Page 1 of 1

Feature request: Confidence intervals for Reliability test

Posted: Wed Jul 22, 2020 7:23 am
by Vit
Hi,

I am glad there are some options how to analyse reliability in jamovi.

Would it be possible to add Confidence interval for their estimates?

Thanks.

Re: Feature request: Confidence intervals for Reliability te

Posted: Thu Jul 30, 2020 7:50 am
by Vit
I'll just add to this. In the current version 1.2.25 and updated modules, I have found three different tools to analysis reliability (in Factor, ClinicoPath, Seolmatrix). Wouldn't it be possible (more convenient) to combine them into one new reliability/interrater agreement module?

Re: Feature request: Confidence intervals for Reliability te

Posted: Thu Jul 30, 2020 8:29 am
by sbalci
Do you have any suggestion for confidence interval calculation. An R code or package? Sometimes there are multiple ways and I am hesitant to add them.

For duplicates:
For ClinicoPath I admit that I have made some duplicates, and I intend to add some more. I think from a clinician/pathologist perspective and try to give options that suits their needs. Actually I add features as I do my pathology research and think if that R code will work in jamovi. It is liberty of R to do same thing in various way, and it is also the curse :grin:
For example DeathWatch is a very nice survival analysis tool, but I made another survival function because that is the template we use in our research.

Re: Feature request: Confidence intervals for Reliability te

Posted: Wed Aug 05, 2020 2:23 pm
by Vit
No, unfortunately, I don't.

I see that for people capable of writing necessary code, jamovi is an excellent toy for their needs :-) and everybody does have slightly different needs based on their fields.

In the long run, I am a bit worried about how the overlap between modules can fragment jamovi... But I guess one needs to keep only the relevant modules for his/her work :-).

Re: Feature request: Confidence intervals for Reliability te

Posted: Fri Aug 07, 2020 2:05 pm
by sbalci
I would not thing it would break jamovi :D but may cause confusion, you have a point :rolling_eyes:

Re: Feature request: Confidence intervals for Reliability te

Posted: Mon Sep 07, 2020 6:12 pm
by farscape1
I use MBESS package in R to calculate confidence intervals. It would be nice if one could do it for item-rest correlations as well when dropping each item for alpha and omega.