We have conducted study with intervention of exercise to measure muscle strength (6 variables) and balance (3 variables). The measurements had been taken at 5 regular intervals (pre, post 1, post 2, post 3 & post 4). Is regression analysis able to interpret the effect of muscle strength on balance measures? If not what would be an effective statistical analysis method?
Dear Sathiyaseelan Ganesan
I do think you can use canonical correlation analysis, focusing on your topic very well --- finding the relationship between two data group. Regression analysis is also OK, but you cannot consider the cross-covariance effects if only using regression approach. I find a manual ppt in English, using R, maybe it will be helpful to you. I have a little experience about canonical correlation analysis but what I used in my research is SPSS and I only have manual in Chinese. The imagine and way of interpretation can be read as the attachment
http://blog.sina.com.cn/s/blog_4a839b9a0100lysx.html
Do you combine your muscle strength and balance variables into new variables, representative for the construct?
Why not use the repeated measures anova, then? You can find this option in SPSS under General Linear Model. Repeated measures. Define two Time Factors with each five measure points, then use your intervention as a between subjects factor and put your combined, dependent variables as time-factors in the measure points.
Hi Ganesan
1. Your question is not very clear.
2. The analyses should be based on your hypotheses and research questions.
3. How many random variables in your research. How many dependent and independent random variables.
4. What is the relation between the random variables, i.e. related or not.
5. Regression analysis required a linear relationship between the independent random variables and the dependent random variable.
6.If the sample size of your random variables >30 you can use most of statistical tests.
7. If the sample size of your random variables
Dear Prof. Zuhair,
Number of Subjects: 24 (single group)
Independent Variable: Exercise
Dependent Variables: Three Balance measures- B3, Six Muscle Strength- M6
Measures: 5 interval measures of DV
Hypothesis: Correlation between muscle strength and balance.
By the way, you can also try this approach: Multinomial logistic regression.
As Z. A. Al-Hemyari noted, after your hypotheses and research questions are made, major idea is that you can define the Y (dependence variable) like this: 0,1,2,3,4 correspond 5 regular intervals (pre, post 1, post 2, post 3 & post 4) . Note that the 5 regular intervals have orders or quantitative scale or not is very important. Sometimes dummy variable definition is needed. Hope this can help you furthermore.
Sorry for those multi-posts. RG showed an error and I just realized that it actually got saved. :/
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Learning