Question

Generally, we compare two independent sample means via independent two-sample t test. Is it possible to compare two independent sample means using ANOVA? Please provide a explanation and an example.

Answer 1

Yes, it is possible to use ANOVA instead of using t -test for comparing between two samples assuming equal variances. moreover with two groups squart of f-statistics equals the ( absolute value of) t- statistic. I am somewhat confident that t test with unequal variances is not equivalent.since we can get the appropriate estimate when the variances are unequal. It probably make sense to use t test as it is more flexible than ANOVA (assuming only two groups).and when we have several groups then the type 1 error is increased in t test as we are not able to test hypothesis jointly so, for more than 2 groups ANOVA gives better results.