[R-meta] Confidence intervals for effect sizes in pretest posttest control group designs

[Viechtbauer, Wolfgang (SP)]

There are tons of ways one can compute CIs for d-type effect size measures. Some of them are just minor variations, some are rather different. One can also compute an 'exact' CI (in the sense that it isn't based on a large-sample approximation and actually guarantees nominal coverage), but this requires iterative methods and usually isn't worth the trouble. In most cases, the approximation given below is just fine (unless sample sizes are very small, but then we are also in trouble when we apply standard meta-analytic models, which assume that the sampling variances are known without error).

And thanks for the kind words.

Best,
Wolfgang

-----Original Message-----
From: Célia Sofia Moreira [mailto:celiasofiamoreira using gmail.com] 
Sent: Monday, 13 May, 2019 22:53
To: Viechtbauer, Wolfgang (SP)
Cc: r-sig-meta-analysis using r-project.org
Subject: Re: [R-meta] Confidence intervals for effect sizes in pretest posttest control group designs

Dear Professor Wolfgang Viechtbauer,

Thank you so much for your great support!
I was expecting the answer to be something complicated because I took a look at your paper (but I didn't have the time to examine it):
Viechtbauer, W. (2007). Approximate confidence intervals for standardized effect sizes in the two-independent and two-dependent samples design. Journal of Educational and Behavioral Statistics, 32(1), 39–60.

I admire that even when you answer this too basic questions (too basic for you), you answer in an elegant way.
Science needs people like you! Thanks.

Kind regards,
csm

Viechtbauer, Wolfgang (SP) <wolfgang.viechtbauer using maastrichtuniversity.nl> escreveu no dia segunda, 13/05/2019 à(s) 09:49:
Hi Célia,

An approximate 95% CI can be obtained with:

dat$yi + c(-1,1) * qnorm(.975) * sqrt(dat$vi)

Best,
Wolfgang

-----Original Message-----
From: R-sig-meta-analysis [mailto:r-sig-meta-analysis-bounces using r-project.org] On Behalf Of Célia Sofia Moreira
Sent: Sunday, 12 May, 2019 23:18
To: r-sig-meta-analysis using r-project.org
Subject: [R-meta] Confidence intervals for effect sizes in pretest posttest control group designs

Dear all,

I would like to compute 95% confidence intervals for the effect sizes in a
pretest posttest control group design, such as the "Morris example"
presented in:
http://www.metafor-project.org/doku.php/analyses:morris2008

Can you please tell me how to do this?

Kind regards,
CSM