The above question is my basis for a randomised controlled study. My DV is working memory my IV is exercise group (2 levels cognitive, physical) if I want to compare with no exercise group. Am thinking along the lines of ANOVA but not sure if it should be one way between groups? New to statistics, or is there a better study alternative?
Dear Lyn,
The authors are here on RG:
FITNESS EFFECTS ON THE COGNITIVE FUNCTION
OF OLDER ADULTS:
A Meta-Analytic Study
Stanley Colcombe and Arthur F. Kramer
Thankyou. However, not having done a meta-analysis study, although I have read up about it. There are several pitfalls from what I can see/understand. My main question here would be how you combine the information on these studies. I understand they should be of similar type ie all randomized controlled studies but if you took 5 such articles, each have say different age groups they are assessing how would you be able to utilise this information as one group may be 65-75, another 60-70?
You can proceed along the way you have thought. With three levels of the IV, i.e., three exercise groups, you can go for one way between groups ANOVA, followed by a post hoc test.
It comes to my mind that Juan Pascual-Leone & coll. have a recent paper on the effect of tai chi. I cannot find that paper here on RG, perhaps it is still in press. But the main author is on RG, so you could contact him for that paper.
ANOVA would be fine, but the design is a bit confusing: do you have two intervention groups (cognitive and physical)? Do you have a passive control group? If you do, then your IV would have 3 levels.
BTW, I'm attaching a recent paper by our team on WM and exercise in case it is of any help.. Good luck
According to your description, you will have three groups: cognitive training, physical training and control). You will have at least a pretest and a postest evaluación. You can run a 2 (pre, post) by 3 (physical, cognitive and control) factor ANOVA.
Hi, you will need a mixed factorial ANOVA. IV - 3 between participant levels (cognitive/physical/none) x at least two within participant levels of DV (if you do not have a baseline measure how will you know if anything has changed?). I would suggest that you may also wish to have more than one type of DV to measure working memory but either way you will need to use factorial analyses.
- Badges
- Science topic
- Similar topics
- Correction