When analyzing mean differences among more than two samples, doing independent t tests on all possible pairs of means
a. Decreases the probability of a Type I error
b. Does not change the probability of a Type I error
c. Increases the probability of a Type I error
d. Cannot be determined from the information provided