[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: correlation coefficient in the tobit model
I'm going out on a limb here, as I have not used the tobit, but my
understanding of the tobit is that there would be two predicted values.
One predicted value would be the probability of the observation being
censored. The other predicted value would be the expected (interval
level) value for the observation assuming it were not censored.
I'm on a limb because I presume one could calculate both predicted values
for every case in the dataset. Afterwards, though, how might one
calculate the correlation between predicted value and actual value--the
censored observations will not have a "true" "as if it was non-censored"
value. So, as I think about it, the challenge is what is the observed
value for censored cases--the usual reason one even estimates the tobit.
So, I guess that means, I am not sure what you would like to be able to
do. But, I'm out here on a limb, and maybe someone else can talk me down.
On Tue, 27 Feb 2007, lamia chourou wrote:
> Dear all,
> The xttobit command does not produce an adj. R-Sqd. or
> pseudo R-Sqd. Is there any cammand which gives the
> correlation coefficient between actual and predicted
> values of the dependent variable.
> Thank you,
> Découvrez une nouvelle façon d'obtenir des réponses à toutes vos questions !
> Profitez des connaissances, des opinions et des expériences des internautes sur Yahoo! Questions/Réponses
> * 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/
* For searches and help try: