jamovi
Paired samples - categorical data
Hi, My next contribution I propose is to improve the paired samples analyses for categorical data. The McNemar test can have improved effect estimates with
- odds-ratio & (95%CI)
- absolute differences & (95%CI)
In this same page I could add agreement statistics:
- observed & expected agreement
- Kappa (95%CI)
- Weighted Kappa for >2x2 can be considered
Though the agreement could be a module instead of jmv
the tension we face is providing features, but not cluttering things too much ... but the mcnemar is so sparse i think we'd be happy to have all these things added.
Hi Jonathon, Do you have a list o new features that that are required, so I can check if I can help with any of them Or these just the requests on the blog and github issues
Victor
hey,
yeah, i think it's mostly on our gh issues page: https://github.com/jamovi/jamovi/issues
if there's things there you could contribute to, that would be awesome! butit might be worth checking with ravi that he's not opposed to something before you do it. ravi and i are the opposite, i want to give people everything, but ravi is more disciplined and principled than i am.
my biggest one would be CIs in descriptives ... it would get a lot of people off our backs!
shall i invite you to our slack group? that might be the easiest way to discuss these things.
cheers
jonathon
Hi Jonathon, Please invite me to slack so we can comment easier Victor