The report from the Department of Labor, Women's Jobs 1964-1996: More Than 30 Years of Progress, provides a graphic, statistical overview of the progress made for women in the workplace in the United States. Even though progress has been made, you can examine this data showing the differences in pay for men and women over the past several years. The table shows that progress has been made and the gap does appear to be closing, although much remains to be done.

The data examined here was recorded by the AFL-CIO. They provide the figures for pay differences for different occupations. Given the data, do you think a wage gap exists? How should you go about examining this data? You want to test whether the average pay for men is equal to the average pay rate for women over the 65 occupations. What kind of test should you use?

The question here is whether to use a paired t-test or a test for independent samples. The statistics for both tests are as follows:

Variable	N	Mean	Std. Dev
women	65	494.34	194.88
men	65	622.55	262.21
gap	65	128.22	91.40

Now, if you use the t-test for independent samples, the associated t value is -3.1641. If you use the paired t-test (testing that average for the variable 'gap' is zero), the t value is 11.31.
Note that you can use the sites in the following Analysis section to check these values.

Although both t values are significant, (is the pay gap real?) there is a large difference between the tests. Which one is correct?