problem-7.32

problem-7.32  The AGE variable is symmetric, so no transformation is needed.
> hist(cfb$AGE)                 # symmetric
> wilcox.test(cfb$AGE, conf.int=TRUE)
...
95 percent confidence interval:
 48 50
...
  
The INCOME variable is skewed. We do a log-transform first.
> hist(log(cfb$INCOME +1))
> wilcox.test((log(cfb$INCOME +1)), conf.int=TRUE)
...
95 percent confidence interval:
 10.49 10.61
...
> exp(c(10.49,10.61)) - 1
[1] 35953 40537