jamovi

Hello everyone! I come to you because, having not had a teaching of statistics last year, I can not understand and use the statistical tests. I have to do a moderation analysis for my thesis. And I have some questions to ask you.

1. Can my moderate variable be categorical? (mine is divided into 4 groups)
2. Can I do a linear regression instead of using medmod to do a moderation?
3. Would it be possible to explain to me what I should conclude from the value of "estimate" "SE" "t" and "p" please?

Thanks a lot for your help !

hi labafr,

other's may be able to help you out here, but there may be other places online where you can get answers to these 'basic statistics' questions, more readily than here. i personally don't know a lot about mediation and moderation, so i'm not a lot of help to you.

again, others may chime in and explain things (and you're welcome to post these sorts of questions) but you'll probably find the participants in this forum are more focused on the use of jamovi than what might be called "statistics in general".

so i'm really just adjusting your expectations in case you don't get much of a response here.

cheers

jonathon

labafr wrote: 1. Can my moderate variable be categorical? (mine is divided into 4 groups)
2. Can I do a linear regression instead of using medmod to do a moderation?
3. Would it be possible to explain to me what I should conclude from the value of "estimate" "SE" "t" and "p" please?

ad.1. No, it should be dichotomous - groups has to have equal distance between them, what is possible only in case with 2 groups

ad.2. Yes, of course

sd. 3. "t" = "estimate" / "SE", "p" is derived from t-distribution and tell You how high is the probability to reject false hypothesis of null effect

More: http://www.quantpsy.org/medn.htm