Isn't our calculation in question 5 an acceptable answer to question 4? Isn't question 5 just asking us to convert the hypothetical effect measure into the association measure that we used in question 4?
Thank you.
Q5 and q4 are related and numerically could be the same. However, in q4, we asked you if the exposure was associated with disease and how you’d assess that with observed data. In q5, we’d like you to make certain assumptions that allow you to take that observed association to compute a causal effect. We suggest you start with the formula for a causal effect, which involves counterfactuals....think about how you go from those equations, to the conditional probabilities you used in 4. The question is more about what are the assumptions or criteria that need to be fulfilled to connect association to causation; or to conceptualize causal effects, and how to get these from observed data with certain assumptions holding. We would like you to show the formulas/notation that reflect these assumptions, name the assumptions (or criteria), and also provide the numeric answer. We recommend reviewing Part 3 of Dr. Graff’s pre-recorded lecture.
Hope that helps!
Hope that helps!