Independent Samples T-test is for testing the statistical significance of difference in means among two groups in a random variable (e.g., difference in scores between men and women). I will also use Michael Lewis-Beck's (1995) sample of 50 college students. I want to test the difference in means of Academic Ability (AA), the college entrance exam score (0 -100), between men and women. After following the top down menu as below, click "Independent-Sample T Test"
In the following pop-up window, put AA in the "Test Variable(s)" box by clicking the arrow button. And put G variable in "Grouping Variables" by clicking the arrow button. G is a gender variable that has the value of 1 for female and 0 otherwise. Now, we have to define the group. Click "Define Groups" button and put values according to the values of gender variable as below.Â
Click "Continue" of "Define Groups" pop-up & Click "OK" of Independent-Samples T Test. Then, the output widow will show you the test outcome like this.
In Group Statistics, we see that the sample has 28 non-female and 22 female students. 22 female student's average score is 72.18 and 28 male student's average score is 69.68. The independence puts the difference between these two means in a significant test. In the test, the Mean Difference (= -2.503) is a test statistics. Because the 95% confidence interval contains zero, we cannot reject the null hypothesis (the parameter value for the difference in means between men and women = 0). In other words, the gender gap in college entrance exam is not statistically significant.