There is a quiet assumption embedded in how most schools operate: that if we measure students carefully enough, and frequently enough, we will eventually understand what is wrong with their learning and be able to fix it.
This assumption is backwards. And the cost of holding it is enormous.
The direction of causation
Outcomes — test scores, grades, performance on standardised assessments — are downstream of understanding. If a student does not understand something deeply, they will not perform well on tests designed to probe that understanding. This is obvious when stated plainly.
But the institutional response to poor outcomes is almost always more measurement, more testing, more data. As if by observing the downstream effect carefully enough, we will change the upstream cause.
What actually changes understanding is something harder to measure: the quality of the reasoning a student is doing. Whether they are making connections or memorising. Whether they are confused and sitting with it or confused and covering it. Whether they can transfer a concept to a new context, or only recognise it in the exact form they were taught it.
The feedback loop that isn’t
A test tells you that a student scored 62%. It does not tell you why. Was it a reading comprehension issue? A gap in prerequisite knowledge? A reasoning error they keep making? Anxiety? A concept they almost understand but not quite?
Each of these requires a different response from a teacher. But the test gives you a number, not a direction.
This means the measurement is not just neutral noise — it is actively misleading. It gives the impression of information while substituting for the harder work of actually understanding how a student thinks.
What this means for how we build
At Superslate, this is the central design question we are working with: what does it look like to give a teacher real-time visibility into how students are reasoning — not just whether they got the answer right?
The answer changes everything about what the product does and what data it collects. We are not building a better gradebook. We are trying to make thinking visible.
I do not think we have solved this. But I think we are asking the right question. And I think most EdTech is not asking it at all.
The harder question underneath
Why do schools measure outcomes instead of understanding? Partly because outcomes are easier to measure. Partly because the people who designed assessment systems were themselves shaped by the same logic. And partly because understanding what a student actually understands requires a different kind of attention — slower, more qualitative, harder to aggregate across a class of 40.
The tragedy is that this is exactly the kind of attention teaching was supposed to be about. Somewhere along the way, it got replaced by data collection. And we call that improvement.