Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: weird margins


From   Austin Nichols <austinnichols@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: weird margins
Date   Fri, 2 Oct 2009 11:23:49 -0400

Just guessing, but do you mean
 predict(ystar(0,.))
instead of
 predict(ystar(o,.))
?

On Thu, Oct 1, 2009 at 3:31 PM, Rich Steinberg <rsteinbe@iupui.edu> wrote:
> I run a tobit because of left censoring at zero, then want to use margins to
> predict the marginal slope for the observable dependent variable atotgiv.
>
> Everything looks ok except in the output appears
> Expression: E(atotgiv*|atotgiv>otherrelh), predict (ystar(0,.))
>
> The appearance of otherrelh in this expression is mysterious.  This is
> another dummy variable used as a control.  I didn't program it to look at
> that, and I think it should read, instead, E(atotgiv*|atotgiv>0), etc.
>
> Is this a bug?  A mistake on my part?  A misunderstanding?
>
> Here is the exact command and headers for the results:
>
> margins, predict(ystar(o,.)) dydx( earn01 assety01 welfare01 nonwelfare01
> annuapprecinh annuownwealthn annuheq )
>
> Average marginal effects                          Number of obs   =
> 4935
> Model VCE    : Robust
>
> Expression   : E(atotgiv*|atotgiv>otherrelh), predict(ystar(o,.))
> dy/dx w.r.t. : earn01 assety01 welfare01 nonwelfare01 annuapprecinh
>              annuownwealthn annuheq
>
> --

*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index