Most of us can spot the student who is waiting to be given the answer. Perry's scheme helps explain what is going on, and being able to measure it turns a hunch into something you can act on. This page sets out the research Greg Stoner and I have done on measuring intellectual development from students' everyday written work, and the training I am setting up so that you can do the same with your own students.

The research

Our paper in Accounting Education asks how first-year students' development towards critical thinking can be measured from the coursework they already write, and what the patterns of development tell us.

Rose, J. and Stoner, G. (forthcoming) Enabling and measuring engaged first year students' intellectual development towards contextually evaluating knowledge and critical thinking. Accounting Education. https://doi.org/10.1080/09639284.2026.2748999

Measuring development from coursework

The usual ways of placing students on Perry's scheme are the Perry interview, the Measure of Intellectual Development (MID) and the Learning Environment Preferences questionnaire. All three are useful, but they ask students to do something extra on top of their studies, the first two take a lot of time, and questionnaires invite students to tell us what they think we want to hear.

With input from the Perry Institute, we developed the Measure of Intellectual Development in Written Work, or MID-WW. It adapts the MID so that a student's position can be read from their ordinary coursework. For our module we built a set of indicators for positions two to five, one set for academic writing such as essay plans and essays and another for self-feedback. Following MID practice, each piece of work receives three ratings. A rating of 2,2,3 means dualism with glimpses of multiplicity, and 4,5,5 means contextual relativism with traces of pure relativism. That detail matters because it shows you the students who are close to a shift as well as those who have made it.

I rated every piece of work as a qualified Perry assessor, talked through the difficult cases with another qualified assessor and the Perry Institute, and Greg moderated the approach.

What we found

We followed the 80 students, from a first-year accounting cohort of 512, who completed all five voluntary formative tasks during their first eight weeks at university. The tasks were an essay plan, group self-feedback on exemplar essays, a draft essay, individual self-feedback comparing their essay with their peers' work, and a redraft.

• 86% moved up at least one position on Perry's scheme in at least one task.

• At the start, 14% showed even glimpses of pure relativism, the position where critical thinking begins to appear. By the end, 74% had shown at least glimpses of it and 48% had reached it or gone further.

• Ten students who began at multiplicity reached contextual relativism, the highest position usually seen in undergraduates.

• 55% reached a higher position in their self-feedback than in their essays, and only 9% the other way round.

Progress was not a straight line. Some students reached a higher position in their self-feedback and then dropped back in the next essay, which is what Perry would lead us to expect. The finding I keep coming back to is the gap between self-feedback and essays. It suggests that the way essays are usually set may hide what students are capable of, and that tasks asking students to explain their own judgements about quality give them more room to show it.

These were engaged students who chose to do all the formative work, and they did better in the final exam than those who did not, so they are not a typical cohort. The paper is open about that, and about the indicators being specific to this module rather than a ready-made tool.

Using the approach with your own students

MID-WW is a framework rather than a fixed questionnaire, so you build indicators that fit your own subject and tasks. A sequence of written tasks in which students have to explain their reasoning is a good place to start, ideally including self-feedback on exemplars or on peers' work, because that is where development shows most clearly. The paper sets out our indicators in full and walks through three students' work, so you can see the reasoning behind each rating. You can read more about self-feedback on the active self-feedback page.

If you plan to publish what you find, you will need ethical approval, and students should be able to opt out of having their work analysed.

I would love to hear from anyone who wants to try this in another discipline or another university. The paper calls for more trained raters and for testing the approach in other settings, and that is much easier to do together. Get in touch