recent research has shown that although the pooled t test does outperform the two-sample t test by a bit (smaller β 's for the same α ) when σ12=σ22, the former test can easily lead to erroneous conclusions if applied when the variances are different. ' (i) What does it mean by smaller β 's for the same α "? Why does this imply that the pooled t test outperforms the two-sample t test? (ii) Why would the pooled t test outperform the two-sample t test when σ12=σ22 ? Discuss from the point of view of equal sample sizes.