Comparing Means Among More Than Two Samples: Analysis of Variance
In chapter 11, we used the t distribution to test hypotheses about means from two independent samples. But what if we are interested in looking at more than two samples at a time? This is a common problem in criminology and criminal justice, where many important questions can be raised across a number of different samples. For example, race is a central concern in criminal justice and criminology, and often it does not make sense to restrict comparisons involving race to just two groups. Similarly, in many criminal justice studies, a number of interventions are compared simultaneously. In such studies, researchers want to compare not just two but three or more means in the context of one statistical test.