The t test is limited when it comes to social research and best used when you have two groups (men and women) to compare. But what if you have many groups, such as people of varying religious backgrounds or beliefs, or even people with different educational backgrounds? We would want to compare, not only the relationship between variables, but also the relationships between the groups (Catholics and Jews, Buddhists and Methodists, Jews and Buddhists, Methodists and Catholics, etc.). One advantage of the ANOVA is that it examines the means, as well as the variances, of the subgroups and eliminates "individual differences" as a source of variability.
Last week you purposefully used GENDER as a variable to see a t test for understanding. This week you will need to choose a variable that has multiple subgroups to use for the ANOVA test of significance. Some examples are religious backgrounds, level of education, SEI, etc. (If you have questions about this, reach out to me ASAP for clarification). See how these variables are related to your variable...
Here are the steps for SPSS:
1. Select Analyze
2. Compare Means
3. One-Way ANOVA
4. Set your variables- your dependent variable goes into the 'Dependent List' box and your Independent Variable goes into the 'Factor' box.
5. Click Options and select 'Descriptive' and 'Continue'
6. Click 'OK'
Reading your output:
Use the steps for hypothesis testing to determine if the relationship between your variables is statistically significant.
1. Restate the question as a research hypothesis and a null hypothesis about the population.
2. Set your significance level at .05
3. Determine the significance level of the sample's ANOVA, which "Sig." identifies exactly.
4. Adhere the following rule: if Sig. is less than the cutoff score (.05) than you reject the null hypothesis and your research hypothesis is supported.
5. Decide whether to reject the null hypothesis and explain your answers.