2 4 6 task You guess the secret rule < yes, fit the secret rule < yes, fits the rule Rule is ascending (increasing numbers)
Another attempt < yes, fits the rule < no < no, do not fit the rule < no < yes < no < yes < no < no < no < yes ANSWER: positive numbers, increasing by 2 Exemplar = set of 3 numbers that you generate
Positive/ negative exemplars Your own hypothesis (H) E.g. H: numbers add to 12 Proposed exemplar: Exemplar fits H “positive” exemplar E.g., if proposed exemplar: Exemplar doesn’t fit H “negative” exemplar
expectations E.g. H: numbers add to 12 Proposed exemplar : Expectation “no” response, doesn’t fit the rule If the response increases confidence in H, then it’s said to be “confirming”
Another expectation E.g. H: numbers add to 12 Proposed positive exemplar: E.g., he expected a “no” response because he had low confidence in H then the hypothesis is “disconfirmed”
Heuristics for induction and hypothesis testing Rule-of-thumb: Not guaranteed to work, but more likely to get the answer –1) form general hypotheses (not specific ones) E.g. H: numbers increasing by 2 (specific) More general version numbers increasing –2) form counterfactual hypotheses (i.e., opposing hypotheses)
More heuristics –2) form counterfactual hypotheses E.g. H: numbers increasing Counterfactual (opposing) H: numbers decreasing –3) use extreme examples (exemplars) E.g., exemplar : Extreme exemplar: -7 9,000 1 million –4) use multiple hypotheses Have at least 2 hypotheses in mind always Pick an exemplar that tells which hypothesis is right (discriminating between hypotheses)
More heuristics –5) seek disconfirming evidence find exemplars that show your hypothesis is wrong Peoples’ natural inclination is to seek confirming evidence (confirmation bias)
Science as hypothesis testing We had H in task; in science, it’s called a “theory” We had exemplars in task; in science, they are “experimental results” In task, we compared the response (yes or no) with our hypothesis; in science, the scientist compares the results to the theory