After you identify a significant Gender effect, a significant interaction between Gender and Courtesy or Gender and Payment, or a three-way interaction between Gender and Courtesy and Payment, the next step is to run some additional analyses to see what's going on.

Main effect of Gender

Analyzing the EF_FRND and EM_FRND variables (how much you'd like to have the female or male characters as friends), we get the following output:

The first row tells us that there is a significant main effect for Gender: ratings of how much people wanted the man to be their friend were significantly different from ratings of how much people wanted the woman to be their friend. But which one was higher?

To answer this question, we need to set up the same analysis we did before but with one change. First, select Analyze -> General Linear Model -> Repeated Measures. If you don't have the box filled in from last time, you'll need to set up the within-subjects factor:

Press "Add" and then "Define". Make sure the within-subjects and between-subjects variables are set up as they were with the first analysis:

Instead of pressing OK to run the analysis, look for a button toward the bottom of the screen above that says Options:

Pressing this button brings up the following menu:

Move the significant effect from the left side window over to the right side window labeled "Display Means for:" Then press Continue to return to the previous screen, and OK to run the analysis.

The output will basically be the same, but there will be a new window at the very bottom:

This window presents the means for the two levels of gender and will allow us to answer the question of "which character was rated higher." One problem is that Gender is labeled just "1" or "2". To determine what it means by 1 and 2, we need to scroll up to the very beginning of the analysis:

Just under the heading "General Linear Model" (which is the beginning of this analysis output), SPSS tells you the identity of the Gender variables: 1 is for the female character, 2 is for the male character. Looking at the Estimated Marginal Means table that we just got, this tells us that the female character was generally rated as being more likely to be considered a friend than the male character. To record this information, open a Word document and label the first section according to the variables you're comparing:

To report the main effect of gender, go to the SPSS output window, select one of the tables we just looked at, press CTRL+C (or select Edit -> Copy) and then go to the Word document and press CTRL+V (or select Edit -> Paste). Do this for both the table showing what Gender 1 and Gender 2 refer to, and the Estimated Marginal Means:

So that's how we'll record main effects for gender. Next we'll look at what to do about significant interactions between Gender and Courtesy or Gender and Payment.