MAJOR correlation (meta-analysis)

Everything related to the development of modules in jamovi
Yucheng
Posts: 3
Joined: Thu Apr 26, 2018 3:13 am

MAJOR correlation (meta-analysis)

Post by Yucheng »

Dear All,

I am a fan of Jamovi. It is user-friendly and awesome! I am now using it for my meta-analysis, so I downloaded the MAJOR from Jamovi library.

There is a question about using MAJOR to run correlation analysis: the moderator column only accepts continuous variables. Are there any ways for me to put categorical (nominal text) variables in that column? Look forward to your reply!

Best,
Yucheng
User avatar
kylehamilton
Posts: 29
Joined: Fri Mar 24, 2017 8:47 am
Location: Merced, California, USA
Contact:

Re: MAJOR correlation (meta-analysis)

Post by kylehamilton »

Hey Yucheng,

Right now MAJOR only supports continuous moderator variables. But I can add categorical variable support to the development version sometime next week. For now, if you have just two categorical variables you can code them as 0 and 1 and MAJOR will produce the same results. This won't work for three or more categories but I'll add the support for that when I get back home.

Are there any other features that you need that are missing?

Kyle
Yucheng
Posts: 3
Joined: Thu Apr 26, 2018 3:13 am

Re: MAJOR correlation (meta-analysis)

Post by Yucheng »

Thank you so much, Kyle! It would be very helpful to add categorical variable support. Since I have quite a few categorical variables, I cannot simply dummy them. I look forward to the new version!

Thank you again for your prompt reply and assistance.

Best,
Yucheng
User avatar
kylehamilton
Posts: 29
Joined: Fri Mar 24, 2017 8:47 am
Location: Merced, California, USA
Contact:

Re: MAJOR correlation (meta-analysis)

Post by kylehamilton »

Hey Yucheng,
I've added the feature you've requested to MAJOR. Right now it can't just take a set of text labels and turn them into a categorical variable but you can use a numbering system to achieve the same results. So if you make a new collum and code the different variables to 0,1,2,3,4,5 then go to the model options menu and change the moderator from continuous to categorical. The updated version is not in the Jamovi library right now but it is on the project github page (https://github.com/kylehamilton/MAJOR) and can be downloaded here https://github.com/kylehamilton/MAJOR/r ... /MAJOR.jmo just download the file and sideload it into Jamovi.

Let me know if there are any problems. I'll try to get the label issue taken care of once the semester is over.

Kyle
Yucheng
Posts: 3
Joined: Thu Apr 26, 2018 3:13 am

Re: MAJOR correlation (meta-analysis)

Post by Yucheng »

Thank you for your quick response and support! I tried to follow your guidance to change the moderator from continuous to categorical. However, the results do not vary with the options I selected. I mean even after I selected categorical, the coefficients in the table still remain the same as before (when I selected continuous). I am not sure whether I got the correct results.

Thanks,
Yucheng
User avatar
kylehamilton
Posts: 29
Joined: Fri Mar 24, 2017 8:47 am
Location: Merced, California, USA
Contact:

Re: MAJOR correlation (meta-analysis)

Post by kylehamilton »

Hey Yucheng, I’m away from my computer right now I’m visiting family but if you send me your data file I’ll take a look when I get home. Email me at kyle.hamilton@gmail.com

Kyle
danacotto
Posts: 2
Joined: Tue Jun 05, 2018 6:20 pm

Re: MAJOR correlation (meta-analysis)

Post by danacotto »

I downloaded the new mod for categorical moderators in meta analysis and now the mod does not work. Just thought you should know. I will use your initial suggestion and code the categories instead.
User avatar
kylehamilton
Posts: 29
Joined: Fri Mar 24, 2017 8:47 am
Location: Merced, California, USA
Contact:

Re: MAJOR correlation (meta-analysis)

Post by kylehamilton »

Hey danacotto,

Yeah, I'm working on this right now. I got a bit sidetracked on a few other features. I'll post when I get the rest of this put together.
danacotto
Posts: 2
Joined: Tue Jun 05, 2018 6:20 pm

Re: MAJOR correlation (meta-analysis)

Post by danacotto »

Hi! Just wanted to check in to see if there was any update on the continuous moderator issue. I have multiple moderators with more than two categories so I'm a bit stuck.
User avatar
kylehamilton
Posts: 29
Joined: Fri Mar 24, 2017 8:47 am
Location: Merced, California, USA
Contact:

Re: MAJOR correlation (meta-analysis)

Post by kylehamilton »

Hey Danacotto,

Sorry for taking so long my time got taken over with grad school. Yes, this is now fixed in the beta version https://github.com/kylehamilton/MAJOR
Post Reply