Online Marking Tools Explained: On-Screen Marking, Comparative Judgement and AI Marking

"Online marking" covers three different ideas: on-screen marking, comparative judgement and AI marking. They answer different questions, so the right tool depends on what you are trying to mark. This guide explains each in plain terms and says where GradeDrive fits.

The short version

On-screen marking

In on-screen marking, scripts are scanned and sent electronically to examiners, who mark them on a screen. Ofqual describes the main benefit as better quality assurance while markers are actually marking: boards can check examiners against pre-marked "seed" items and send some scripts to a second examiner. Scripts also make only one journey from the school to the scanning centre.

Ofqual describes exam boards using it for national exams. It is built around trained examiners marking at scale, so it answers a different problem from a teacher with a pile of mock papers.

Comparative judgement

Comparative judgement does not use a mark scheme. Instead, judges look at two pieces of work side by side and decide which is better. After many comparisons, the results are modelled statistically and every piece of work is placed on a scale of relative quality.

No More Marking is one example used in UK schools. Its own page describes the approach as suited to open-ended work and gives writing as the example, and says research finds it as reliable as double marking but much quicker. The teacher sets a task, uploads the scripts and judges pairs on screen, and the school gets a report to benchmark attainment and progress.

It suits work where there is no single right answer to check against a mark scheme. The output is a ranking, so it will not give you marks against a scheme.

AI marking

AI marking reads a student's answer, checks it against the mark scheme a teacher uploads, and suggests a mark for each criterion. A teacher reviews every suggestion before it counts.

GradeDrive is one example. You scan the class set into one PDF, upload it with the mark scheme, review each question with the script on one side and the criteria on the other, and export marks, feedback sheets and a class analysis.

Picking the right tool for the job

• Ranking open-ended writing. Comparative judgement was built for that; GradeDrive marks against a scheme instead.
• GCSE and A-level Biology, Chemistry, Physics and Maths. That is what GradeDrive officially supports today, with English still experimental.
• National, high-stakes marking at scale. That is the job of the exam boards and their on-screen marking.

When it fits

You have mark-scheme papers in the four supported subjects, a class set to get through, and you want to review the marks and not produce them from scratch. For a sense of what teachers do with the suggestions once they see them: 99.2% were left unchanged last summer, measured from our own usage rather than an outside audit. See the accuracy page for the methodology.

Common questions

Is online marking the same as AI marking? No. Online marking is a broad term. AI marking is one kind of it.

Can these be combined? A school can use more than one. For example, comparative judgement for extended writing and AI marking for science mock papers.

What should I check before choosing a tool? What it outputs (marks or a ranking), which subjects it supports, and who decides the final mark.

Sources

• Ofqual: Exam marking, how technology is improving the quality of marking (20 April 2018)
• No More Marking: Comparative Judgement

Related reading

• Exam marking software for UK schools
• What AI marking means for GCSE and A-level
• How the GradeDrive workflow runs