problem-4.15

problem-4.15  The boxplots are easily made with the formula interface:
> boxplot(attendance ~ year, data = MLBattend)
    
Looking at the drops in attendance shows that 1972, 1981, and 1994 are likely candidates.