Independent Samples t-Test

Use this test when you have a continuous outcome measured in two separate groups and you want to know whether the group means differ. For example, comparing blood pressure between a treatment group and a control group, or comparing test scores between two classrooms.

• The dependent variable is continuous (interval or ratio scale).
• Observations in each group are independent of one another.
• The dependent variable is approximately normally distributed within each group, or sample sizes are large enough for the Central Limit Theorem to apply (roughly N >= 30 per group).
• If you choose the pooled method, group variances should be reasonably similar. easyCris also reports a Folded F variance check as supplemental context.

• If the p-value (Pr > |t|) is less than your chosen alpha (typically 0.05), the difference between group means is statistically significant.
• easyCris reports both the Pooled and Satterthwaite rows. Interpret the row that matches the method you selected in the app.
• The Folded F table provides context about variance similarity between the two groups, but easyCris does not silently switch your chosen method for you.
• The mean difference tells you the direction and magnitude of the effect. Always report the confidence interval alongside the p-value.
• A confidence interval that does not include zero is consistent with a significant result.
• Statistical significance does not imply practical importance. Consider whether the observed mean difference is meaningful in your research context.

• Choosing the pooled method when group variances are clearly unequal can inflate the Type I error rate; in that situation, the Satterthwaite method is usually safer.
• Do not assume the app automatically overrides your chosen method based on the variance check. The guide and the result table should be read together.
• The t-test is sensitive to outliers, which can distort group means and inflate or mask real differences.
• Small sample sizes reduce statistical power. A non-significant result with a small sample does not mean there is no difference, only that you could not detect one.
• Running multiple t-tests across many group pairs without correction inflates the family-wise error rate. Use ANOVA with post-hoc tests instead.

1. Calculate the mean and variance for each group.
2. Compute both t-statistic rows shown in the app: the pooled version uses a common variance estimate, while the Satterthwaite version uses separate group variances.
3. Determine the degrees of freedom. For pooled: N1 + N2 - 2. For Satterthwaite: the unequal-variance approximation based on each group's variance and sample size.
4. Compute the Folded F variance comparison from the two sample variances and report it in the Equality of Variances table.
5. Compare the t-statistic to the t-distribution with the appropriate degrees of freedom to obtain the p-value.