Howdy, Stranger!

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

Supported by

interpreting and reporting results

Dear forum,

I am quite new in Bayes analysis and want to make sure that I am interpreting and reporting my results correctly. Let me briefly explain my study. I have two different tests with normative scores.

In my analysis, I have compared the normative scores of each test with my observers' scores. For that I have conducted two one-sample t-test. The results revealed a BF01 of 0.00006244185 for the first test and a BF01 of 8.98 for the second test.

My interpretation: These results reflect that the differences between my observers' scores and the normative scores are 16014.9 (1/0.00006244185) more favored than the null model, for test A. For test B, the lack of differences between my observers' scores and the normative scores is 8.98 more favoured than the differences.

A couple of question about my interpretation. First, is it right? Second, I find confusing reporting BF01 for test A. Would it be possible to report BF10 for test A and BF01 for test B, or is it better to keep consistency? How do you report a big (or small) BF value? I don't find very elegant to report a large number such as BF10 = 16014.9 (or BF01 = 0.00006244185. Could this be reportrted like BF01 < .001, or in a similar way?

Thanks in advance for your help!

Comments

  • Dear Ajestudillo,

    Your interpretation is right (as long as, when you speak of one hypothesis being "favored", you mean to say that the data are more likely under that hypothesis than under the other). And yes,. I always transform the BFs such that they exceed 1. For clarity, you could simply explain what you did (as you did on this forum). Finally, I would simply report those large BFs. It may not be elegant, but it is precise.

    Cheers,
    E.J.

  • Thanks a lot for your reply EJ!

Sign In or Register to comment.