From the course: SPSS for Academic Research

Unlock the full course today

Join today to access over 22,400 courses taught by industry experts or purchase this course individually.

Assumptions and hypotheses

Assumptions and hypotheses - SPSS Statistics Tutorial

From the course: SPSS for Academic Research

Start my 1-month free trial

Assumptions and hypotheses

- [Instructor] A lot of the assumptions for this test are going to look similar to the other tests. For example, you can only have one dependent variable and it must be continuous. Ours is salary and that fits the bill. Second, the observations must be independent of each other. There should be no relationships between the participants in any of the groups. Third, you should try to get a balanced design that's equal sample sizes for each level. This very rarely happens in real life, so don't worry too much if you violate this assumption. There are ways to control for it. Fourth, the variance of the dependent variable residuals should be approximately equal. This is commonly known as maintaining homogeneity of variance. We can determine if this assumption is met using Levene's test of equality. Finally, there should be no outliers in any cell of the design. That can be checked using boxplots. You'll have three null hypotheses, and as such, three alternative hypotheses that are the…

Contents