Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: RE: P-value

From   Roger Newson <>
Subject   RE: st: RE: P-value
Date   Fri, 11 Nov 2005 18:19:36 +0000

If Raphael wants to see P-values to a greater precision than 3 decimal places, then Raphael might like to use the -parmest- package, downloadable from SSC. For instance, if you install -parmest-, and type

regress y x1 x2 x3
parmest, list(parm estimate min* max* p) format(p %-8.2g)

then the P-values will be displayed to 2 significant figures, rather than to 3 decimal places. Raphael will therefore know whether a P-value of 0.000 means 4*10^-4, 4*10^-5 or 4*10^-9. This distinction might be important if you have calculated a lot of P-values, and want to know whether the smallest ones are credible, given the number of P-values that you have calculated. After all, 5 percent of P-values will be significant at the 5 percent level, even if all null hypotheses are true. On the other hand, if all null hypotheses are true, then you would have to calculate a very large number of P-values to find a P-value of 4*10^-9 just by chance.

A possible alternative to -parmest- is -estout-, which is also downloadable from SSC and does a lot of similar things.

I hope this helps.


At 17:27 11/11/2005, you wrote:

>-----Original Message-----
>From: []On Behalf Of Raphael Fraser
>Sent: vrijdag 11 november 2005 18:17
>Subject: Re: st: RE: P-value
>That's what I think too but I keep seeing people interpreting 0.000 as
><0.0001. Am I missing something here?

very small is very small, your null hypothesis is rejected by any conventional confidence level, which is all that P-values are supposed to say.

>>On 11/11/05, Maarten Buis <> wrote:
>> as very small (less than 0.0005)

* For searches and help try:

Roger Newson
Lecturer in Medical Statistics
Department of Public Health Sciences
Division of Asthma, Allergy and Lung Biology
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

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

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index