Please answer the following two questions; each question around half page long:
1. Imagine a story on the 6 o’clock news reports on a recent survey that found adults who drink a glass of red wine every night have lower blood pressure. The story also reports that the difference in average blood pressure across groups of adults who do and who do not drink every night is statistically significant. Evaluate this evidence in terms of the causal effect of drinking red wine on blood pressure. Would evidence from a randomized controlled experiment provide more compelling evidence of a causal effect of red wine on blood pressure? If so, why? Finally,list two statistical techniques that can be used to conduct hypothesis tests on experimental data and briefly comment on the similarities or differences between the two techniques.
2. Econometric models estimated using observational data are often biased. Two types of bias are particularly common: omitted variables bias (OVB) and simultaneous causality bias (SCB). For both of these types of bias, illustrate the problem by writing down a regression equation,describing the dependent and independent variables, and explaining which of the coefficient estimates is biased and why. Then, describe two econometric techniques that can be used to ameliorate OVB and SCB in your examples. (So, if you begin with OVB, write and explain an equation illustrating OVB and then discuss two possible solutions to OVB, and then write and explain a separate equation illustrating SCB, and then discuss two solutions to SCB.)
