Skip to main content

NS5108 Research Methods Handbook

11 JASP workship - Factorial ANOVA

In the video below, I walk through examples of analysing factorial ANOVAs. If you want to follow along, please find the corresponding question sheets and datasets available in the NS5108 module on Moodle.

11.1 “Where to Click” Guide - Conducting Factorial ANOVAs in JASP

Sometimes you just want to know where to click to run the test. Below is a step-by-step guide for performing factorial ANOVAs in JASP. Refer to the video above for more context regarding these steps.

11.1.1 Conducting a fully Independent Samples Factorial ANOVA

  1. Open JASP and Load Your Data:
  • Click on the File tab at the top left.
  • Select Open, and navigate to the folder containing your data file.
  1. Explore data:
  • Due to the factorial design, you will need to filter the data as you are only able to split data by one variable. To filter there is a funnel symbol to click in the top right of the screen.
  • Once you have the filter box up you will need to bring in one of your IVs, click “=” and then type one of the levels. Then click “apply pass through filter” After this everything you run will be just on that level of the variable.
  • Go to Descriptives -> Descriptive Statistics.
  • Move your dependent variable into the Variables box and your independent variables (factors) into the Split box.
  • Use Plots to create histograms or boxplots for each combination of factors to check the distribution and identify potential outliers.
  • Clear change the filtered level to the other level of the IV and repeat
  • Once finished make sure to clear the filter entirely.
  1. Check Assumptions:
  • Normality:
  • Check for normality within each combination of the independent variables. Again you will need to filter to account of the design.
  • In the Descriptive Statistics window, under Statistics, check Skewness and Kurtosis for each condition.
  • Use the Shapiro-Wilk Test to assess normality within each condition if sample sizes are small (typically less than 50 per group).
  • Homogeneity of Variance:
  • Navigate to ANOVA -> ANOVA.
  • Move your dependent variable to Dependent Variable and your independent variables to Fixed Factors.
  • Under Assumption Checks, check Homogeneity tests to perform Levene's Test.
  • A non-significant p-value (p > .05) in Levene's Test indicates that the assumption of homogeneity of variances is met across all conditions.
  1. Run the Factorial ANOVA:
  • Ensure your dependent variable and independent variables are correctly specified.
  • Select to report effect sizes
  • Select to create line graphs to illustrate the relationship between variables
  • Select to run post hoc tests
  1. Post Hoc Tests:
  • If significant main effects or interactions are found, go to the Post Hoc Tests tab.
  • Move your independent variables or interactions into the Post Hoc Tests box.
  • Select a correction method (e.g., Bonferroni or Tukey) for multiple comparisons.
  • This will help you identify where the significant differences lie among your levels of the independent variables.

11.1.2 Conducting a Factorial ANOVA with a repeated measures variable

  • Name your repeated measures variable by clicking where it says RM Factor 1 and renaming it.
  • Bring each of the levels of your repeated measures variable into where it says level 1 and level 2, adding more levels if needed.
  • Add new repeated measure IVs in if needed and repeat
  • Bring in any independent samples IV in to the between subject factors box
  • Repeat all other steps form the independent samples factorial ANOVA

11.2 APA Style Guide for Reporting Factorial ANOVAs

Here's how to format your hypotheses and report results from your factorial ANOVAs in APA style.

Hypotheses for an Independent Samples Factorial ANOVA:

  • Null Hypothesis (H₀): There wil be no significant main effect of IV1], no significant main effect of [IV2], and no interaction effect between [IV1] and [IV2] on [dependent variable].

  • Alternative Hypothesis (H1): There will be a significant main effect of [IV1] whereby ...

  • Alternative Hypothesis (H2): There will be a significant main effect of [IV2] whereby ...

  • Alternative Hypothesis (H3): There will be a significant interaction effect whereby the change in IV1 will be ... as compared to the change in IV2

11.3 Example APA Report for an Independent Samples Factorial ANOVA

"An independent samples factorial ANOVA was conducted to examine the effects of [IV1] and [IV2] on [dependent variable]. Assumption checks confirmed normality and homogeneity of variance. The results revealed a significant main effect of [IV1], F(df_between, df_within) = F-value, p = p-value, η² = effect size, and a significant main effect of [IV2], F(df_between, df_within) = F-value, p = p-value, η² = effect size. Additionally, there was a significant interaction effect between [IV1] and [IV2], F(df_between, df_within) = F-value, p = p-value, η² = effect size. Post hoc analyses using Bonferroni correction indicated that [specific group differences], suggesting that [interpretation of results]."

The following is an example analysis for a fully repeated measures ANOVA design:

11.4 Example APA Report for an Repeated Measures Factorial ANOVA

"A repeated measures factorial ANOVA was conducted to examine the effects of [Repeated Measure IV] and [Between-Subjects IV] on [dependent variable]. Mauchly's Test indicated that the assumption of sphericity had been met/violated (χ²(df) = value, p = p-value). The results showed a significant main effect of [Repeated Measure IV], F(df_between, df_error) = F-value, p = p-value, η² = effect size, and a significant main effect of [Between-Subjects IV], F(df_between, df_error) = F-value, p = p-value, η² = effect size. There was also a significant interaction effect, F(df_between, df_error) = F-value, p = p-value, η² = effect size. Using the Greenhouse-Geisser correction (if sphericity is violated), the effect remained significant/non-significant. Post hoc pairwise comparisons showed that [specific condition differences], indicating that [interpretation of results]."