The One-Way ANOVA is an extension of the independent-samples t-test and is used when comparing more than 2 independent groups. The ANOVA will tell you if there are any significant differences between the groups, but you will need to conduct a post hoc analysis to determine which groups are significantly different.
Assumptions
- One continuous (interval or ratio) dependent variable and one categorical (nominal or ordinal) independent variable with more than two levels.
- Independence of observations - usually evaluated based on the research design
- No significant outliers - can be assessed using boxplots, scatterplots, and other methods
- The dependent variable is approximately normally distributed for each level of the independent variable.
- Homogeneity of variances - can be tested using Levene's Test
Running a One-Way ANOVA in SPSS
- Analyze > Compare Means > One-Way ANOVA
- Move the continuous variable into the "Dependent List" box and the categorical variable into the "Factor" box
- Click on the Post Hoc button
- choose the appropriate post hoc analysis for your study
- Select Continue
- You may select additional output, such as descriptive statistics, using the Options button
- Select OK to run the analysis
Interpreting the Output
- Descriptives (if you opted to include them)
- provides measures of central tendency and dispersion based on levels of the IV
- ANOVA
- provides the results of the statistical test
- test statistic = F-ratio
- associated probability = Sig.
- used to make a decision about the null hypothesis
- Multiple Comparisons
- provides the results of the post hoc analysis
- allows you to determine exactly which groups are significantly different than each other
- compare the Sig. to your level of significance (i.e. .05)
Reporting Results in APA Style
There was a statistically significant difference between groups as determined by the one-way ANOVA (F(2,62) = 5.467, p = .012). A Tukey post hoc test revealed that the time to respond was significantly lower for those that received the reaction time training (M = 1.3, SD = .5) and those that were experts in the field (M = 1.7, SD = .8) compared to those that had no training (M = 2.8, SD = .7). There was no statistically significant difference between the training group and the expert group (p = .214).