A clinical researcher in Mumbai compared the mean systolic blood pressure (SBP) between two independent groups of hypertensive patients: those on antihypertensive therapy for 6 months (Group A, n=45) and those on therapy for 12 months (Group B, n=50). Both groups had approximately normally distributed SBP values with similar variances. Which test is most commonly used to determine if there is a statistically significant difference in mean SBP between the two groups?

Explanation

## Test Selection for Continuous Data — Two Independent Groups ### Study Design Analysis The researcher is comparing: - **Two independent groups** (different patients, different treatment durations) - **One continuous outcome** (systolic blood pressure in mmHg) - **Approximately normal distribution** with **similar variances** This is the classic scenario for the **independent samples t-test**, the most commonly used test in clinical research for this design. ### Why Independent Samples t-Test? **Key Point:** The independent samples t-test (also called unpaired t-test) is the most common parametric test for comparing means between two independent groups when the outcome is continuous and approximately normally distributed. **High-Yield:** Assumptions for independent samples t-test: 1. Two independent groups (different subjects) 2. Continuous outcome variable 3. Approximately normal distribution in each group (robust to mild violations with n ≥ 30) 4. Homogeneity of variances (Levene's test; if violated, use Welch's t-test) 5. Independent observations (no clustering) ### Test Formula $$t = \frac{\bar{X}_1 - \bar{X}_2}{SE(\bar{X}_1 - \bar{X}_2)}$$ where SE is the pooled standard error of the difference between means. ### Decision Tree for Choosing t-test vs. Alternatives ```mermaid flowchart TD A[Comparing means between groups?]:::decision A -->|Yes| B{How many groups?}:::decision B -->|2 groups| C{Independent or paired?}:::decision B -->|3+ groups| D[Use ANOVA]:::action C -->|Independent| E{Normal distribution?}:::decision C -->|Paired| F[Paired t-test]:::action E -->|Yes + equal variances| G[Independent t-test]:::action E -->|Yes + unequal variances| H[Welch's t-test]:::action E -->|No + small n| I[Mann-Whitney U test]:::action G --> J[Most common in clinical research]:::outcome ``` ### Comparison of Related Tests | Test | Data Type | Groups | Assumption | Use Case | | --- | --- | --- | --- | --- | | **Independent t-test** | Continuous | 2 independent | Normal, equal variances | **Most common** — comparing means between two different groups | | **Paired t-test** | Continuous | 2 paired | Normal differences | Same subjects measured twice (before-after) | | **Welch's t-test** | Continuous | 2 independent | Normal, **unequal variances** | When Levene's test p < 0.05 | | **Mann-Whitney U** | Continuous/Ordinal | 2 independent | Non-normal | Non-parametric alternative | | **ANOVA** | Continuous | 3+ independent | Normal, equal variances | Comparing means across multiple groups | **Clinical Pearl:** With sample sizes of 45 and 50 (both ≥ 30), the t-test is robust to mild departures from normality due to the Central Limit Theorem. The independent samples t-test is the gold standard and most frequently used test in clinical trials and observational studies comparing two treatment groups. **Mnemonic:** **I-C-I** for Independent samples t-test: **I**ndependent groups, **C**ontinuous outcome, **I**ndependent observations.