Re: Re: st: Question on estimation procedure

[Nick Cox <njcoxstata@gmail.com>]

This looks like the same question in essence, even though your model
changes from post to post.

If you disagree with Kit's analysis, please explain why he is wrong.

Otherwise if something is impossible in principle, you can't do it in
Stata [NB].

Nick

On Tue, Nov 13, 2012 at 4:56 PM, Elena Quercioli
<elena.liquorice@gmail.com> wrote:

> My question was meant to get at STATA (estimation) capabilities and
> not identifiability; here, in essence, is what I want to know. I want
> to simultaneously estimate the common base and other coefficients of
> this geometric function:
>
>  Y = c_0 + c_1[X_1+a^d2 X_2 + a^d3 X_3]
>
> any ideas?

On Mon, Nov 12, 2012 at 9:15 PM, Christopher Baum <kit.baum@bc.edu> wrote:

>> Elena said
>>
>> Oops, my bad. I omitted to state that d1=0.
>>
>> Would you have any suggestions on how to estimate the non-linear model
>> I illustrated?
>>
>>
>>
>> No, the model
>>  Y = c_0 + c_1  X1 + c_2 a^d2 X2
>> is still underidentified, even with the restriction that removes the alpha^0 term. Consider
>>
>> sysuse auto, clear
>>
>> forv c2 = 0.75(0.05)1.2 {
>>         nl (price = {c0} + {c1} * mpg + `c2' * {alpha}^2 * weight)
>>         nlcom `c2' * ([alpha]_cons )^2
>> }
>>
>> As you can see, there is a locus of c2 and alpha values that render the same product (c2 * alpha^2),
>> so the data cannot identify c0, c1, c2, alpha from three data vectors: iota, mpg, weight.
>>
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/