Answers to Practice Questions for Final:
Part IV, V: INTERPRETING SPSS OUTPUT
IV: Under group statistics, circle the Ns for the two groups (a small n study, so limited power) and the means. Make a note of which mean is higher (FTF) so you'll know what direction the difference is in if the test is significant. Next part, circle the significance level. NOT significant at .05 level, which is GOOD news -- in this case you are hoping that the null hypotheses of "equal variances" is NOT rejected. Note in margin something like "NS -- passed test" or "Good--equal variances assumption okay"
Now you know which set of results to look at for the actual t-test (by default, SPSS always runs it both ways, for equal variances assumed or not). Mark which row we should look at. Circle the t, df, and significance, checking to see if the test came out significant, and if so, at what level. Write "NS" next to the significance level. If the test WAS significant, I would put a star next to it or write "p < .05 ... or .01 or .001 or however it turned out [there's not a "fixed" way to annotate -- what's important is that you draw attention to the key information here]
There was no significant difference between the two groups, using a t-test for independent samples, t (28) = .227, NS. [taking info from output and turning into APA notation] The FTF groups did have higher performance scores overall than CMC groups (mean of 63.6, SD 13.15 versus 57.7, SD 13.17) but this difference was not significant at the .05 level. This was a small N study, however, (14 FTF groups, 16 CMC groups) so the power is low.
[Your paragraph might be worded differently: the important thing here is that it should convey all the key information about the results of the test]
V: Under descriptives, circle the Ns and the means. Ns not equal in each sample, note the ordering of means (Protestant > Catholic > Muslim). Test of homogeneity ok (circle sig., note "passed test). For ANOVA table, circle F and sig. Note that the F is significant at .05, .01 level.
Look at Tukey results to see which means are different. Muslim and Catholic are in the same subset, Protestant in a different subset, so Protestant is different from the other two, but Muslim, Catholic not significantly different at .05 level.
The ANOVA shows a significant difference in urbanization based on the main religion of the country, F (2, 79) = 5.71, p < .01. Tukey posthoc comparisons show that Protestant countries were more urbanized (M = 72.63 percent living in cities, SD = 16.03) than either Catholic (M = 56.13, SD = 23.06) or Muslim (M = 48.70, SD = 24.84) countries. Muslim and Catholic countries did not differ in urbanization rates at the .05 level.