A professor asked her research intern to perform an ANOVA test. Instead, the research intern ran a series of t tests thinking that would be more efficient than conducting the complex ANOVA test. What errors might occur as a result of conducting t tests instead of an ANOVA test?

1.The familywise error rate will increase, making the work fraught with unacceptably high levels of error.

2.The within-group variance will decrease to such low levels, the validity of the work will come into question.

3.The levels of Type II error will increase rapidly, making the analyses meaningless.

4.The between-group variance will decrease as the t tests are performed, making it seem as though a relation exists when there isn't one.