This piece was originally published by The Justice Gap.

A new analysis of close to a thousand Court of Appeal cases over the last seven years has found that more than one in five decisions by lower courts (22%) were argued unsafe because they contained misleading evidence. The University College London study, which looked at the transcripts of 996 cases, also revealed that more than three-quarters of successful appeals (76%) were based on reinterpretations of the same materials available in the original trial rather than new information.

Scientists Nadine Smit, Ruth Morgan and David Lagnado maintained that when forensic evidence misled judges and juries, it did so because of a misinterpretation of its relevance, probative value or validity. Their paper called on lawyers and expert witnesses to bring more transparency to the relationship between evidence and hypothesis, taking care to avoid ‘an erroneous understanding of the evidential value of evidence’.

Belief in a hypothesis, the researchers conclude, is ‘sometimes presented disproportionately to the actual probative value of an item of evidence’. According to scientists at the European Forensics Genetics Network of Excellence, as The Justice Gap reported in January, it is a myth to believe that DNA evidence is ‘infallible’.

In 235 Court of Appeal cases, the UCL study found examples of misleading evidence. Over one-third (85) of these cases involved problems with the presentation of evidence in court, and in 26% (60) judges had made errors in their directions on issues like probative value, standard of proof, relevance and validity.

Although the researchers conceded that their sample size limited their scope and the conclusions they could draw, they concluded that ‘the identified cases are only the tip of the iceberg and can no longer be attributed to simple individual “bad apples” in the system’. In 2013, the Ministry of Justice admitted in response to a Law Commission report on expert evidence that ‘there is no robust estimate of the size of the problem to be tackled’.

The study calls for increased scrutiny and analysis of case transcripts, urging for improvements to ‘access to case documents post-conviction’ to avoid unnecessary repetition of past mistakes. ‘Accepting that uncertainty will remain present,’ the study explains, ‘it is important that the uncertainties are not caused by erroneous arguments and judgements but by inherent properties of the knowledge base, which, in turn, can be presented transparently’.

The study goes further to reveal that, on average, it took 2.9 years for the Court of Appeal to overrule a case. However, 18 rulings took over 10 years to be overturned. ‘The consequences of these are severe’, the report maintains, ‘and have caused many defendants to be wrongfully incarcerated.’

In the coming years, the researchers hope ‘that this research becomes more than just an “academic exercise” as these datasets grow, and that these results allow for the much-needed continuous evaluation of the use of forensic evidence in court and their contribution to both justified and unjustified rulings.’