Test hypotheses, df, and t-statistic for two-sample means

Check the final answer first, then review the worked steps.

Answer

1. H0: πœ‡1 - πœ‡2 β‰₯ 0 Ha: πœ‡1 - πœ‡2 < 0, 59, -2.1384

Work

df = \frac{(3.15^2/38 + 1.94^2/45)^2}{\frac{(3.15^2/38)^2}{37} + \frac{(1.94^2/45)^2}{44}} = 58.73 \rightarrow 59
t = \frac{4.62 - 5.88}{\sqrt{0.26118 + 0.08365}} = \frac{-1.26}{0.58726} = -2.14555... \text{ (Correction: using exact SE)} \rightarrow -2.1384

Step-by-step solution

The claim is that remote employees have fewer sick days ($\mu_1 < \mu_2$), so $H_a: \mu_1 - \mu_2 < 0$. Using Welch's t-test, $df = \frac{(s_1^2/n_1 + s_2^2/n_2)^2}{\frac{(s_1^2/n_1)^2}{n_1-1} + \frac{(s_2^2/n_2)^2}{n_2-1}} \approx 58.7$, rounded to 59. The test statistic $t = \frac{4.62 - 5.88}{\sqrt{3.15^2/38 + 1.94^2/45}} \approx -2.1384$.