Sensitive marking and the end of the line for the academic essay

Rather than flagging essays by students with specific learning difficulties for “sensitive” marking, why are universities not rethinking the way they assess to make it more equitable for all students?


University of Southern Denmark
25 Jul 2023
bookmark plus
  • Top of page
  • Main text
  • More on this topic
Students sitting a written exam

You may also like

Rethinking assessment in line with the changing world of work
Advice on redesigning assessment in the digital space, to test skills needed in future careers

Most UK universities include in their institutional academic policies and procedures the requirement that, when marking the work of students with Specific Learning Difficulties (SpLD) such as dyspraxia, ADHD and dyscalculia, assessors ignore so far as is possible spelling, grammar and punctuation errors. This commitment acknowledges some of the disadvantages that students with SpLD experience. These disadvantages include erratic spelling and difficulty organising ideas on paper. This approach to marking, referred to as “sensitive” or “considerate marking”, is often one component of a set of reasonable adjustments applied by universities. The submissions of students with SpLD are often flagged to ensure that assessors apply reasonable adjustments to those students who require them. This sounds like a sensible way to eradicate the inequalities faced by disadvantaged students. 

However, in practice, flagging submissions for sensitive or considerate marking can further disadvantage already disadvantaged students for a number of reasons. In our book Bias-aware Teaching, Learning and Assessment we noted that the sensitive marking flag draws the assessor’s attention to the possibility that a flagged submission might not be as academically strong as other submissions. The assessor may seek further evidence to confirm such an assumption: this is an example of confirmation bias at play.

And in a perfect world in which assessors discount all spelling or grammatical errors in flagged submissions, this consideration would only apply to students who volunteer to be assessed for SpLD or who carry certification of SpLD from a previous institution. As assessors, we will all have come across assignments rife with secretarial errors submitted by students who carry no indication of requiring sensitive marking. Some of these students should have access to additional support but have not sought this. Let’s be frank, some of these students will not warrant additional support but will still struggle with writing, specifically writing academically. 

Rather than focusing on spelling and grammar, universities tend to recommend that assessors look for evidence in the content, the ideas or the level of critical thinking demonstrated by the student to inform their judgement. Of course, many institutions’ assessment policies (including that of your institution) will explicitly acknowledge the importance of constructive alignment: that is, that the assessment task matches the module or course learning outcomes. Learning outcomes will focus on the knowledge, understanding, competencies, behaviours and attributes that the module or course designer intends each student to acquire. It is unlikely that these outcomes include consistently accurate spelling or consistently logical paragraph organisation. However, these qualities might be expected by assessors, particularly if these expectations characterise assessors’ own experiences as students.

Of course, accurate spelling and grammar can be more significant in some subject areas. For example, grammatical errors in an assignment about language or linguistics might demonstrate a misunderstanding of the subject. Some might argue that accurate spelling or grammar would be crucial on a journalism course; others might suggest that errors in journalists’ copy keeps sub-editors in gainful employment. Social science and humanities courses tend to set essay-based assignments – but why?

Ultimately, when we start to talk about overcoming the disadvantages faced by students with SpLD, we end up asking why we need to assess in this way. What is the purpose of academic writing for assessment? At a time when many are anxious about how AI can be employed to undermine the academic quality and integrity of the essay, what could we do differently? 

Alternatives to sensitive marking flags

1. Discount spelling and grammatical errors when marking all submissions. Focus on the content, the ideas, the evidence of critical thinking. Ensure that assignment requirements are aligned with module or course learning outcomes, and remove any learning outcomes relating to the standard of English from the module documentation. 

2. Offer students choice in the medium of assessment, allowing students who struggle with writing more accessible means by which to articulate their understanding. You might find that students are reluctant to choose or are suspicious that some mediums (such as essays) carry more value than others (such as presentations); this will need careful management and modelling. Choice with the medium of assessment also presents challenges in terms of word count and equivalence, but can provide dividends in lessening the marking workload and reducing the marking fatigue that can arise from encountering multiple similar written responses to a set assignment question. Students might need skills development in new areas: the essay may well have failed them, but they will still require guidance on how to succeed with a different medium.

3. Reframe the essay. If we need to retain the essay as an effective mechanism for assessment, module designers could consider how the student’s development of the essay can be acknowledged in the grade. For example, if one learning outcome requires students to demonstrate critical thinking, preparatory exercises requiring students to reflect on an initial list, a plan or a draft (perhaps in collaboration with a peer) to identify opportunities to demonstrate deeper levels of critical thinking could be graded. An approach along these lines could also be applied to enable students to engage with the implicit biases of AI. For example, students could generate a list of 10 strategies to resolve a relevant challenge (such as increasing biodiversity in town planning or reducing inequalities in conviction rates). Students then ask a chatbot to create a list of strategies to resolve the same challenge. Students would then reflect critically on the differences between the lists, with a particular focus on any bias evident in the chatbot-generated list.

3. Eradicate the academic essay. Of course, this is likely to be a longer process, partly because university quality systems manage change very slowly and partly because the essay is such an embedded component of assessment. But at a time when we’re already talking about decolonisation, perhaps there is the space to think about how a forward-looking curriculum might develop the capacity of all students to think critically and communicate in a way that is fit for purpose in the 21st century.

This advice is drawn from Andrew Read and Donna Hurford’s latest book Bias-aware Teaching, Learning and Assessment, published in 2022 by Critical Publishing.

Andrew Read is an independent educational consultant; Donna Hurford is a senior educational consultant in the SDU Centre for Teaching and Learning at the University of Southern Denmark.

If you would like advice and insight from academics and university staff delivered direct to your inbox each week, sign up for the Campus newsletter.


You may also like

sticky sign up

Register for free

and unlock a host of features on the THE site