PHI 31 Lecture 4:
Document Summary
Predictions are about what to expect of data extracted from an experiment or observation, not about the world. Negative evidence: data and prediction dont agree. Positive evidence: go to step 6 and ask if the evidence is appropriate to judge if the model fits. Does not have to explain how strong it fits, just yes or no. Inconclusive evidence: data and prediction agree, but not good enough to say the model fits. Prediction is what the data will be if the model fits. Crucial experiments: data will take you down one road or the other. One data-set tests 2 models that make distinct predictions. More ideal than reality bc the 2 models are not typically evaluated at the same time but rather historically separated. Provides evidence that one model does not fit the real world (step 5) and at the same time, evidence that the other model does fit the real world (step 6)