Hello members,
I have a problem in running my regression with dummy variable. can someone help me out ? because i google it, most of it teach how to create dummy.
thank you.
What is the problem you are facing? Also please include your regression equation.
What problem? Also what is your regression equation and variables?
If the outcome is continuous or number, using general linear model,
For any software, it should have a function to classify the independent variable as continuous or categorical, for example, in SAS
proc GLM;
class gender;
model y=gender x;
here x is continuous, gender is dummy.
If you use proc reg, which doesn't take categorical predictor, you need generate
numerical dummy variables for categorical data,
for gender, it has two levels, you need generate one dummy;
if sex='Female' then iffemale=1; else iffemale=0;
then you can use
proc reg;
model y= iffemale x;
you will get same results as you did by using Proc GLM.
If you have three levels, say race= white, black, others,'
you need generate two dummy variables:
ifwhite=1 race=white, otherwise ifwhite=0
ifblack=1 race=black, otherwise ifblack=0
proc reg;
model y=ifwhite ifblack x;
or
proc GLM;
class race;
model y=race x;
if outcome is categorical, you need use different model, such as logistic, etc.
Thank you for the respond.
I will create a dummy based on prices. Through the prices of the stock for six years, i will categorize it below $0.10 and above $0.10. So what would be the model can be used or how to regress it using eviews
this is my simple regression equtation
Y_i=β_0 + β1 X1i+ԑi
y=trading volume
β1 = prices
β2= dummy
hope someone out there can help..thank you
if x1>0.1 then phigh=1; else phigh=0;
the model is
Y= β0 + β1 phigh+ԑ
in fact β1 is the mean difference of stock(>0.1)-stock(
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression