Dummy Variable Help

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

Becca88222
Posts: 9
Joined: Fri Apr 30, 2010 4:11 pm

Re: Dummy Variable Help

Postby Becca88222 » Sat May 01, 2010 9:38 am

YAY! Thanks!!!


Okay my final question is:

For dummy variables can I only use the numbers 0 and 1 or can I have 4 different education levels (college, advanced, high, junior) and number them 0,1,2,3 respectively?

startz
Non-normality and collinearity are NOT problems!
Posts: 3796
Joined: Wed Sep 17, 2008 2:25 pm

Re: Dummy Variable Help

Postby startz » Sat May 01, 2010 9:45 am

0 and 1 only.

Dataminer
Posts: 11
Joined: Thu Jul 23, 2009 2:01 am

Re: Dummy Variable Help

Postby Dataminer » Sat May 08, 2010 7:31 am

Becca88222
can I only use the numbers 0 and 1 or can I have 4 different education levels (college, advanced, high, junior) and number them 0,1,2,3 respectively?
You can certainly do this as far as OLS is concerned. The issue is its interpretation. Implicitly you are saying that the effect of this education variable on the depvar. is the same difference between cases with values 0 and 1 ('college' an 'advanced') as it is is between 2 and 3 ('high' and 'junior').This is seldom credible!
More typically one creates a dummy variable for each of the mutually exclusive and exhaustive categories (here 4?). For each case these dummies will either take value 1 if the category applies or 0 if it doesn't. Then you can either put all dummmies into the estimating equation and no constant so that the coefficient estimates indicate the effect of each category on the depvar. Also putting the intercept into the model here creates the 'dummy variable trap' (see any econometrics text) i.e. perfect multicolinearity. Alternatively use the intercept and dummies but leave one dummy out. The intercept measures the effect of the omitted category and the other dummy coefs measure the marginal effect of each category over the omitted one. Both methods produce identical (interpretative) results for a common dataset.


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 2 guests