Jamovi Tutorial Guide: Assumption-Based Analysis Workflow
Introduction
This guide provides a comprehensive step-by-step tutorial on how to use Jamovi for statistical analyses. It
begins by checking assumptions, followed by instructions for running analyses when assumptions are
fulfilled and not fulfilled. Covered analyses include: - Descriptive Statistics - Independent Samples t-Test -
Paired Samples t-Test - Correlation - One-Way ANOVA - Two-Way ANOVA
Part 1: Assumption Checking in Jamovi
A. General Steps for Assumption Checks
1. Open Dataset: Launch Jamovi and open your dataset via “☰ Menu > Open”.
2. Descriptive Exploration:
3. Go to Exploration > Descriptives
4. Select variables of interest
5. Enable: Mean, Median, SD, Histogram, Q-Q plot, and Shapiro-Wilk test
6. Interpretation: A Shapiro-Wilk p-value > .05 suggests data is approximately normally distributed.
7. Boxplots:
8. Also in Descriptives, check boxplots to detect outliers visually
9. Interpretation: Extreme outliers may violate assumptions and impact analysis results.
Part 2: Analysis Procedures Based on Assumption Results
1. Descriptive Statistics
No assumption violations apply. - Go to: Exploration > Descriptives - Select variables - Enable plots
(Histogram, Boxplot) and stats (Mean, SD, etc.) - Interpretation: Use descriptive stats to understand
distribution, central tendency, and variability of your data.
2. Independent Samples t-Test
A. If Assumptions Are Fulfilled
• Go to: T-Tests > Independent Samples t-Test
• Dependent Variable: Continuous variable
• Grouping Variable: Categorical (2 groups)
• Check: Assumption Checks > Shapiro-Wilk, Levene's
• Interpretation:
• Shapiro-Wilk p > .05 = normal distribution
1
• Levene's p > .05 = equal variances
• t-test p < .05 = significant difference between groups
B. If Assumptions Are Not Fulfilled
• If Levene's fails (p < .05): Use Welch's t-test (Jamovi auto-adjusts)
• If Shapiro-Wilk fails: Use Mann-Whitney U Test:
• Go to: ANOVA > Nonparametric Tests > Independent Samples
• Interpretation: Mann-Whitney p < .05 = significant difference without assuming normality or equal
variance.
3. Paired Samples t-Test
A. If Assumptions Are Fulfilled
• Go to: T-Tests > Paired Samples t-Test
• Add variable pair
• Check: Assumption Checks > Shapiro-Wilk (on difference)
• Interpretation:
• Shapiro-Wilk p > .05 = normal distribution of differences
• t-test p < .05 = significant change between paired values
B. If Assumptions Are Not Fulfilled
• Use Wilcoxon Signed-Rank Test:
• Go to: ANOVA > Nonparametric Tests > Paired Samples
• Interpretation: Wilcoxon p < .05 = significant median change between paired samples
4. Correlation
A. If Assumptions Are Fulfilled
• Use: Pearson Correlation
• Go to: Exploration > Correlation Matrix
• Select variables
• Enable: Pearson
• Enable scatterplots to verify linearity and check assumptions visually
• Interpretation:
• Pearson r indicates strength/direction of linear relationship
• r close to +1 or -1 = strong relationship, p < .05 = statistically significant
B. If Assumptions Are Not Fulfilled
• Use: Spearman Correlation
• Go to: Exploration > Correlation Matrix
• Select variables
2
• Enable: Spearman
• Interpretation:
• Spearman rho measures monotonic relationship (not necessarily linear)
• rho close to +1 or -1 = strong rank-based correlation, p < .05 = statistically significant
5. One-Way ANOVA
A. If Assumptions Are Fulfilled
• Go to: ANOVA > One-Way ANOVA
• DV: Continuous variable
• Factor: Categorical variable (3+ levels)
• Check: Assumption Checks > Shapiro-Wilk, Levene's
• Add Post Hoc: Tukey if significant
• Interpretation:
• ANOVA p < .05 = at least one group differs
• Tukey test identifies which groups differ significantly
B. If Assumptions Are Not Fulfilled
• Use Welch's ANOVA:
• Check Welch's box
• Or use Kruskal-Wallis Test:
• Go to: ANOVA > Nonparametric Tests > One-Way
• Interpretation:
• Welch's or Kruskal-Wallis p < .05 = significant group differences
6. Two-Way ANOVA
A. If Assumptions Are Fulfilled
• Go to: ANOVA > ANOVA
• DV: Continuous
• Factors: Two categorical variables
• Include interaction term under Model Terms
• Check: Assumption Checks > Shapiro-Wilk, Levene's
• Interpretation:
• p < .05 for main effects = significant factor influence
• p < .05 for interaction = one factor's effect depends on the other
B. If Assumptions Are Not Fulfilled
• No built-in nonparametric equivalent in base Jamovi
• Options:
• Consider transformation of DV
• Use robust ANOVA in the WRS2 module (add-on)
3
• Interpretation: Interpret robust methods cautiously, especially with interaction terms or
transformed variables.
Summary
Always check assumptions before proceeding with analyses. Jamovi provides easy-to-use tools for both
checking assumptions and switching to appropriate alternatives when assumptions are not met. Use non-
parametric or robust methods when in doubt. Interpret p-values carefully and combine statistical findings
with theoretical understanding.
End of Guide
