[R-meta] Meta analysis subgroups

[Viechtbauer, Wolfgang (SP)]

Hi Julie,

If that is the hypothesis, then looking at the association between the SMD for X and the SMD for Y doesn't answer it. You should be looking at the association between X and Y within each study (and group) directly.

Best,
Wolfgang

-----Original Message-----
From: Julie Webbs [mailto:webbsjulie using gmail.com] 
Sent: Thursday, 05 September, 2019 9:14
To: Viechtbauer, Wolfgang (SP)
Cc: Michael Dewey; r-sig-meta-analysis using r-project.org
Subject: Re: [R-meta] Meta analysis subgroups

Thank you,  Wolfgang

The research hypothesis is that the variable Y does not affect the variable X. Or does the variable Y  affect the variable X? 

Hope this clears things up.

Best Wishes,
Julie

On Wed, Sep 4, 2019 at 9:59 PM Viechtbauer, Wolfgang (SP) <wolfgang.viechtbauer using maastrichtuniversity.nl> wrote:
Hi Julie,

You have clarified the data structure, but I personally still have no idea what you mean by: "Now I want to run a met-analysis for X differences associated with Y." What exactly do you mean by that? What is your research question/hypothesis?

Best,
Wolfgang

-----Original Message-----
From: R-sig-meta-analysis [mailto:r-sig-meta-analysis-bounces using r-project.org] On Behalf Of Julie Webbs
Sent: Thursday, 29 August, 2019 17:00
To: Michael Dewey
Cc: r-sig-meta-analysis using r-project.org
Subject: Re: [R-meta] Meta analysis subgroups

Dear Michael

Thank you for your email. and sorry about that.

I have to variables, say X and Y.  

1. The X has two categories, let's say  a and b, so you  can calculate Mean and SD for group a and group b. 

2. The Y has two categories, let's say c and d. So you can calculate Man and SD for group a and group d.

I have run a meta-analysis separately for both X and Y using the meta package. 

Now I want to  run a met-analysis for  X differences associated with Y. Is it possible? or I need to consider other statistical procedures? 

Here is the excel file

Thank you
Julie

On Thu, Aug 29, 2019 at 1:28 PM Michael Dewey <lists using dewey.myzen.co.uk> wrote:
Dear Julie

Sorry bit that does not make much sense to me. See comments in-line

On 28/08/2019 21:42, Julie Webbs wrote:
> Dear All
> 
> I have two continuous variables and each has two categories as follows:

If they are continuous then they do not have categories so I am at a 
loss here.

> X = a and b
> Y= c and d
> 
> I have calculated  the standardised mean difference (SMD)  for  *X* and *Y
> using meta-analysis.*
> I want to calculate *X* differences associated with *Y*.

That also seems opaque here.

> Would you please advise me how I can do it. Here is a sample of data. Thank
> you, Julie

I am afraid your data-set did not arrive here well since you posted in 
HTML and such posts get mangled when they pass through the list. It 
might also help if you posted an example of what the differences you are 
seeking would look like.

> X Y
> Mean SD Mean SD Mean SD Mean SD
> a a b b c c d d
> 12 1 14 2 10 0.3 18 0.1
> 11 2 13 0.5 11 0.1 14 0.3
> 10 1 12 1 14 1 16 0.2