After referring yesterday to the use of statistics in the public debate surrounding climate change, I learn today of the death by statistics of Sally Clark, a 42 year-old lawyer who in 1999 was sentenced to life imprisonment for the murder of her two children: Christopher, aged 11 weeks, and Harry, eight weeks. Clark spent three years in jail before her conviction was quashed by the Court of Appeal.
Sally Clark’s death is now in the hands of the coroner, and we do not know for certain exactly how she died. But the implication behind the news reports circulating today is clear enough. Clark’s family have said that Sally “never fully recovered” from the effects of the “appalling miscarriage of justice” she suffered.
Clark’s conviction rested on the supposedly expert testimony of paediatrician Professor Sir Roy Meadow, who was called upon often by the courts to provide expert scientific opinion, and is best known for identifying the (now controversial) condition commonly called Munchausen Syndrome by Proxy.
In the case of the Crown vs. Sally Clark, Meadow took the then accepted figure for the chance of a single cot death – 1/8543 – and squared this to calculate the probability of two cot deaths occurring in the same family. The figure Meadow gave the court was 1 in 73 million, which makes it look as if Clark was guilty as sin.
The flaw in Meadow’s analysis should be obvious to anyone with even the most basic knowledge of statistics, and should certainly have been apparent to Meadow himself.
Multiplying two probabilities together is valid only if the two events are totally unrelated, but one cannot make this assumption in the case of cot deaths. The chances of two such deaths occurring could be linked by genetic, environmental and other factors, and when these factors are taken into account, it turns out that the probability of two cot deaths in the same family is 1 in 200. This is the figure arrived at by a number of independent experts following Sally Clark’s conviction.
When Meadow’s scientific incompetence came to light, and the Court of Appeal heard previously suppressed evidence that Harry Clark had been suffering from a brain infection, Sally Clark was set free and all eyes turned to Roy Meadow. Following an investigation by the General Medical Council, Meadow was found guilty of serious professional misconduct, and struck off the medical register.
Meadow appealed to the High Court, was cleared of the most serious charges, and reinstated. The reasons for the judge’s decision are not entirely clear, but, at the time, the Society of Expert Witnesses had expressed fears that hanging Meadow out to dry would lead to many witnesses withdrawing their services, depriving the legal system of much-needed expert counsel. The latest twist is that in October of last year, the Appeal Court overturned the High Court decision against the disciplining of expert witnesses, but upheld Meadow’s reinstatement by a 2:1 majority decision.
To my knowledge, Meadow never expressed remorse for his false testimony and the human damage he caused. Not only to Sally Clark, but to others such as Angela Cannings who were also wrongly convicted of murdering their children. Throughout this whole sorry saga, Meadow’s only concern has been to protect his personal reputation.
It is said that “There are lies, damned lies, and statistics.”. I don’t agree. Statistics are just numbers. They are completely value-free, and, when misused, entirely information-free. Statistics help us to make sense of data and assess their significance, but this is true only when the scientific assumptions behind the analyses are made clear, and the questions asked of the data properly framed. The assumptions made by Roy Meadow were hidden, and the figure of 73 million-to-one handed down from on-high. The number has about as much significance as 42 being the answer to the meaning of life.
At the time of her release from prison, Sally Clark said: “Today is not a victory. We are not victorious. There are no winners here. We have all lost out.”