PG851: GLM--ANOVA

This blog is for the use of students enrolled in either section of PG851. It is intended to be a forum for asking, discussing, clarifying, and helping.

Sunday, January 24, 2010

Results section for ANOVA w/ multiple groups

When reporting which groups were found to be different in the post-hoc tests, do we need to report the p value if there is no significance?

For example:

It was found that there was no difference between group 1 and group 2, p = .42.

Or do we just exclude the p value?

Thanks,
Lauren

posted by Lauren B @ 3:39 PM  

3 Comments:

  • At 6:10 PM, Blogger Andrew said…

    I know that when you report that a relationship is not significant you should include the numbers to back up your statement.
    However I'm not sure if you need to state every non-significant relationship found using a post hoc test. Mari please comment.

     

  • At 7:36 PM, Blogger Mari said…

    For ANOVAs themselves, you must report the F, df, and p of every effect actually tested. Always.

    For post-hocs, this can get really messy if you have a ton of groups. Three, however, is not a ton, so you should report the ps associated with the conclusions you make (or, in Andy's words, "include the numbers to back up your statement").

     

  • At 8:09 PM, Blogger Lauren B said…

    thank you

     

Post a Comment

<< Home