In: Statistics and Probability
True or False:
1) ____ Only experiments are useful to capture the causal effect of interest.
2) ____ The regression of y on x will necessarily yield the best predictor of y given x in a minimum mean squared error sense.
3) ____ The conditional independence assumption (CIA) asserts that selection bias is independent from any observed characteristics.
4) ____ A "proxy control" is a variable that might help control for omitted factors but is itself influenced by the variable of interest.
1)
Answer:
True
Explanation:
There are two types of studies, 1) observational and, 2) experimental.
The observational study tells the correlation between the variables such that whether the two variables are correlated while the experimental study tells the causation such that whether the one variable causes the change in another variable.
2)
Answer:
False
Explanation:
The least-square regression analysis minimizes the mean squared error to estimate the best fit line. However, it will not necessarily yield the best predictor of y given x, for example, a potential outlier data point can affect the best fit line and thus the predicted value of y for given x.
3)
Answer:
True
Explanation:
The conditional independence assumption says that the sample selection is independent of the possible outcome.
4)
Answer:
True
Explanation:
A "proxy control" variable is included in the model to control the effect of an omitted variable but they also affected by the variable of interest.