This is my first attempt at a meta-analysis. Advice would be much appreciated!
I am looking at the effects of a specific mindfulness intervention on emotion regulation looking at the DERS questionnaire as the measure of emotion regulation.
I am analysing both the subscales scores and total scores.
I have 5 out of 10 papers that have provided the pre and post means and SDs for the subscale scores and 8/10 papers who have provided the pre and post means and SDs for the total scores.
How would I go about analysing this data? 3 studies have more than one mindfulness intervention condition and one has no control group but multiple mindfulness intervention conditions.
Hi Michelle,
One meta-analysis would not be sensible for your data. I would definately seperate the subscores and total scores in different analyses. I would also only considering pooling data if they have a similar control group (e.g. doing nothing or standard of care). And lastly, I would perform seperate analyses per type mindfulness intervention conditions. The multiple arm studies share a control group and are therefore dependent (correlated). I thinkt that it would be wrong to include the multiple arms as independent studies in the same analyses. The variance in the model would be underestimated.
As every study uses the DERS questionnaire, and if the studies demonstrate baseline equivalence, you can easily use the mean difference of the post means (control vs intervention group) as effect size.
Good luck,
Bastiaan
Hi Bastiaan!
Thank you for taking the time to provide guidance. That was really very helpful.
Wishing you a lovely week!
Obviously reviewing and analyzing the maximum number of scientific publications of this type: whether or not they have a "Control Group" is not your problem or responsibility, you only do the Meta-analysis OF WHAT IS, NOT OF WHAT YOU WOULD LIKE TO HAVE ... but, be careful, YOU CANNOT MIX THOSE THAT DO HAVE A CONTROL GROUP WITH THEM!
- Badges
- Science topic
- Similar topics
- Psychology
- Emotion