Three thoughts:
(1) Read almost any book on regression - that will cover it.
(2) If you are using SPSS, using the GLM - Univariate command. If you
have categorical predictor variables, stick them into the factors box,
if you have continuous variables, stick them in the covariates box.
Click on Options and request parameter estimates. Hurrah! SPSS dummy
codes all the variables for you. Similarly, in logistic, just declare
your variables as categorical.
(3) There is SPSS code to create dummy variables at:
http://www.spsstools.net/Syntax/LabelsAndVariableNames/CreateDummyVariables.txt
Jeremy
Colette Reynolds wrote:
> Hi there
> Hope everyone has survived January alright.
>
> I am currently doing Regression analysis on my data (multiple linear &
> logistic) and have been asked by my supervisor to recode my predictor
> variables using dummy variable coding. I cant seem to find much information on
> this- particularly why it should be used and how to do it. My supervisor isn’t
> approachable at all so I’ll have to figure this out myself. Can anyone offer
> some help regarding where I can find out more about dummy variable coding.
> Many thanks
> Colette
>
>
>
--
Jeremy Miles
mailto:[log in to unmask] http://www-users.york.ac.uk/~jnvm1/
Dept of Health Sciences (Area 4), University of York, York, YO10 5DD
Phone: 01904 321375 Mobile: 07941 228018 Fax 01904 321320
"No fair, you changed the outcome by measuring it!"
- Professor Hubert J. Farnsworth
|