Exercise 7.22 introduced a study conducted about antibiotic consumption in China. One aim of the study was to develop a prediction model for predicting monthly average antibiotic consumption based on county-level data. The following plots show the association between monthly average antibiotic consumption and four potential predictor variables: proportion female inhabitants (female), average life expectancy in years (lifeexp), proportion of population illiterate (illiterate), and population density in 1,000 people \(/ \mathrm{km}^{2}\) (popdensity).

(a) Summarize what you see.

(b) Identify any predictor variables that might benefit from a natural log transformation and briefly justify your choices.

Data From Exercise 7.22

40 40 10 Consumption (DID) 20 30 30- 0 Consumption vs Female 10 40 Consumption (DID) 30 20- 10- Consumption vs Life Expectancy 0 T T T T 46 48 50 52 54 72 73 74 75 76 77 78 Percentage Female Pop Life Expectancy (yrs)