1. State whether each of the following statements is true or false. For the statements that are false, explain why they are false.
2. Suppose an untransformed regression model satisfies all but one LINE assumption. A transformation of the response variable results in a model that satisfies the assumption that failed to be satisfied by the untransformed regression model. Since the untransformed regression model satisfied the other assumptions, there is no need to re-check these assumptions for the model with the transformed response variable.
3. Suppose we've fit a simple linear regression model with response variable Y and predictor variable ln(X) and obtained a prediction interval for Y when X = 5. To interpret the interval in the measurement units of Y we need to back-transform the prediction interval by exponentiating the interval bounds.
4. Suppose we've fit a simple linear regression model with response variable ln(Y) and predictor variable X and obtained a prediction interval for ln(Y) when X = 5. To interpret the interval in the measurement units of Y we need to back-transform the prediction interval by exponentiating the interval bounds.
5. The hierarchy principle in regression modeling states that if a fitted model includes a statistically significant interaction term, then the model should also include the individual predictors that make up the interaction term, whether or not the coefficients for these predictors are significant.

An appropriate regression modeling strategy is to continue adding higher and higher powers of X as predictor terms in a polynomial regression model until R² reaches 90%.