the policy statements at universities suggest that compensation ... OR AT LEAST RAISES ... are based on "merit" ... but, while this may have SOME relevance WITHIN UNITS ... it clearly does not have validity across units and everyone knows this
if so ... how could you explain the fact that liberal arts faculty at u of t (just like at penn state) ... perhaps have salaries 1/2 or less than those in the faculty of business ... at the same professorial ranks? merit can't possibly explain these kinds of salary differentials ...
now, as to the original post requesting help on doing a t test to determine if there are male and female differences in salary ... i claim that my position is no more absurd than radford's
in fact, the last post from the original poster suggests a much simpler way to deal with this matter ... since EEOC by definition says that if there is a 2 sd difference in ACTUAL salaries ... there is a problem ... they define this as "statistically significant" but ... that is totally a separate matter
line up all the salaries (male and female) ... convert to z ... then find the mean z for males and females ... and see if they differ by 2 sds ...
. . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
