The Scientific Method - Homework 3 - Causal Reasoning

HINT: You may find it especially useful to read through Prof. Fisher's notes causal reasoning.

Part 1. Causation without Correlation.

Dr. Lumpemall is interested in the hypothesis that high levels of testosterone are a cause of breast cancer. She gathers a random sampling of 100,000 American adult men and women. In this sample, there is no correlation between high levels of testosterone and breast cancer. Should this convince her that high levels of testosterone are not a cause of breast cancer? If not, what further data should Dr. Lumpemall look at?

Part 2. Correlation without Causation.

Dr. Illchurch randomly sampled 20 American adults and noted that in his sample, church attendance is correlated with better health. On the basis of this correlation, Dr. Illchurch concludes that church attendance causes better health. What alternative explanations for this correlation should we consider? (We talked about three in class -- be sure to consider each.) What further sorts of evidence could Dr. Illchurch use to determine whether one of these alternative hypotheses is correct?