Q4 What is one advantage of using the F ratio when assessing the fit of our model?

a) It takes into consideration how many degrees of freedom were spent in order to reduce the error.

b) It tells us whether or not the model is correct.

c) It gives a better indication than the PRE about whether the model reflects the true population parameters.

d) It is the only statistic that can be used when working with a quantitative explanatory variable.

Q5 Suppose a linear model to predict Height (in) based on Weight (lbs) has a linear correlation coefficient very close to 1 (meaning the data is strongly correlated). Does this mean that increasing a person's weight will increase their height?

a) It may seem odd, but this is essentially what the linear model says.

b) No, since there is a strong linear correlation, we would not think that increasing a person's weight will increase their height.

c) No, despite a strong linear correlation between the two variables, increasing a person's weight will not increase their height. Correlation is not the same as causation - it may simply be the case that taller people tend to weigh more.

d) No, a correlation coefficient close to 1 usually indicates no relationship between the variables.