View Full Version : Hierarchical Regression Results


nadia
04-26-2006, 05:04 PM
Hi John (and to everyone esle as well, but thusfar John has been the one answering my questions :) ),

I have finally finished running the regression analysis for my report. I am unsure of my results and I am not sure if it is a flaw in my model or whether my results are accurate, just not as strong as I wanted.

I used a hierarchical model to predict my dependent variable Political Participation (additive index) by my three independent variables (formal group involvement, voluntary association membership, and orgranized group interactions) while controlling for gender, age, race, income, and interest in politics.

The R Square is .420, which I interpret as my model predicting 42% of the total variance in Political Participation, and a moderate result. The R Square change is .056, which I interpret as the addition of my various controls to the model only accounts for an additional 5.6% of the variance in Political Participation. Therefore, am I to assume that the controls did little to the prediction and my three dependent variables account for the majority of the variance in political participation in this model?

Your assistance has help me several times before, and I do appreciate this resource. I am so close to finishing my paper and I do thank you for all of the help.

Regards,
Nadia

JohnM
04-26-2006, 05:17 PM
Nadia,

Actually, 42% is pretty darn good in social/behavioral science studies....humans are really hard to predict....especially considering that the additive index was your own idea, and there were only 3 predictors. Nice job!

Yes, the controls didn't change much, but I would briefly mention it.....

nadia
04-26-2006, 05:29 PM
John thank you, thank you, thank you!

I was so worried that I had done all of this work and I wasn't interpreting it correctly, or that my model was flawed (i.e., my paper would fail and graduation goes down the drain).

There are so many resources out there that give you the nuts and bolts of statistical methods, but it is very hard to find somewhere that tells you how to apply and interpret your results.

You have helped me in several steps of this project and I do thank you whole heartedly.

Warm regards,
Nadia