Start with two questions: what are you trying to find out, and what type of data do you have?
Comparing groups
| Goal | Data | Parametric test | Non-parametric alternative |
|---|---|---|---|
| 2 independent groups | Continuous | Independent t-test | Mann–Whitney U |
| 2 related measurements | Continuous | Paired t-test | Wilcoxon signed-rank |
| 3+ independent groups | Continuous | One-way ANOVA | Kruskal–Wallis |
| 3+ related measurements | Continuous | Repeated-measures ANOVA | Friedman test |
| Groups vs categories | Categorical | Chi-square test of independence | Fisher's exact test (small samples) |
Relationships and prediction
| Goal | Data | Test |
|---|---|---|
| Strength of relationship | 2 continuous variables | Pearson correlation (or Spearman if not normal) |
| Predict a continuous outcome | Several predictors | Multiple linear regression |
| Predict a yes/no outcome | Several predictors | Binary logistic regression |
| Test a theoretical model | Latent constructs | SEM / PLS-SEM (AMOS, SmartPLS) |
Before you run anything
- Check assumptions: normality, equal variances, independence, outliers.
- Report effect sizes and confidence intervals, not only p-values.
- Decide your analysis plan before looking at results.
Not sure which applies to your study? Bring your research questions to a free consultation.
Your brand here
Reach students and researchers with a clearly labelled placement.
Get featured
Reach students and researchers with a clearly labelled placement.
Questions & discussion