Academic Methodology Resources & Digital Stats Toolkit
The Wilcoxon Signed-Rank Test is a non-parametric powerhouse used to determine whether there is a median difference between paired observations. It is the essential alternative to the Dependent T-test when your data violates the assumption of normality.
Unlike parametric tests that use exact values (which outliers can skew), Wilcoxon converts differences into Ranks. This prioritizes the relative position of data, making it resilient against extreme scores.
Wilcoxon does not assume your data follows a "Bell Curve." It is "Distribution-Free," meaning it is ideal for ordinal scales (Likert) and small sample sizes where normality cannot be proven.
| Decision Metric | Paired t-Test (Parametric) | Wilcoxon (Non-Parametric) |
|---|---|---|
| Data Type | Interval / Ratio | Ordinal / Skewed Interval |
| Normality | Required (Bell Curve) | Not Required |
| Sample Size | Large (n > 30) preferred | Excellent for Small Samples |
| Central Tendency | Mean Difference | Median Difference |