Re: st: explanation of stdf

[Kit Baum <baum@bc.edu>]

On Apr 8, 2010, at 2:33 AM, Ariel wrote:

> I am not sure if it is a matter of nomenclature, but I am a little confused
> on the meaning of the stdf (found in predict). More specifically, I am not
> sure what I get if I build 95% CIs using this value. It is not really a
> Confidence Interval, it is not a Prediction Interval, and it is definitely
> not a Tolerance Interval.
> 
> Here is the stata definition: 
> 
> 
> "stdf calculates the standard error of the forecast, which is the standard
> error of the point prediction for one observation.  It is commonly referred
> to as the standard error of the future or forecast value. By construction,
> the standard errors produced by stdf are always
> larger than those produced by stdp; see [R] regress postestimation Methods
> and formulas."


from IMEUS, chapter 4 section 6:

This alternate interval estimate may be computed with predict’s stdf (standard error of forecast) option, which is applicable for a point prediction for a single observation (see [R] regress postestimation). In contrast to stdp, which calculates an interval around the expected value of y for a given set of X values (in- or out-of-sample), stdf takes the additional uncertainty associated with the prediction of a single y value (i.e., \sigma^2_u) into account. A confidence interval formed with stdf may be used to evaluate an out-of-sample data point, y0, and formally test whether it could plausibly have been generated by the process generating the estimated model. The null hypothesis for that test implies that it should lie within the interval yˆ0 ± t stdf.

Kit Baum   |   Boston College Economics & DIW Berlin   |   http://ideas.repec.org/e/pba1.html
                              An Introduction to Stata Programming  |   http://www.stata-press.com/books/isp.html
   An Introduction to Modern Econometrics Using Stata  |   http://www.stata-press.com/books/imeus.html


*
*   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/