After gathering the data for an experiment with a one factor ...

[dejastew]

After gathering the data for an experiment with a one factor between groups design, if we were to conduct several t tests instead of an ANOVA, the probability of _____ would be unacceptably high.
 
  a. a Type I error c. null results
  b. a Type II error d. both a and b

Question 2

In a one factor between groups design, if we want to know whether the difference between group means was produced by chance, we should analyze the data by conducting _____.
 
  a. an analysis of variance
  b. separate t tests for each pair of means
  c. either a or b
  d. none of these - we can tell by looking at the results on a bar graph

[SVictor]

Answer to Question 1

A
Such a procedure would increase the probability of a Type I error by alpha for each comparison made. The result is that the experiment wise probability of a Type I error would be way too high.

Answer to Question 2

A
Separate t tests would raise the probability of a Type I error to an unacceptable level.

[dejastew]

Excellent

[nguyenhoanhat]

Wow, this really help