the tools students built that their university refused to

why this is worth studying

gpa decides scholarships, admission to capacity-constrained majors, graduate school, and in some cases immigration status. but grading practice varies between instructors and departments, and the gap between how objective a grade looks and how it was actually produced is wide. students live inside that gap with very little visibility into it.

hci has looked at grade data tools before, but mostly treats grades as an objective measure and mostly studies institutional tools. the ones students build for each other have gone largely unexamined — which is odd, because building one is about the most legible statement of an unmet informational need there is.

the questions

• what are the motivations, processes, and challenges behind creating these tools — for institutional and non-institutional makers alike?
• how do students actually use and perceive them?

what i did

semi-structured interviews at one large public r1 university with both sides of the artifact: the people who built the tools — institutional and student-made alike — and the students who use them. creators alone tell you what was intended; users alone tell you what was wanted. the pairing is what shows the distance between the two.

the study also follows how these tools are maintained, and what happens to them when the person keeping them alive graduates.

a tool built out of frustration is a very precise description of what the official system refused to show.

how i’m framing it

i treat the grading system itself as a data assemblage — kitchin’s term for the socio-technical amalgam of elements, practices, processes and institutional arrangements that together produce a data formation. not the data. everything around it that makes the data what it is.

grading is a particular kind of assemblage, and the two conditions are what make it worth studying:

• the primary data collection is mandated by an institution;
• there is a clear separation between the measured and the measurer.

an assemblage can be made of other assemblages, and this one is. there’s the institution-aligned part, where grades are produced, recorded and presented under institutional control — and a second, much less institutionally controlled part made of the tools themselves, built by students and alumni out of the institution’s own data. other data drifts in at the edges: ratemyprofessor reviews, course evaluations. the two are deeply interlinked rather than opposed.

the lens i use on that is conviviality — illich’s term for tools that widen autonomy instead of deepening dependence on rigid systems. the contribution i’m arguing for is treating conviviality as an affordance: not asking whether a tool is convivial, but how much of it a given tool affords, and to whom. following davis, that reframes a yes-or-no question into a design question.

this work hasn’t been through peer review yet, so i’m keeping findings off this page for now — the poster below is what i’ve presented publicly so far. happy to talk it through: get in touch.

how it connects

this grew directly out of undergraduate student critical literacy with grades, where students kept describing workarounds i hadn’t asked about. it’s also the clearest bridge to where the work is heading: peer-built tools are one way of navigating an opaque measurement system, and ai assistants are rapidly becoming another. i want to know how those differ — in what they teach people, and in whose interests they serve.