In the one-sample t-test \( t = \displaystyle\frac{\bar{x} - \mu_0}{s/\sqrt{n}} \), the sample standard deviation s of a sample of n observations is calculated as
Answer & explanation
Correct answer: option 2
The correct answer is Option (2) → \( s = \sqrt{\displaystyle\frac{\sum (x - \bar{x})^2}{n - 1}} \)
In the t-test the population standard deviation is unknown, so it is estimated from the sample. When the sample mean \(\bar{x}\) is used, only n − 1 deviations are free to vary, so the sum of squared deviations is divided by the degrees of freedom n − 1:
\( s = \sqrt{\displaystyle\frac{\sum (x - \bar{x})^2}{n - 1}} \)
Why the other options are incorrect
Option (1) \( s = \sqrt{\displaystyle\frac{\sum (x - \bar{x})^2}{n}} \) — Incorrect. Dividing by n gives the standard deviation of the sample itself, not the estimate s used in the t-test.
Option (3) \( s = \displaystyle\frac{\sum (x - \bar{x})^2}{n - 1} \) — Incorrect. This is the variance \(s^2\); the square root has not been taken.
Option (4) \( s = \sqrt{\displaystyle\frac{\sum (x - \bar{x})}{n - 1}} \) — Incorrect. The deviations are not squared, and \(\sum (x - \bar{x})\) is always 0.