Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

Re: st: Comparing coefficients across sub-samples

 From "Fitzgerald, James" To "" Subject Re: st: Comparing coefficients across sub-samples Date Fri, 3 Aug 2012 02:37:28 +0000

```Thanks again David.

On 3 Aug 2012, at 03:35, "David Hoaglin" <dchoaglin@gmail.com> wrote:

> James,
>
> The concept underlying "Lisa's formula" is quite basic; you should be
> able to find it in almost any introductory book on mathematical
> statistics or probability ("variance of a sum or difference [or, more
> generally, a linear combination] of independent random variables").
> It is not a "big" enough concept to be called a theorem, let alone be
> named after someone.  I should be able to dig up a reference if
> needed.
>
> The only part of the formula that might need justification is the use
> of the estimates for var(B1) and var(B2) and then referring z to a
> standard normal distribution.  That step relies on having "large"
> samples, but it should be OK, and it is very commonly done.
>
> David Hoaglin
>
> On Thu, Aug 2, 2012 at 9:51 PM, Fitzgerald, James <J.Fitzgerald2@ucc.ie> wrote:
>> David,
>>
>> One final thing; if I use this formula how would I reference or justify my use of it? Is it a basic formula I'd find in an econometrics text, or does it have a specific author, or is it founded on a particular theorem?
>>
>>
>> James
> *
> *   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/

*
*   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/
```