Overview:
T-tests are the perhaps one of the most simple types of statistical tests. The goal here is simply determining if one mean is statistically different than another mean.
When to use an independent samples t-test:
You perform an independent samples t-test (also known as a between subjects t-test) when comparing two continuous, unpaired means. Means are unpaired when the scores in one group do not effect the scores in another group, or when the observations in group 1 are unrelated to those in group 2. This assumption would be violated if, for example, you were looking at before- and after- scores, or if the same individual were giving measurements of two different outcomes (for example, if you were measuring preferences for chocolate vs. vanilla ice cream, but each individual sampled both flavors).
Running an independent samples t-test:
From the main toolbar, click Analyze
Click Compare Means
Click Independent-Samples T-Test…
Next, move the dependent variables from the variable list on the left to the Test Variable(s) window using the arrow.
Move female to the Grouping Variable box.
Click Define Groups… In this case, we specify the groups as 0 and 1, since our female variable lists males as 0 and females as 1.
Click Continue
Click OK to run the analysis and generate the output.
Interpreting the t-test output:
- Levene’s Test for Equality of Variances (Sig) tells us that the means in our sample meets a basic parametric assumption, that the variances of each of our samples are similar to one another. However, Levene’s test is sensitive to several other parametric concerns so we often ignore it. If it is significant though, we might want to consider sampling more, or simply proceed to the remaining interpretations with caution.
- The Sig. (2-tailed) in the t-test for Equality of Means section provides the actual test for the means, by testing the null hypothesis that the means are equal. At alpha = .05, our result is significant, p < .001.
Reporting the t-test results:
A between subjects t-test indicated that males (M = 13.64, SD = 7.01) scored significant higher in their math scores than females (M = 11.95, SD = 6.65), t(7183) = 10.48, p < .001.