Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Supported by

Polynomial contrasts

edited April 2016 in JASP & BayesFactor

I was wondering why the polynomial contrasts (after an ANOVA) come as t-tests instead of F tests. I also noticed that the significance of those doesn't match the significance obtained with Statistica or SPSS. Any ideas on this?


  • edited 8:53PM

    I see that SPSS provides t-tests and an omnibus F test which is the same as the result in the ANOVA table. As the F test is already provided in the regular ANOVA table in JASP I don't really see a reason to give the same exact result again for the contrasts. About the differences in significance: can you send me the data set where this happens? I just did a polynomial contrast using SPSS and it gives me the same exact result as JASP.

  • edited 8:53PM

    Thanks Ravi! I understand what you mean, but I think I really got different results. But this is the first time I'm using polynomial contrasts so I might be messing up at some point. Anyway, I noticed that usually the F statistics is what's reported. But here's a link to my data set (it's a repeated measures design) and I also included both the JASP and SPSS outputs. contrasts_elle.xlsx?dl=0
    What do you think?

  • edited 8:53PM

    Ah, I thought you were talking about the regular ANOVA. I'm not completely sure how SPSS produces its results so I will look into it and get back to you.

Sign In or Register to comment.