Re: st: OLS versus Ordered Probit

[Maarten Buis <maartenlbuis@gmail.com>]

On Fri, Dec 9, 2011 at 3:59 AM, Venkiteshwaran, Vinod wrote:
> This question is not about stata per se but more of an econometrics query.

Sounds to me more like a statistical question...

> What are the strengths/drawbacks of using OLS, as opposed to ordered probit or logit, to estimate a model of ordered choices?

That depends on the nature of the dependent variable. One potential
problem with linear regression (linear regression is the model, OLS is
only the method used to compute the coefficient) can be defining a
meaningful scale for your dependent variable. Some variables are a
coarsened version of a continuous variable others are truly ordered
variables. In the former case you can often derive a more or less
meaningful scale for your dependent variable, in the latter case that
is usually impossible. Say you have a question in the form of a
statement and the respondent can choose between 5 categories from
strongly agree to strongly disagree, than assigning those categories
the values 1 till 5 and using linear regression will assume that the
distance between the 5 categories are all equal, which is usually
problematic. Ordered model relax that assumption.

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany


http://www.maartenbuis.nl
--------------------------
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/