[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

From |
David Airey <david.airey@Vanderbilt.Edu> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
st: Stata not recommended for exact 2x2 test |

Date |
Sun, 9 Nov 2008 10:40:53 -0600 |

Int J Epidemiol. 2008 Aug 18. [Epub ahead of print]

Ludbrook J. Department of Surgery, The University of Melbourne, Parkville, Victoria, Australia.

The most useful are StatXact and Testimate. and from the discussion: Which statistics program? It is safe to use any of the statistics programs listed in Table 4 for the asymptotic form of Pearson’s chi2-test and for Fisher’s exact test. But for an exact test on OR =1, it is safe only to use the Testimate single-conditioned option. For the exact test on RR =1, and for the exact test on the difference in proportions (p2 – p1 =0), it is safe to use StatXact and Testimate, but not Stata. and later in the appendix: Commercial software packages for exact tests Stata 10.0 (Stata Corporation, College Station, TX, USA) provides ‘exact’ 95% CIs for the test statistics OR, RR and the difference in proportions. However, the CI for OR is a good deal wider than that given by StatXact, which suggests that this is the result of (inappropriate) double conditioning. The only P-values that Stata gives are those that result from Fisher’s exact test (that is, with double conditioning). -- David C. Airey, Ph.D. Pharmacology Research Assistant Professor Center for Human Genetics Research Member Vanderbilt University School of Medicine * * 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/

- Prev by Date:
**Re: st: Interval variables as independent variables** - Next by Date:
**st: stcompet updated** - Previous by thread:
**st: sum across observations until a certain cutpoint** - Next by thread:
**Re: st: Stata not recommended for exact 2x2 test** - Index(es):

© Copyright 1996–2016 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |