problem-8.18
problem-8.18
The log() function can be used directly, as in
> wilcox.test(log(exec.pay), mu = log(22), alt="greater")
Wilcoxon signed rank test with continuity correction
data: log(exec.pay)
V = 10966, p-value = 0.004144
alternative hypothesis: true mu is greater than 3.091
We see a similarly small p-value as in the previous question,
indicating that the null hypothesis is not supported.
If you tried to do a histogram to check for symmetry, you may have
gotten an error message. This is because of the data values for
which the compensation is 0. (The logarithm of 0 is treated as
-¥.) To avoid this error, you can exclude this case first
as follows:
> ep = exec.pay[exec.pay > 0]
> hist(log(ep)) # fairly symmetric