Get your stats straight

The case of Sally Clark ("Acquittal was just a chance in 73 million", THES , June 28) has interested me since the publicity given during the trial to the "1 in 73 million" odds that two children from one family could suffer cot deaths.

Ray Hill fails to see the true point of the prosecutor's fallacy - that those odds would apply to a mother picked at random . Clark was in the dock because there were two dead babies; that in itself was not evidence of guilt. The Appeal Court ruled that this argument did not affect the jury's deliberations. It criticised defence lawyers for not challenging the probability, ruling that the barrister could have dismissed it as "damned lies and statistics".

The derogatory attitude towards professional data analysis is widespread. It is assumed that anyone who has attended a basic course or has access to a personal computer is competent and that a researcher may spend years collecting data, but complete the analysis in a day.

R. Allan Reese
Chartered statistician
Graduate Research Unit
Hull University