st: New version of the -parmest- package on SSC

[Roger Newson <roger.newson@kcl.ac.uk>]

Dear All

Thanks to Kit Baum, there is now a new version of the -parmest- package 
downloadable from SSC. Type -ssc describe parmest- inside Web-aware Stata 
to find out more.

The -parmest- package converts estimation results to a data set with one 
observation per parameter, or one observation per parameter per by group, 
and data on confidence intervals, P-values and other parameter attributes. 
This new data set can be listed to the Stata log, saved to disk, or written 
to memory (overwriting any pre-existing data). The new version is still 
Stata 7, but has a new option -clnumber-, controlling the numbering of the 
variables in the output data set containing the confidence limits. The 
-clnumber- option has 2 possible values, -level- (the default) and -rank-. 
If -level- is specified, then the confidence limits are numbered by their 
confidence levels. If -rank- is specified, then the confidence limits are 
numbered by their ranks in ascending order of confidence. Therefore, if the 
user types

parmby "regress mpg weight foreign",norestore level(90 95 99 99.5) clnum(level)

or, equivalently,

parmby "regress mpg weight foreign",label norestore level(90 95 99 99.5)

then the 90%, 95%, 99% and 99.5% lower confidence limits will be in 
variables named min90, min95, min99 and min99_5, and the corresponding 
upper confidence limits will be in variables named max90, max95, max99 and 
max99_5. However, if the user types

parmby "regress mpg weight foreign",norestore level(90 95 99 99.5) clnum(rank)

then the lower 90%, 95%, 99% and 99.5% confidence limits will be named 
min1, min2, min3 and min4, and the corresponding upper confidence limits 
will be named max1, max2, max3 and max4. The -clnum(rank)- option can be 
useful if the confidence levels have a lot of digits after the decimal 
point, which might be the case if the user is defining Bonferroni-corrected 
or Sidak-corrected confidence intervals. For instance, if the confidence 
intervals are Sidak-corrected for 5 multiple confidence intervals, then the 
corrected confidence level will be equal to

100*(1-0.05)^(1/5) = 98.979378

In this case, the user might prefer the corrected confidence limits to be 
called min2 and max2, instead of the default names min98_979378 and 
max98_979378.

Best wishes

Roger

--
Roger Newson
Lecturer in Medical Statistics
Department of Public Health Sciences
King's College London
5th Floor, Capital House
42 Weston Street
London SE1 3QD
United Kingdom

Tel: 020 7848 6648 International +44 20 7848 6648
Fax: 020 7848 6620 International +44 20 7848 6620
  or 020 7848 6605 International +44 20 7848 6605
Email: roger.newson@kcl.ac.uk

Opinions expressed are those of the author, not the institution.

*
*   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/