Exam marking reliability is an issue that crops up regularly as analysts try to decipher the intricacies of exam results and what they mean for attainment and exam difficulty.
This has been complicated in recent years by the alterations to the GCSE system that have altered the way league tables are compiled and therefore make it harder for comparisons to be made.
However, in terms of exam marking, plenty of progress has been made over the years to eliminate bias from the process, which should lessen concerns about bias in formal exam marking.
Increase in on-screen marking has improved reliability
There has been a deliberate move in recent years away from traditional paper marking to on-screen marking for national tests.
Often, this involves papers being trimmed down into different items and then marked by different examiners. Other papers, especially those with essay-style responses, are marked by one examiner, but double marking is sometimes used to check quality.
The regulator Ofqual explained the benefits of on-screen marking in a detailed blog post about the process, highlighting that:
Each student may get a single final mark per qualification, but for most subjects it will have come about through the combined effort, experience and dedication of many different examiners.
This necessarily removes bias from individual papers and seemingly improves the overall quality of the marking process. Anonymising exam scripts removes further elements of bias, especially related to gender.
An analysis of higher education anonymised marking changes published in 2017, concluded that:
There seems to have been a very slight narrowing of gender differences since implementation of anonymous marking. Additionally, there do appear to be larger gender differences post-implementation in exam marks than in coursework, though these differences favour female students.
Whether these findings could be extrapolated to GCSE anonymisation is unclear, but, taken with the rise in on-screen marking, it seems to eliminate much potential for gender bias.
Regional biases in context
On-screen marking has also significantly improved the prospects of eliminating regional bias because exam papers are sent electronically rather than transported in the traditional sense.
Theoretically, this could lead to papers being marked by examiners anywhere, as Pearson Edexcel admitted in 2013 when they said that some of their simple papers were marked in Melbourne.
While the potential for different types of problems remain, the very fact that exam papers can be anonymously marked by an examiner in any location seems to reduce the possibility of regional bias.
Gender and regional divides
What remains clear from 2018 GCSE results is that there are distinctive divides based on gender and location.
Girls are still more likely to receive top marks, although boys are narrowing the gap. Equally, another analysis demonstrates that there is a 4% difference in the likelihood of students in the South receiving higher grades than their counterparts in the North.
So, while marking is switching to on-screen anonymity, the wider problems of gender and regional divides remain fairly constant.