Problem with CI in independent samples t-test

Discuss the jamovi platform, possible improvements, etc.

by mdelacre » Wed Mar 18, 2020 11:18 am

I think that you reversed Confidence limits for mean differences when the alternative hypothesis is mu1>mu2 and mu1<mu2, when performing an independant Sample t-test. If our expectations are that mu1>mu2, we need to check the upper limit (and lower limit should then be -Inf).

Best Regards,

Marie Delacre
mdelacre
 
Posts: 5
Joined: Fri Jul 12, 2019 9:39 am

by jonathon » Thu Mar 19, 2020 12:01 am

hi,

if our expectation is mu1 > mu2, then we expect that (mu1 - mu2) > 0, in which case we want to check the lower limit? (the confidence interval is on the difference).

kind regards

jonathon
User avatar
jonathon
 
Posts: 1115
Joined: Fri Jan 27, 2017 10:04 am

by mdelacre » Fri Mar 27, 2020 11:03 am

Hi, you're perfectly right! I confused with CI around 0, my bad :) sorry for the waste of time, and thanks for your answer!
mdelacre
 
Posts: 5
Joined: Fri Jul 12, 2019 9:39 am

by jonathon » Fri Mar 27, 2020 10:44 pm

haha, no worries. it's a good exercise for me to go back and think about these things from time to time.

cheers

jonathon
User avatar
jonathon
 
Posts: 1115
Joined: Fri Jan 27, 2017 10:04 am


Return to General