T-test: Salaries of Female and Male Human Resource Managers
Christy Newman
Argosy University

T-test: Salaries of Female and Male Human Resource Managers

An independent-samples t-test using the raw measurement data presented was completed to check the difference, if any, of the salaries between female and male human resource managers, t(18) = -0.408, p = 0.688, but no significant difference was found (Female Mean = 62.2; Male Mean = 63.7).
The preparation for a statistical analysis test began with a well-developed, clear research question: What is the difference, if any, between the salaries of female and male HR Directors? Next, the null and alternative hypotheses must be defined. The null H₀ is that the salaries of female human resource managers = salaries of male human resource managers. The alternative Ha is that salaries of female human resource managers ≠ salaries of male human resource managers.
The next step in the process is to determine the appropriate statistical test and sampling distribution. Since σ is unknown and the number of salaries being tested is less than 100, the t-test will be used. Because we are comparing data from two different groups, the two-sample t-test will be used. We do not know if the variance is different or equal, so we will use the two-tailed t-test for two-sample assuming unequal variances.
Subsequently, we have to choose the Type 1 Error rate. For this t-test, α = 0.05, which is the standard rate used by statisticians for research purposes. After that, we have to calculate the degrees of freedom: df = (n₁ - 1) + (n₂ - 1) = (10 – 1) + (10 – 1) = 9 + 9 = 18. Before we can begin the t-test, we also need to state the decision rule: +/- 0.444 will either reject or fail to reject H₀.
Once we have determined the information above, we can calculate test statistic. Many different software programs and websites exist to compute