On 3 Apr 2003 19:13:43 GMT, trw7atixdotnetcomdotcom (Tim Witort) wrote: > I'm developing a report in an analysis program. > This report examines employee salaries - comparing > the salaries of men to those of women in a particular > job title in a particular company. The goal is to > determine if the difference in their mean salaries > is statistically significant. > > I have been directed to the t-test to gather this > information. When I look at the t-test, however, > it appears to be geared toward *estimating* the > difference in the means of a population based on > a *sample* of the population. Since I am using > the entire population, can I still use the t-test > to determine if the difference in the means is > statistically significant? Is there another test > that should be used instead? >
Use the regular t-test. There is such a thing as correction for "finite population" but it is hardly ever appropriate. It certainly is not, in the instance *you* describe: With 100% sampled, there is no "correction"; *you* would simply conclude A is greater than B whenever A is measured as greater than B. Taking votes and ordering supplies... you might use a finite-sample correction to further those purposes, if you have a survey that is (a) incomplete; and (b) randomly chosen, for what is complete. You can use groups.google.com , advanced, to look in the sci.stat.* groups for the topic. -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
