I am recently started my empirical thesis work in some addiction related topic. I have define 6 independent variable 3 moderating variable and 1 dependent variable. So, i want to perform MANOVA instead logistic regression. I think Logistic regression is used to study impact but i want to study about the online game addiction in student.
My supervisor suggest me to do MANOVA and restric me to perform ANOVA and logistic regression. Is there possible of MANOVA in my case (6 independent, 3 moderating, 1 dependent vairable)? I have defined "game addiction tendency" as my dependent variable.
So please suggest me how to perform multivariate analysis ?
If MANOVA is possible please suggest me a best and step by step procedure example of MANOVA.
Thank you in advance.
Everything you wrote is very confusing.
1) If you have only one dependent variable, you can not perform a MANOVA (even if the rest of your design would fit ), since it estimates the effect of your independent variables in more than 1 DV simultaniously. So, if you only have 1DV this would reduce to an ANOVA.
2) what is the point with the logistic regression? If your DV is categorical, this would be the correct approach. If it is continous, it wouldnt be. So, how does your DV look like?
3) your design is not clear. You have 9 IVs (including moderators)??? Are they all categorical? A 9 factorial ANOVA would be absolutely ludicrous!! Not only a useful sample size, but also the interpretation would be crazy.
I think what you are looking for is some kind of Structural Equation Modeling, but since you apparently lack fundamental skills, you should consult a local statistician to help you. Basic education is beyond the scope of this forum.
Dear Som Nath Adhikari ,
Rainer Duesing is right about it not being MANOVA. The alternative to a factorial ANOVA is a multiple linear regression. If you were to use an automated technique you might be able to reduce the number of IVs. You can also test for multicollinearity.
If the data structure is like this: enter the dependent variable along the six independent variables as columns and the three moderating variables as rows, then this is a case of MANOVA. You may use MINITAB for it. Click on DATA. Click on CODE. Click on TEXT to NUMERIC if you called the set of three moderating variables SET with text names. Recode SET by SET1 with numeric values. Then click ANOVA then GENERAL MANOVA. Put VARIABLES as the six independent variables and METHOD as SET1. You are done. Good luck!
Lets be clear, the M for "multiple" in MANOVA applies to multiple dependent variables. If there is only one dependent variable, then MANOVA is not appropriate.
- Badges
- Science topic