‘If you’re worried about the accuracy of your marking, you’re probably right to be’

Marking is hard. Marking open tasks such as essays and extended writing is particularly hard, because, unlike more closed tasks, the marker has to make difficult judgements about quality. Daisy Christodoulou, education author and  Tes columnist reports.

If the mark scheme tells you that the right answer for 34 x 58 is 1972, then you know what to do. But if the mark scheme tells you that the best types of answer for a piece of creative writing will use “original and varied vocabulary”, the judgements you have to make are more difficult. 

And because marking is hard, we use mental shortcuts to make it easier. Instead of only looking at the actual quality of work a pupil has produced, we inadvertently and unconsciously use other information about the pupil to help us make our judgements. When we’re marking the work of pupils in our own class, we have a lot of information about the pupil in question, which can potentially bias our judgements. 

Ethnicity bias

Perhaps the most well-explored of these biases is ethnicity. A series of studies have shown that unconscious bias creeps into teachers’ assessments of their own pupils. A study by Simon Burgess and Ellen Greaves in 2009 showed that Black Caribbean and Black African pupils were “under-assessed”: that is, that they did worse in their teacher assessments than in their tests. Indian and Chinese pupils were “over-assessed”: they did better on teacher assessments than on tests. Burgess and Greaves suggest that we carry stereotypes about the performance of different ethnic groups and these unconsciously influence our judgements of members of those groups.  

Externally-marked tests solve some of these problems because markers don’t know anything about the pupils or their backgrounds. Even with externally marked tests, though, problems can creep in. For example, when external markers know the names of the candidate, they can be biased in favour of more “attractive” first names.  

Of course, it is fairly straightforward to eliminate such bias by anonymising scripts.

The contrast effect

It’s harder to eliminate the bias caused by the contrast effect. This is when an examiner is influenced by the standards of the scripts they’ve just marked. For example, if they see an average script after seeing a series of really good ones, they will tend to mark the average script down. If they see an average script after seeing a series of poor scripts, they’ll mark the average script up. One way to try and get around this is to read a series of scripts before starting to mark. 

So, if you’re worried about the accuracy of your marking, you’re probably right to be. But you are also not alone: everyone is struggling with a similar problem. 

Read the full article ‘If you’re worried about the accuracy of your marking, you’re probably right to be’

Do you feel worried? Are you only too aware of the bias surrounding marking? Please tell us your thoughts in comments or via Twitter ~ Tamsin

Are you a trainee teacher, NQT, teacher, headteacher, parent or  just someone who cares about education and has something to get off  your chest in a Schools Improvement Guest Post? Follow this link for more details at the bottom of the page.

Don’t forget you can sign up to receive our daily email bulletin (around 7am) with all the latest schools news stories. Your details will never be given to anyone else and you can unsubscribe at any stage. Just follow this link.

We now have a Facebook page - please click to like!


Universities Minister breaks ground at the Dyson Institute
Sadiq Kahn bans fast food restaurants opening within 400 metres of London schools
Categories: Exams, Primary and Secondary.

Let us know what you think...