Beyond Computer Science

Examples from Other Disciplines

Digital humanities Economics Chemistry Biology History Psychology & statistics

Nothing in coLearn-AI is specific to computing. A group works through a model and a sequence of questions, takes turns in roles, and gets the same accept or revise response from the AI, whatever the subject. The author decides what a good answer looks like and how the AI should respond to a weak one, so the expertise is the instructor’s, in their own field.

What carries over to any field Written answers, multiple choice, tables to fill in, images and figures, and optional Python for working with data. Roles, turn-taking, the AI’s focused questions, and the complete record of how each group got to its answer work the same way everywhere. The map activity in Part 2 is itself a non-CS example.

Each example below shows one question, one answer a group might submit, and the kind of question the AI sends back. As in Part 1, the AI never gives the answer: it points at one thing to reconsider.


Digital humanities

What is this passage about? Word frequency with Python

The model

A short Python program counts the words in a passage from a novel. A second version removes common “stopwords” such as the and of before counting.

The question

Run both versions. Which three words best describe what the passage is about, and why are they more useful than the top three from the first version?

coLearn-AI responds“Sea and ship look like content words. Does and tell you what the passage is about? Check whether it is in your stopword list.”

The author writes a sample answer with the expected content words, and feedback guidance: accept when all three are content words with a reason; if a function word slips through, ask about the stopword list.

Economics

Is demand elastic? Price elasticity from data

The model

A table: when a café raises the price of coffee from $4.00 to $5.00, daily sales fall from 100 cups to 90.

The question

Between these two prices, is demand elastic or inelastic? Show the calculation that supports your answer.

coLearn-AI responds“Sales did fall, but that happens for almost any good. Elasticity compares the size of two changes. By what percent did quantity change, and by what percent did price change?”

The author writes: midpoint method, quantity −10.5%, price +22.2%, elasticity about 0.47 in absolute value, so inelastic. Feedback guidance: if the group only notes that quantity fell, ask them to compare percentage changes; don’t give the numbers.

Chemistry

Which reactant runs out first? Limiting reagents

The model

The balanced equation 2 H2 + O2 → 2 H2O, and a flask that starts with 4 mol of H2 and 3 mol of O2.

The question

Which reactant is limiting, and how many moles of water can form?

coLearn-AI responds“Having fewer moles isn’t the test; the equation’s ratio is. How many moles of O2 would 4 mol of H2 need?”

The author writes: 4 mol H2 needs only 2 mol O2, so H2 is limiting, 1 mol O2 is left over, and 4 mol of water form. Feedback guidance: if the group compares raw amounts, point them to the mole ratio.

Biology

Does music help plants grow? Experimental design

The model

A student puts one plant next to a speaker playing music and another in a closet, and after three weeks reports that the music plant is taller.

The question

Name two things other than music that differ between the plants, and redesign the experiment so music is the only difference.

coLearn-AI responds“More plants would help with chance differences. But compare the two locations: what else is different between a closet and the spot next to the speaker?”

The author writes: confounds such as light, temperature, and air flow; a design with several plants per group in the same conditions, randomly assigned, with music the only difference. Feedback guidance: accept sample size only alongside at least one confound.

History

Two accounts of one event: reading primary sources

The model

Two newspaper reports of the same labor strike, one from a labor paper and one from a business paper, provided in the activity.

The question

Find one fact both reports agree on and one detail they report differently. How might each paper’s readers explain the difference?

coLearn-AI responds“Both papers have a point of view. Which specific detail do they report differently, and whom would each version persuade?”

The author writes the agreed facts and the key differences in the two sources, and feedback guidance: accept when the answer cites a specific detail from each source; if the group labels one source “biased,” ask for the evidence.

Psychology & statistics

Correlation is not causation

The model

A scatter plot of monthly ice-cream sales against drowning deaths in a coastal town, with correlation r = 0.8.

The question

Does this data show that eating ice cream increases the risk of drowning? Explain, and describe a study that could test a causal claim.

coLearn-AI responds“0.8 is a strong correlation. What else rises in the summer months that could drive both numbers?”

The author writes: a third variable (hot weather, more swimming) explains both; correlation alone can’t establish cause. Feedback guidance: if the group equates strength with causation, ask about a common cause.


Under the hood

The Same Markup in Any Field

Here is the economics question as an author writes it. The sample answer and the feedback guidance are for the AI; students see only the question and an answer box. (See the Markup page for the full language.)

\question{Between $4.00 and $5.00, is demand for coffee elastic or inelastic? Show the calculation that supports your answer.} \textresponse{4} \sampleresponses{Midpoint method: quantity changes 10/95, about 10.5%; price changes 1/4.50, about 22.2%. Elasticity is about 0.47 in absolute value, less than 1, so demand is inelastic.} \feedbackprompt{Accept if the group compares percentage changes and concludes inelastic. If they only note that quantity fell, ask them to compare the sizes of the two changes. Do not give the numbers.} \endquestion
Try it in Part 3 When you build your own activity, choose a topic from your own course. The AI Builder and the Visual Editor work the same way for any subject.