Chapter 12 SPSS^®

In this chapter we conducted a one-way between-groups ANOVA to compare people in four different types of societies in terms of how fairly they behaved in a game, as assessed by the proportion of money that they gave to a second player in that game. The type of society was a nominal independent variable, and the proportion of money that they gave to the second player was a scale-dependent variable. To conduct a one-way between-groups ANOVA using SPSS, we enter the data so that each participant has one row with all of her or his data. For example, a person would have a score in the first column indicating the type of society, perhaps a 1 for foraging or a 3 for natural resources, and a score in the second column indicating fairness level, the proportion of money he or she gave to a second player. The data as they should be entered are visible behind the output on the screenshot shown here.

Then we can instruct SPSS to conduct the ANOVA by selecting Analyze → Compare Means → One-Way ANOVA. Now select the variables from the list. The independent variable, named “society” here, goes in the box marked “Factor,” and the dependent variable, named “fairness” here, goes in the box labeled “Dependent List.” To request a post hoc test to compare the means of the four groups, select “Post Hoc,” then “Tukey,” and then click “Continue.” Click “OK” to run the ANOVA.

On the screenshot, we can see the source table near the top. Notice that the sums of squares, degrees of freedom, mean squares, and F statistic match the ones we calculated earlier. Any slight differences are due to differences in rounding decisions. The last column, titled “Sig.,” says .006. This number indicates that the actual p value of this test statistic is just 0.006, which is less than the 0.05 p level typically used in hypothesis testing and an indication that we can reject the null hypothesis. Below the source table, we can see the output for the post hoc test. Mean differences with an asterisk are statistically significant at a p level of 0.05. The output here matches the post hoc test we conducted earlier.

SPSS Statistics