How to Study STEM Subjects (When Memorisation Isn't Enough)
The worst advice you can get for a STEM exam is "just memorise it." Memorisation gets you through recall questions. It collapses completely on application questions — which are the ones that determine your grade.
STEM exams test whether you can use knowledge, not whether you can repeat it. That requires a fundamentally different study approach.
Why STEM is different
In a humanities course, understanding a concept and being able to reproduce it on an exam are relatively close together. In STEM, they're not.
You can fully understand how to balance a chemical equation and still fail five problems in a row because you haven't practiced the procedure enough to execute it under pressure. Understanding and fluency are different things — and STEM exams require both.
The method that works: worked examples then problems
The most evidence-backed approach for STEM learning is studying worked examples first, then solving problems without looking at them.
Here's the sequence:
- Read the worked example fully — understand each step and why it was taken
- Close the example and try to reproduce the solution from memory
- Check your work — not just the answer, but the reasoning
- Try a similar problem you haven't seen before
- Repeat with a new worked example
This sounds slow. It is. It's also significantly more effective than doing 50 practice problems by looking at the answer when you get stuck.
The most common STEM study mistake
Passive problem reviewing. You do a practice problem, get it wrong, look at the solution, understand why — and think you've learned it. You haven't.
Understanding a solution is not the same as being able to generate a solution. The only way to know if you can do it is to do it again, cold, on a different problem.
How to use AI tools for STEM
AI study tools are genuinely useful for STEM, but only if you use them correctly.
Use them for:
- Generating practice problems on specific topics ("give me 5 problems on enzyme kinetics")
- Explaining why a concept works the way it does ("explain why Le Chatelier's principle applies here")
- Checking your reasoning on a problem you've already attempted
Do not use them for:
- Getting answers to problems you haven't tried
- Having them explain solutions you haven't worked through yourself first
- Replacing practice with reading
The difference is whether you engage your brain before getting help, or instead of getting help.
Vera's approach to STEM
Vera generates questions at three levels: recall, understanding, and application. For STEM subjects, the application questions — the ones that give you a scenario and ask you to solve it — are the most important.
If you're consistently scoring high on recall and understanding but low on application, that's your signal. More worked examples, more cold problem solving, less re-reading.
The goal in STEM isn't to know the material. It's to be able to use it when a problem you've never seen before is sitting in front of you.