A right-tailed one-sample t-test of \(H_0: \mu = \mu_0\) against \(H_1: \mu \gt \mu_0\) is to be done at 5% level of significance. For every sample size n considered, the sample mean exceeds \(\mu_0\) by 3, and the sample standard deviation (calculated with divisor n − 1) is 8. The table values of t at 5% level of significance (one-tailed) are
| Degrees of freedom | 19 | 20 | 21 | 22 |
| Table value of t | 1.729 | 1.725 | 1.721 | 1.717 |
The smallest of the following sample sizes for which \(H_0\) is rejected is (Use \(\sqrt{20} = 4.472\), \(\sqrt{21} = 4.583\), \(\sqrt{22} = 4.690\), \(\sqrt{23} = 4.796\))
Answer & explanation
Correct answer: option 3
The correct answer is Option (3) → 22
For a sample of size n,
\( t = \displaystyle\frac{\bar{x} - \mu_0}{s/\sqrt{n}} = \frac{3\sqrt{n}}{8} \)
and the table value is read for \(df = n - 1\). \(H_0\) is rejected when t is more than the table value.
n = 20 (df 19): \(t = 3 \times 4.472 \div 8 = 1.677 \lt 1.729\), not rejected
n = 21 (df 20): \(t = 3 \times 4.583 \div 8 = 1.719 \lt 1.725\), not rejected
n = 22 (df 21): \(t = 3 \times 4.690 \div 8 = 1.759 \gt 1.721\), rejected
n = 23 (df 22): \(t = 3 \times 4.796 \div 8 = 1.799 \gt 1.717\), rejected
So the smallest sample size among these for which \(H_0\) is rejected is n = 22. The table value must be read for df = n − 1, not for df = n.
Why the other options are incorrect
Option (1) 20 — Incorrect. t = 1.677 is less than 1.729 (df 19), so \(H_0\) is not rejected.
Option (2) 21 — Incorrect. t = 1.719 is less than 1.725 (df 20), so \(H_0\) is not rejected.
Option (4) 23 — Incorrect. \(H_0\) is rejected for n = 23, but it is already rejected for the smaller size n = 22.