[R-meta] Effect Sizes From Pretest-Posttest-Control Group Designs and effect sizes from Pretest-Posttest designs

Yingkai

Mon, Dec 13, 2021 11:36 PM

Dear Dr. Wolfgang and all,


After reading Morris and DeShon, 2002, I have some new questions.



Morris provided an example of a meta-analysis based on raw-score and
change-score effect sizes (Table 3), however, this is an example of
independent-groups posttest design and single-group pretest?posttest
design. What about independent-groups pretest?posttest design? Can we
transform the effect size of independent-groups pretest?posttest design to
effect size of independent-groups posttest design or to effect size of
single-group pretest?posttest design? If we can, should we use equation 11
or 12 in Morris et al., 2002? The effect size of Independent-groups
pretest?posttest design is d (IG) or d (RM)?



Sorry for the interruption, just feel very confused about this.

Yingkai <yykjiyisuipian at gmail.com> ?2021?12?13??? 22:37???

Thanks a lot!

Viechtbauer, Wolfgang (SP) <wolfgang.viechtbauer at maastrichtuniversity.nl>?2021?12?13?
????10:36???

Dear Yingkai,

I don't think there is a single answer here. Of course one can do this --
whether one should in any particular case depends on the circumstances. And
not sure why combining effect sizes from pretest-posttest-control group
designs and effect sizes from between-group designs is out of the question.
In any case, a canonical reference here is:

Morris, S. B., & DeShon, R. P. (2002). Combining effect size estimates in
meta-analysis with repeated measures and independent-groups designs.
Psychological Methods, 7(1), 105-125.
https://doi.org/10.1037/1082-989x.7.1.105

Best,
Wolfgang

-----Original Message-----
From: Yingkai [mailto:yykjiyisuipian at gmail.com]
Sent: Monday, 13 December, 2021 15:08
To: r-sig-meta-analysis at r-project.org
Cc: Viechtbauer, Wolfgang (SP)
Subject: Effect Sizes From Pretest-Posttest-Control Group Designs and

effect

sizes from Pretest-Posttest designs

Hi Dr. Wolfgang and all,
   I am wondering whether one can combine effect sizes from

pretest-posttest-

control Group designs and effect sizes from pretest-posttest designs(no

control

group) in a meta-analysis?
I know that one should not combine effect sizes from

pretest-posttest-control

Group designs and effect sizes from between-group designs in a

meta-analysis.

Best,
--
Yingkai, Associate Professor.
Southwest University, ChongQing, CHINA.

--
Yingkai, Associate Professor.
Southwest University, ChongQing, CHINA.

*If you receive this email outside your office hours, please do not feel
obligated to respond immediately.*

Yingkai, Associate Professor.
Southwest University, ChongQing, CHINA.

*If you receive this email outside your office hours, please do not feel
obligated to respond immediately.*

	[[alternative HTML version deleted]]

[R-meta] Effect Sizes From Pretest-Posttest-Control Group Designs and effect sizes from Pretest-Posttest designs

Thread (4 messages)