I just had a look at the abstract. Reading more deeply the paper I found that authors used symmetrically trimmed least squares (STLS) estimator proposed by Powell (1986). To compare the parametric and semi-parametric estimations, they used Hausmans type specification test proposed in Newey (1987).
Sorry, but it is definitevely more a practical application of existing statistics than the development of new ones. I can provide you with the paper - obviously off the list - if you still find it interesting.
NEWEY, W. K. (1987), Specification tests for distributional assumptions in the Tobit model, Journal of Econometrics, 34 : 124�145.
POWELL, J. L. (1986), Symmetrically trimmed least squares estimation for Tobit models, Econometrica, 54 : 1435�1460.
P.S. Thanks for responding to my question on using different commands with the same name
Nicola
At 02.33 30/01/2007 -0500, "Nick Cox" wrote:
>I don't have access to this journal. What significance
>procedure justifies their claim?
>
>Nick
>[email protected]
>
>[email protected]
>
>> Elizabeth said:
>> Using former posts from Statlist, I've found I need to use
>> the command intreg because a Tobit will not produce robust
>> st. errors.
>>
>> I want just let you know something related to this issue
>> (althought no Stata code is provided):
>>
>> Quote from the abstract: "this paper attempts to apply
>> symmetrically trimmed least squares estimation as a
>> semi-parametric estimation of the Tobit model in order to
>> model firms' R&D expenditures with zero values. The result of
>> specification test indicates the semi-parametric estimation
>> outperforms the parametric ML estimation significantly."
>>
>> Full reference:
>> Scientometrics, Vol. 69, No. 1 (2006) 57�67
>> A semi-parametric modeling of firms' R&D expenditures with zero values
>> SEUNG-HOON YOO & HYE-SEON MOON
*
* 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/
