Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Predicted probabilities after oprobit w/robust standard errors


From   Matt Barmack <[email protected]>
To   [email protected]
Subject   Re: st: Predicted probabilities after oprobit w/robust standard errors
Date   Fri, 2 Jun 2006 10:36:16 -0700 (PDT)

Richard,

Thanks for your detailed responses.  I guess I am having trouble
translating my intuition from linear models.  My understanding is
that  the use of robust in a linear context results in an uu' matrix
in which the diagonal elements are potentially different, i.e. the
error term for each observation is drawn from a normal distribution
with mean zero and a variance that may differ from observation to
observation.  In oprobit w/robust (leave aside cluster), what does it
mean to say that all of the u_j's are normal(0,1) but, when it comes
to calculating standard errors for parameter estimates, the diagonal
elements of the uu'are potentially different from one another?

Thanks again,

Matt

--- Richard Williams <[email protected]> wrote:

> At 06:55 PM 6/1/2006, Matt Barmack wrote:
>>Specifying cluster or robust does not seem to change the predicted
>>probabilities from oprobit.  Does it?  Shouldn't it?
>>Intuitively/naively, I am thinking that for an observation for
> which
>>the variance of the random part of the latent index is high, there
> is
>>a greater chance of ending up further away from what the
>>deterministic part of the latent index alone might suggest.
>
> One other clarification: by "variance of the random part of the
> latent index" I assume you mean the residual.  In Probit the
> residual
> is assumed to have a Normal(0, 1) distribution.  The linear
> prediction is an estimate and is subject to sampling variability.
> I
> suppose if one were so inclined, you could compute the confidence
> interval for that estimate, and then based on that estimate come up
>
> with a range for the predicted probabilities.  I don't recall ever
> having seen that done.  Nor am I exactly sure how you would do it,
> since the estimates of the cutpoints are themselves subject to
> sampling variability.
>
>
> -------------------------------------------
> Richard Williams, Notre Dame Dept of Sociology
> OFFICE: (574)631-6668, (574)631-6463
> FAX:    (574)288-4373
> HOME:   (574)289-5227
> EMAIL:  [email protected]
> WWW (personal):    http://www.nd.edu/~rwilliam
> WWW (department):    http://www.nd.edu/~soc
>
> *
> *   For searches and help try:
> *   http://www.stata.com/support/faqs/res/findit.html
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index