The victims’ minister has been accused of using “factual inaccuracies” about the effectiveness of video evidence in criminal trials by one of the UK’s leading justice experts.

Vulnerable witnesses, such as children and rape victims, have been allowed to record video evidence rather than be cross-examined in person during crown court trials since a pilot in 2013, and across England and Wales since 2022.

Alex Davies-Jones MP, also the minister in charge of tackling violence against women and girls, told the Commons justice committee in a letter on 7 March that the policy, section 28, had not led to fewer guilty verdicts.

Davies-Jones based her comments on research conducted by the Ministry of Justice, which she said was the “most robust evaluation” of the use of video cross-examination since it was first piloted. She wrote: “Overall, our impact evaluation suggests the measure helps victims to give evidence earlier and avoid the stress of a live trial without reducing the chance of a conviction.” However, the MoJ findings were directly at odds with a separate, far wider, study led by Cheryl Thomas KC, professor of judicial studies at UCL and the UK’s leading expert on judges and juries.

Thomas’s analysis, which was published last year, concluded that when section 28 was applied, conviction rates were lower than when the measures were not adopted: 61% compared to 70%. In rape cases, there was a 20 percentage point difference.

Thomas’s work compared 11,798 jury verdicts in section 28 cases from 2016 to 2023 with verdicts where the measures were not used. In contrast, the MoJ research focused on a three-year period, examined 2,169 verdicts and its method, according to Thomas, “inflated” the number of convictions and “under-counted” acquittals.

In a letter to the committee last week, Thomas said the MoJ’s approach was “misleading, narrow and bureaucratic” and claimed Davies-Jones had got her facts wrong six times. “Vulnerable witnesses need accurate information about special measures in order to make important decisions about how to give their evidence in criminal cases. It is concerning that these factual inaccuracies are in an official letter,” she wrote.

Mary Prior KC, chair of the Criminal Bar Association, said Thomas’s findings “mirror perceptions” among barristers, and described the MoJ study as being “spun as proof that the policy is the way forward”.

In a message to her members, Prior said: “The MoJ has produced a piece of statistical analysis on the use of S28 hearings which it is using to pretend that all is well for victims who are advised to have their cross-examination pre-recorded in advance of trial.”

Pre-recorded cross-examination is one of a series of “special measures” that judges can authorise to lessen the courtroom ordeal for victims and witnesses, and reduce the time they have to wait to testify. Under section 28, evidence from vulnerable witnesses is recorded in a closed courtroom before the trial starts and is played to the jury when the proceedings are under way.

Speaking to the Observer, Thomas said the MoJ had pushed ahead with the video evidence policy, so the results of her study, showing a drop in conviction rates, had been “problematic” for the department.

“I can only assume they decided to do their own analysis because the results were not what they would like to see from a policy that had that level of prominence,” she said.

Thomas is conducting further research to understand why conviction rates are lower in these cases, but believes one possible factor is that jurors have less empathy when witnesses appear on a screen than in person.

Prior said there was “no point” promoting such measures if victims still faced long delays. “For S28 to be increased in use, it requires significant investment in the reduction of the court backlog, in the criminal barristers who undertake this work and in the judiciary who preside over these cases,” she said.

An MoJ spokesperson said: “Pre-recorded cross-examination helps vulnerable and intimidated victims give evidence in court … The findings from our evaluation have been peer-reviewed and show there were benefits for victims from pre-recording evidence during the period analysed.”