Natural Sciences: Immediate targeted feedback

What you'll be able to do: find out which model produced your wrong answer, not just that the answer was wrong, which in the sciences is almost always the real information.

Why the diagnosis matters more than the correction here

In mathematics a wrong answer usually comes from a wrong rule. In the sciences it usually comes from a wrong model of how the thing works, and the same model will keep producing wrong answers across topics that look unrelated.

So the feedback that matters isn't "you said the equilibrium shifts right, it shifts left." It's "you appear to be treating a catalyst as something that changes the position of equilibrium rather than the rate of approach to it", because that belief is also about to cost you marks on rates, on energy profiles, and on anything involving activation energy.

One misconception, several topics. The correction that names the model is worth five that fix instances.

The prompt

Here's a question and my answer with my reasoning. Do NOT give me the correct answer yet. Tell me: (1) the first step in my reasoning that's wrong, (2) what model of how this works I appear to be using, and (3) one question that would show me it's wrong.

Item 2 is the whole method in this subject. Item 3 is what dislodges it, being told the right model leaves the wrong one intact (article 01).

Worked examples

Chemistry: equilibrium. Student says adding a catalyst shifts equilibrium right. "Your first error is treating the catalyst as affecting position. You appear to believe a catalyst makes the forward reaction happen more, rather than both directions equally. What does it do to the reverse rate?"

Chemistry: energetics. Student says energy is released when bonds break. "You appear to hold that bonds store energy like a compressed spring. If breaking a bond released energy, what would hold the molecule together?" One question, and the model usually collapses.

Physics: mechanics. Student includes a forward "force of motion" on a coasting object. "Your free-body diagram has a force with no source. You appear to believe motion requires a force to sustain it. What's exerting that force?"

Physics: circuits. Student says the bulb nearest the battery is brightest. "You appear to believe current is consumed as it flows. What would that imply about the current returning to the battery?"

Biology: natural selection. Student writes that giraffes evolved long necks to reach leaves. "Your reasoning has the need preceding the variation. You appear to hold that organisms change in response to requirements. Where did the first long neck come from?"

Biology: osmosis. Student says water moves to where there's more solute. "That gives the right answer here and the wrong one in a different setup. You appear to be tracking solute rather than water potential. What happens if I apply pressure to the concentrated side?" This is the right-answer-wrong-model case, which is the most valuable and the hardest to catch.

Earth science: seasons. Student says summer is when Earth is closer to the Sun. "You appear to hold a distance model. What's happening in Australia at the same time?"

Astronomy: moon phases. Student says phases are Earth's shadow. "You appear to be describing an eclipse. How often would that happen, on your model?"

Notice the shape of every one: a question that makes the student's own model produce an obviously wrong prediction. That's the format to ask for.

Batch it weekly

Here are five science questions I got wrong this week with my reasoning. Don't correct them individually. What single model of how things work would explain the most of them?

In the sciences this collapses hard. Errors across chemistry and physics frequently share one root, usually something about energy, or about what "cause" means at the molecular scale.

Pitfalls

1. Asking for the correct answer. You'll get one and learn nothing about the model.
2. Accepting a correction you can't follow. Demand the case where your model fails visibly.
3. Fixing instances. Five corrected answers with the model untouched is five wasted corrections.
4. Missing the right-answer-wrong-model case. Only stating your reasoning catches it, and it's the one that bites later.
5. The tell: you have a corrected problem set and couldn't name one belief about how the world works that changed.

Try this today

Take three science questions you got wrong. Paste each with your reasoning, not just your answer.

Ask only: what model do I appear to be using? Write the three models down.

There's a good chance two of them are the same model in different clothes.