Effect sizes + confidence intervals for ANOVA posthoc effect

Discuss the jamovi platform, possible improvements, etc.
Post Reply
filination
Posts: 11
Joined: Fri May 12, 2017 5:18 am

Effect sizes + confidence intervals for ANOVA posthoc effect

Post by filination »

Dear JAMOVI,


(we're using JAMOVI with students in our mass replication project http://mgto.org/pre-registered-replications/ , and this is a common repeating nuisance JAMOVI request with almost every project, so this is on behalf of us all, not just myself).

I would like to request that you calculate and report Cohen's d effects and Cohen's d confidence intervals for posthoc/contrasts/marginal-means in an ANOVA. The np2/w only gives the overall differences between all conditions, but often times the most important effect of interest is the effect of the contrast between two effects. Sure, one could run a t-test and get that (alas, with no CIs, see comment below), but with 3 it's already 3 t-test to run, which gets even more complicated if you have an additional factor (two way ANOVA) or 4 or more conditions in a single factor, becoming a real chore. Simply adding d and d CIs to the existing tables would be terrific.

[I would also like to remind and reiterate the need for CIs for all effects reported (viewtopic.php?f=2&t=495). It's a shame to have to ask students to use online calculators to compute Cohen's d CIs outside of JAMOVI.]

Thanks,
Gilad
User avatar
Ravi
Posts: 194
Joined: Sat Jan 28, 2017 11:18 am

Re: Effect sizes + confidence intervals for ANOVA posthoc ef

Post by Ravi »

Hi Gilad,

Could you make an issue for this on our github: https://github.com/jamovi/jamovi/issues. I can definitely see the use-case for this feature, but unfortunately our time is also limited so have to see how quickly we can implement it.

Cheers,
Ravi
DrKnux
Posts: 1
Joined: Tue Nov 10, 2020 9:35 am

Re: Effect sizes + confidence intervals for ANOVA posthoc ef

Post by DrKnux »

Hi,

Very much a newbie here but, can I ask whether this has been remedied please?

Many thanks in advance
Knux
User avatar
jonathon
Posts: 2613
Joined: Fri Jan 27, 2017 10:04 am

Re: Effect sizes + confidence intervals for ANOVA posthoc ef

Post by jonathon »

hi,

it's better if you start a new post, rather than posting on a very old thread.

take a look in the 1.6 and see if it does what you need.

jonathon
Post Reply