In Example 8.3, suppose the organization wanted to compare the mean test scores of Catholic priests and Methodist ministers. Note that it appears based on the data that these two groups have the same variance. What is the gain in using a two-sample t-test having s2W in the denominator as opposed to using the conventional pooled t test with s2p, the average of the sample variances for the Catholic priests and Methodist ministers?