I am reading a research paper about the relationship between service quality and consumer behavior. One hypothesis is: customers experiencing no service problems have the strongest levels of favorable behaviors and the smallest levels of unfavorable behaviors. ANOVA was operated, results are:

enter image description here

The author intepreted:

"The pay more intentions of customers experiencing no service problems are not significantly higher than those of customers experiencing service problems that were resolved."

But the significance level for this factor was ,036, less than ,05. It should be significant, right? (The author used ,05 as standard for other analysis )

  • 1
    $\begingroup$ It's clearly explained under the table what was done and why. The 0.05 level was the overall type I error rate; the per-test type I error rate was therefore lower. $\endgroup$
    – Glen_b
    Commented Jun 20, 2014 at 9:13

1 Answer 1


If you read the details printed below the table it tells you that they used a Bonferroni correction to yield a critical alpha level of 0.006.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.