[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

From |
"Carlo Lazzaro" <carlo.lazzaro@tin.it> |

To |
<statalist@hsphsun2.harvard.edu> |

Subject |
st: R: Anova with not variance homogeneity |

Date |
Wed, 3 Oct 2007 18:06:09 +0200 |

Dear Rachele, perhaps a possible suggestion would be to perform a non-parametric bootstrap ANOVA (frequently used in the economic evaluation of healthcare programmes). Please take a look at what follows: -----------------start example--------------- use "C:\Programmi\Stata9\ado\base\a\auto.dta" oneway price foreign, bonferroni tabulate return list scalar F=r(F) summarize price if foreign ==0, mean replace price = price -r(mean) + 6165.257 if foreign ==0 summarize price if foreign ==1, mean replace price = price -r(mean) + 6165.257 if foreign ==1 summarize price by foreign, sort : summarize price bootstrap r(F), reps(1000) nodots verbose seed(999) : oneway price foreign, bonferroni tabulate estat bootstrap, all verbose count if r(F)>=F -----------------end example--------------- By dividing the result of count if r(F)>=F for the number of bootstrap replications (1000 in this example), you will get a pbootstrap that you can compare with the base case ANOVA p-value. The pbootstrap may confirm or deny the base case ANOVA p-value. Fir further details (however about the application of non parametric bootstrap to t-test and not to ANOVA) I would refer you to: Efron B, Tibshirani RJ. An Introduction to the Bootstrap. Chapman and Hall: New York, 1993. Desgagné A, Castilloux AM, Angers JF, LeLorier J. The use of bootstrap statisticalmethod for the pharmacoeconomic cost analysis of skewed data. PahramacoEconomics 1998; 13: 487-497. I do hope this helps. Kind Regsrds, Carlo -----Messaggio originale----- Da: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] Per conto di Rachele Capocchi Inviato: mercoledì 3 ottobre 2007 15.49 A: statalist@hsphsun2.harvard.edu Oggetto: st: Anova with not variance homogeneity Dear Statalisters, Can anybody help me to do Anova test if there is difference on the 3 groups' means with Stata, considering that the variances of 3 groups are not homogeneity ? Thanks a lot, Rachele * * 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/ * * 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/

**References**:**st: Anova with not variance homogeneity***From:*Rachele Capocchi <rachele.capocchi@arsanita.toscana.it>

- Prev by Date:
**Re: st: local macro tempvar troubles** - Next by Date:
**st: Lincom after devcon, and another minor devcon question** - Previous by thread:
**st: RE: Anova with not variance homogeneity** - Next by thread:
**st: Test for the difference in the coefficients of two models that usedifferent samples - Stochastic Frontier Analysis** - Index(es):

© Copyright 1996–2015 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |