confidence interval in one sample t test

General help and assistance with jamovi. Bug reports can be made at our issues page: https://github.com/jamovi/jamovi/issues . (If you're unsure feel free to discuss it here)

by Corabiche » Tue Jan 21, 2020 9:11 am

Hello,
I'm introducing psychology students to Jamovi. They have to use the free book "Learning statistics with jamovi".
When they follow the tutorial page 250 for one sample t test, they don't have the same results as mentionned in the book.
They obtain some strange results about confidence interval. I try myself and I have the same issue : the IC don't contain the sample mean. I looks like the calculating IC is not the mean IC (see screenshot attach).
Can you explain me if we're doing something wrong ?
Thanks a lot for your answers.
Attachments
jamovi_issue.JPG
jamovi_issue.JPG (73.62 KiB) Viewed 374 times
Corabiche
 
Posts: 4
Joined: Tue Jan 21, 2020 8:57 am

by jonathon » Tue Jan 21, 2020 9:45 am

hi,

the 95% confidence interval is on the difference between the population mean, and the test value. we changed this behaviour in the 0.9.5 version of jamovi, so the screenshot from the book must be from before we made this change.

i'll let david foxcroft know.

cheers

jonathon
User avatar
jonathon
 
Posts: 1059
Joined: Fri Jan 27, 2017 10:04 am

by David » Tue Jan 21, 2020 10:31 am

Hi, thanks for the heads-up. I've updated the book with the revised CI from version 1.05. This should be avaliable for download right away on learnstatswithjamovi.com.

Best.
D
David
 
Posts: 30
Joined: Wed Jun 13, 2018 10:18 am

by Corabiche » Tue Jan 21, 2020 10:43 am

I notice that if I use 0 in my test, i have some results more coherent for the IC (same as in the book). But if my test is not about 0, can I still see the IC for mean only ?

Thanks a lot.
Attachments
jamovi_issue2.JPG
jamovi_issue2.JPG (131.86 KiB) Viewed 368 times
Last edited by Corabiche on Tue Jan 21, 2020 12:47 pm, edited 1 time in total.
Corabiche
 
Posts: 4
Joined: Tue Jan 21, 2020 8:57 am

by Corabiche » Tue Jan 21, 2020 10:50 am

Hi,
Thanks for your answers.
I understand you change the IC type, and now we have no more the mean IC ?
So how can I have it ?
Thanks

(sorry for my bad english, I hope I am understandable)
Last edited by Corabiche on Tue Jan 21, 2020 12:48 pm, edited 1 time in total.
Corabiche
 
Posts: 4
Joined: Tue Jan 21, 2020 8:57 am

by David » Tue Jan 21, 2020 12:15 pm

Hi, if you check the 'Mean difference" option does that give you what you need?
David
 
Posts: 30
Joined: Wed Jun 13, 2018 10:18 am

by Corabiche » Tue Jan 21, 2020 12:45 pm

Mhhh no, sorry.
Actually I'm just confused about the value of the IC.
Why it is not the IC of the mean ? And how can I obtain it ?

If I understand you well, this is the IC of "sample mean - testing mean". So we have to check if 0 belongs to the IC. It's intersting, but if I want to check the mean IC, can I do it with the "one sample t test" procedure, or can I just do it using R ?

(I think it's ok if I write "0" in the test value, but I'm not completly sure about that ...)
Thanks again a lot !
Corabiche
 
Posts: 4
Joined: Tue Jan 21, 2020 8:57 am

by jonathon » Tue Jan 21, 2020 9:44 pm

yup, writing 0 in the test value will give you the CI you want.

cheers

jonathon
User avatar
jonathon
 
Posts: 1059
Joined: Fri Jan 27, 2017 10:04 am


Return to Help

cron