may seem.
According to a report into DNA testing conducted in Texas, “one in every hundred forensic tests performed on the DNA of suspected criminals may give a false result”, this was based on a study of error rates, most of which are caused by human error. The UK were very quick to assure the public that the DNA system worked. Professor Derek Pounder of the Department of Forensic Medicine at the University of Dundee, said: ‘The increased quality controls that have been introduced to laboratories over time are a reaction to errors being found.’.
This response is interesting for two reasons: First it implies that that there have been errors within the UK system, in fact the errors were such a problem that they were common enough to be detected and then to introduce new systems to combat this. Secondly, while the forensic scientests in the UK have been very keen to assure that there are now no longer errors, it is the same body of people who assured the public that they were not wrong about fingerprints. While it is not fair, nor responsible, to tar every forensic scientist with the same brush, it should be recognized that major mistakes have been made and it took years to rectify them.
In 2008 it was reported that there were over 550,000 false, miss spelt or incorrect names on the UK DNA database, which is 1 in 8 entries.
The DNA company Promega, have an old paper on their site that shows the chance of a random match (i.e an error) varying form 1 in 100 billion to 1 in 100,000, this huge variation appears to depend on the technology being used, and does not allow for human error.
- In 1995, in the US, a man was charged with a sexual assault based on DNA evidence. While in the lab DNA samples became mixed. The swab taken from the the victim’s body had the label on it switched with the label of the DNA for the suspect. From this the lab conducting the work, CellMark, concluded that the swab taken from the victim contained the suspects DNA, after matching it with a known DNA of the suspect, therefore the suspect’s DNA was found on the victims body. A Cellmark employee, Charlotte Word, caught the error while testifying in court.
- In 1999 there was a DNA error at the Philadelphia police, which was almost identical to the errors listed above. The picture shows how this information recored by the police.
- In 2000 Raymond Easton was arrested for burglary, based on DNA evidence. Following a second test, the man was released as it was realized that the wrong man
- In 2001 a man was informed, following a DNA paternity test, that he was not the father of his child.
- In 2002 a man was jailed for a year for sex offences based on DNA evidence. It was later found that there was an error in the DNA labs, which resulted in a fale positive.
- In 2003 Peter Hamkin, from Liverpool, was arrested for a murder in Italy, based on a DNA sample from the crime scene and a DNA sample taken from him years before (he was never charged or convicted for any previous offence). Peter had never been to Italy, had proof he was at a different location, and did not commit the offence. The error occurred due to an error with his DNA profile on the database.
- In 2006 Kevin Reynolds was arrested for the murder of Sally Anne Bowman, based on DNA evidence. The DNA evidence was latter found to be an error, he was released and later Mark Dixie was found guilty of the murder. Kevin Reynolds later started legal action against the Met Police.