The one-sample, independent-samples, and paired-samples t-tests constitute a family of parametric statistical procedures used to determine whether observed differences between sample means and reference values or between groups are statistically significant under the assumption of normality. These tests formalize hypothesis testing within inferential statistics by comparing calculated t-values against critical distributions derived from degrees of freedom, where the test statistic is defined as the ratio of the mean difference (or deviation) to its standard error. The selection among these variants depends strictly on the sampling design—specifically whether groups are independent or dependent—and requires verification that metric variables exhibit normal distribution and homoscedasticity for independent comparisons.
The one-sample, independent-samples, and paired-samples t-tests constitute a family of parametric statistical procedures used to determine whether observed differences between sample means and refere…