Examiners disagreed with each other on the marks they gave nearly three quarters of the time during a recent trial of exam papers.
And according to a report from Ofqual, monitors did not agree with the scores awarded to papers marked a second time by their original examiner 60 per cent of the time.
The trial saw senior examiners investigating the marks given to 173 exam papers, covering GCSEs in English literature and Maths, and AS-level in Biology, sat last year. Incredibly, they only agreed with the initial mark in 28 per cent of cases.
Twelve per cent of the time, the moderators were unable to agree on a final mark amongst themselves.
Most of these disagreements were in English literature. In total, about 19 re-marked scripts for each of nine units across the three subjects were analysed.
Almost half (43.8 per cent) of all scripts sent back for review by schools and colleges in 2016 had their marks changed.
The report found that the majority of review examiners were found to be sticking to Ofqual’s new changes to review marking introduced in 2016, which stipulated that they can only correct a marking error, rather than give a new mark based on their own judgement.
The change was introduced because Ofqual found examiners were more generous with an exam script if it had been submitted for a re-mark.
For 93 per cent of the sample exam papers, review examiners were found to be re-marking correctly by only correcting a marking error.
In a “small number of cases”, less than seven per cent, a change was made even though there was no error in the original marking, and in a “very small number of cases” errors were not corrected.
Sally Collier, the chief regulator at Ofqual, said it was “pleasing” to see the new rules being implemented as “intended” last summer.
The system is still in a period of “adjustment”, so Ofqual is “working with exam boards to identify what can be done this year to be even more confident that students are getting the results their performance deserves”, she added.
The reviewed exam papers in the sample Ofqual analysed included those papers with the largest increase in marks, the largest decrease in marks, as well as some which experienced no change in marks.