0
$\begingroup$

I have an experiment with two conditions: Control and treatment groups. I am measuring how confident the treatment groups felt while answering the question. It is a three level indicator- not sure, unsure, really sure. The experiment is done in pairs(2*2).

Each pair play the game together but only have access to their partners assurity levels. The pair's total adds up for the team score. Each team only gets the team score and no information of an individual partner's score.

How do I treat these levels in regression. I am measuring Team performance vs confidence(how sure they were).

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

It seems that you need to use effect coding, or dummy coding, in this analysis. You need to recode the different levels of your "assurity levels" into 1's and 0's. Once that is done, you can run a regression analysis that uses the dummy variables in different combinations, to hone in on the comparisons you are most interested in.

More details are here.

Improve this answer
$\endgroup$

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.