Statistical test selection — quick reference table
| Situation | Normal data | Non-normal data |
|---|---|---|
| Compare 2 independent group means | Independent t-test | Mann-Whitney U |
| Compare paired / before-after means | Paired t-test | Wilcoxon signed-rank |
| Compare 3+ group means | One-way ANOVA | Kruskal-Wallis |
| Repeated measures (3+ time points) | Repeated-measures ANOVA | Friedman test |
| Compare proportions / categories | Chi-square (Fisher's exact if expected count < 5; McNemar if paired) | |
| Correlation between two continuous variables | Pearson's r | Spearman's rho |
| Predict binary outcome | Logistic regression | |
| Predict continuous outcome | Multiple linear regression | |
Examiner tip: State the normality test in your methodology — "Normality was assessed using the Shapiro-Wilk test; parametric tests were applied for normally distributed variables." This one sentence pre-empts the most common statistics viva question.
Frequently asked questions
How do I check whether my data is normal?
In SPSS: Analyze → Descriptive Statistics → Explore → Plots → Normality plots with tests. Use Shapiro-Wilk for n < 50. If p > 0.05, treat as normal.
Chi-square or Fisher's exact — which one?
Use Fisher's exact test when any expected cell count in your 2×2 table is below 5; otherwise chi-square. SPSS prints both — report the right one.
What is a p-value of 0.05 actually saying?
It means there is a 5% probability of seeing a difference this large (or larger) purely by chance if no real difference exists. p < 0.05 is the conventional threshold for "statistically significant" in medical theses.
Related free tools
- Sample Size Calculator — with ready-to-paste protocol paragraph
- Study Design Finder — the design decides the test
- How to Write an MD/MS Thesis — full roadmap