Doubt regarding Multiple Regression

#1
Hi,
I have some doubt in Multiple regression.
I have Y as the dependent variable (mutual fund returns)

I have many independent variables such as

1)Age of Manager
2)Sat score
3)MBA/Non-MBA.


Using multilple regression, I have to develop a relationship between the value of mutual fund and the qualities of manager (i.e age,sat score,mba/non mba).

My doubt is that I have interval data for age and sat score but for MBA/Non MBA , the data is nominal i.e (0-Non MBA , 1 -MBA).

How do i consider this nominal data while calculating multiple regression.

I am using excel.Do I need to consider MBA/Non MBA.IF so ,how?

Kindly help
 
#2
Its not an issue for prediction. And may not be an issue for statistical significance; you have to evaluate residual assumptions there.

Catagorical variables in general do not upset the regression framework in anyway continous variables do not when they are in the predictors. Predicting catagorical variables is a whole different ball of wax because the normality assumption on the error term is automatically violated.
 
#3
To use a categorical variable as a predictor in Multiple Regression, you must create a dummy coded variable.

Basically, create a single new variable and call it dummy_MBA or something relevant.

Now, within this column, assign a value of "1" if the person has an MBA, and a "0" if they don't have the MBA. Then use this new variable in the multiple regression analysis.

good luck.