• If we had data from a randomized controlled experiment, then why can the difference in means across the two groups (treatment and control) serve as an estimator for the treatment effect? That is, why can we say it's an estimate of a causal relationship?
• Suppose you ran a bivariate regression to assess the relationship between age (your x variable) and average weekly earnings and got your estimators for beta0 and beta1. What sort of information would you look at to interpret whether your estimates were "economically significant"? That is, assuming you guessed correctly, how would you try to find out if the relationship was meaningful in the real world? Be specific about what objects you'd look at in your data, what you'd look for on a graph, etc.