Natural Sciences: Misconception elicitation

What you'll be able to do: Find out which documented wrong models you're carrying, before you spend a term building on them.

Why this is the first method for the sciences

The natural sciences have something no other subject group has: a well-documented, stable, near-universal catalogue of wrong models. Decades of education research have mapped them, and the same ones appear in students across countries and curricula.

That's an enormous advantage and almost nobody uses it. You don't have to discover what you might believe wrongly. You can go and look it up, and then test yourself for it specifically.

The reason this matters more here than elsewhere is that these models are occupied, not empty. A gap announces itself, you know you don't know. A misconception produces confident, coherent, wrong answers and feels exactly like knowledge. Ordinary revision does not touch it, because ordinary questions don't discriminate between the right model and the wrong one.

The catalogue worth testing yourself against

Physics - A moving object has a force in its direction of motion. - Current is used up as it goes round a circuit. - Heavier objects fall faster. - Heat and temperature are the same quantity. - There's no gravity in space. - Action and reaction cancel out, so nothing accelerates.

Chemistry - Atoms "want" full shells. - A catalyst shifts the position of equilibrium. - Electrons orbit the nucleus like planets. - Bonds "store" energy and release it when broken. - Mass is lost when something burns. - Equilibrium means the reaction has stopped.

Biology - Organisms change in order to adapt. - Evolution is progress toward something better. - Plants take their mass from the soil. - Genes directly cause traits, one for one. - Blood is blue inside the body. - A theory is a guess.

Earth science and astronomy - Seasons are caused by distance from the sun. - Moon phases are Earth's shadow. - Plates float on liquid magma. - Weather and climate are the same claim at different scales.

Several of these (bonds store energy, atoms want, evolution is progress) are actively taught in simplified form at school and then have to be dismantled, That's a simplification-to-unlearn (simplify, then restore, the textbook-versus-literature gap), not a mistake you made.

How to test for them

What are the documented common misconceptions about [TOPIC]? For each, what wrong model produces it, and what does that model predict that the correct one doesn't?

Then:

Give me eight questions designed to detect those misconceptions, each should give a DIFFERENT answer depending on which model I'm using. Don't tell me which misconception each targets. I'll answer fast, without working it out.

Answer fast. This is the rule that makes it work. Given time you'll construct the correct answer; the fast answer is the model you actually run, and the exam is a fast-answer environment.

Then:

Don't just mark my answers. Tell me which wrong MODEL each of my wrong answers is consistent with.

Dislodging one, which is the hard part

Being told the right answer does not remove a misconception. The wrong model was never activated, so it's still there, and it will run again the moment the question is unfamiliar.

What works is a case where your model gives an obviously wrong answer and you can see it yourself:

I appear to hold [misconception]. Give me one concrete case where my model gives an answer that's obviously wrong in a way I can check.

Bonds store energy dies when you're asked why breaking bonds in an exothermic reaction doesn't release the energy. Current is used up dies when you measure the same current either side of a bulb. Seasons are caused by distance dies when you're asked why the southern hemisphere has summer at the same time.

Each of those takes one question and does what a chapter doesn't.

Pitfalls

1. Reasoning carefully before answering. You'll construct the right answer and never find out what you run under pressure.
2. Marking instead of diagnosing. "Wrong" tells you nothing; the model does.
3. Correcting without dislodging. Demand the killing case.
4. Testing once. These come back, especially when tired or rushed.
5. Assuming you're exempt. These are near-universal, including among people who teach the subject.
6. The tell: you get them all right, slowly.

Try this today

Pick the topic you're studying this week. Ask what its documented misconceptions are, then for eight discriminating questions with the targets hidden.

Answer each in under ten seconds. The ones you get wrong fast are the models you're actually carrying, and they were going to cost you marks on any unfamiliar question.