Shirley McKie case – logical and statistical aspects

The Y7/Shirley McKie Case – Logic and Likelihoods

The only evidence of the act it was alleged Shirley McKie was lying about (an improper entry to a crime scene) was the same fingerprint that she denied depositing. Because of this I don’t think there was any time when it was reasonable to be confident that Ms McKie was dishonest. Even now the logic behind this observation seems to be poorly understood by criminal justice professionals.

The relevant facts of the case are described in the first 5 paragraphs here

Fingerprint identification appears to be highly reliable but it would be a serious error of logic (known as the prosecutor’s fallacy) to conclude on the basis of this reliability alone that the identified person probably deposited the fingerprint.

Explanation

We must not confuse general and specific likelihoods. The chances of winning the lottery are one in millions but if somebody claims to have won we should not assume they are lying just because, in general, a win is very unlikely.

The chances of a verified fingerprint or DNA identification being wrong could be one in millions but if an identified person denies depositing the item we should not assume they are lying just because, in general, errors are very unlikely. We can expect misidentifications to be disputed so if identifications are selected for further action just because they are disputed then errors will be picked out no matter how rare they might be. Fingerprints of people who are lying will also be picked. To reach a view about someone’s honesty in these situations a judgment must be made of the relative likelihood of the different explanations for the facts. The prosecutor’s fallacy occurs if attention gets fixed on the absolute likelihood of one explanation.

Normally if it is suspected that someone is lying when they deny depositing a fingerprint it is also suspected that they committed the crime that is under investigation. A crime brings with it the certainty that somebody committed it and every fingerprint in the crime scene has a good chance of being the perpetrator’s, who can be expected to lie. A disputed fingerprint that, after investigation, cannot plausibly be linked to a crime is in a very different situation. We would not have any specific reason to conclude that it is more likely that the fingerprint is being disputed because the identified person did something wrong in the location and is lying about it, rather than because a misidentification occurred.

In Shirley McKie’s perjury trial the disputed identification from mark Y7 was not linked to any other incriminating evidence such as a crime. This is highly abnormal if not unique in the history of forensic evidence presented in a court of law.

Steve Horn.

(Computer programmer working in the field of statistics for industry)

West Lothian

sz@hornsc.clara.co.uk

----------------------------------------------------------------------------------

Comment

Using the above logic it is questionable whether there was any time when it is was clear, prima face, that Shirley McKie had lied.

We should not assume that this logic is obvious or intuitive. The Y7/McKie case has been discussed by forensic experts and criminal justice professionals for 14 years and has been the subject of numerous inquiries and investigations. In all this time nobody from the criminal justice professions has suggested that there was anything abnormal or risky about accusing someone of lying because they challenged a forensic finding that had no connection with a crime. Nobody pointed out the abnormality to the jury at Ms McKie’s perjury trial and no guidance was given on how it might affect the probative value of the evidence.

I have discussed this with many people including police officers, lawyers and criminal justice professionals since the Scottish Parliament Inquiry in 2006. None of the professionals seemed to be aware that linking, or failing to link, disputed forensic evidence with a crime alters the likelihood that the evidence is erroneous.

Scottish appeal ruling Hamilton v HMA 1934 JC 1 is the judgement that allows a prosecution to proceed with a fingerprint as the only evidence against the accused. The evidence in that case consisted of the crime, the fingerprint and a connection between the two (Hamilton lived nearby). If this represents the minimum evidence required for prosecution then the McKie case fell below that minimum.

----------------------------------------------------------------------------------

In Bayesian terms

Using Bayes theorem the likelihood of a suspect's guilt after an item of evidence is taken into account (posterior likelihood) is dependent on the likelihood before it is taken into account (prior likelihood) and the strength of the item of evidence.

When a crime occurs somebody did it. It would be reasonable to believe that most people with a connection to the crime (even if just by proximity and opportunity) have some prior likelihood of being that person before any forensic evidence is taken into account.

When a disputed fingerprint is the only reason to believe that wrongdoing has occurred as well as identifying the perpetrator, the prior likelihood that the identified person did something that they would lie to the police about and commit perjury over could be extremely low (without the fingerprint there would be no reason to believe that anything improper or unusual had happened in the location). Even forensic evidence of a type that is generally extremely reliable may not modify the prior likelihood enough to provide confidence of guilt.

----------------------------------------------------------------------------------

I became interested in this case because it struck me while watching news clips of the Scottish Parliament inquiry into the Shirley McKie case that if fingerprint identifications generally carry a small risk of error, then a confident belief that McKie deposited Y7 could be a case of The Texas Sharpshooter Fallacy.

When looking at expert evidence we have to be aware of Transposing the Conditional

LINKS:

I think that the principle described here is important and should be clearly understood by justice professionals. I have joined discussions in the CLPEX forum for fingerprint examiners and the linked pages below are copies of some of my contributions.

Fingerprinting Innocentville is a way of looking at the above logic using a few simple “back of envelope” type calculations.

From a logical point of view using a denied elimination identification as a means of detecting a hidden act of wrongdoing has similarities to using mass medical screening to detect hidden diseases. In mass screening a medical test can be very accurate yet someone who tests positive will probably not have the disease.

There are logical similarities with the Sally Clark case (the Royal Statistical Society used the term “serious error of logic” in connection with that case).

More about relative likelihood and the McKie case.

An exercise to see how varying the error rate alters the likelihood of guilt produced some interesting results.