Looking at differences

When we want to know if differences exist between the means of continuous variables, we can use Student’s t-test (for 2 variables) or an ANOVA (for more than two variables).

To be able to validly use both of these tests, the data must meet the parametric assumptions, else you will have to use their non-parametric siblings, the Mann Whitney U and Kruskal Wallis test respectively.

In the independent samples t-test or Student’s t-test, the results of two independent samples (two results of two separate groups) are compared and the mean differences reported, using the t-statistic.

In an ANOVA, the results of groups of independent samples ( results of at least 3 separate groups) are compared and the mean differences are reported using the F-ratio statistic.

For all tests (parametric and non-parametric), significant differences only exist if the p-value is less than .05.

For more information on statistical tests, we refer you to the statistics manuals in the Recommended Reading Section of this module.