Exact small-sample methods for discrete data use probability distributions that do not depend on unknown parameters. However, they are conservative inferentially: The actual error probabilities for tests and confidence intervals are bounded above by the nominal level. This article discusses ways of reducing the conservatism. Fuzzy inference is a recent innovation that enables one to achieve the error probability exactly. We present a simple way of conducting fuzzy inference for discrete one-parameter exponential family distributions. In practice, most scientists would find this approach unsuitable yet might be disappointed by the conservatism of ordinary exact methods. Thus, to use exact small-sample distributions, we recommend inferences based on the mid-P value. This approach can be motivated by fuzzy inference, it is less conservative than standard exact methods, yet usually it does well in terms of achieving desired error probabilities. We illustrate this and other small-sample methods for the case of inferences about the binomial parameter.

Reducing Conservatism of Exact Small-Sample Methods of Inference for Discrete Data / A.Agresti;A.Gottard. - STAMPA. - (2006), pp. 245-260. (Intervento presentato al convegno COMPSTAT 2006 tenutosi a Roma nel 28/8 - 1/9 2006) [10.1007/978-3-7908-1709-6_19].

Reducing Conservatism of Exact Small-Sample Methods of Inference for Discrete Data

GOTTARD, ANNA
2006

Abstract

Exact small-sample methods for discrete data use probability distributions that do not depend on unknown parameters. However, they are conservative inferentially: The actual error probabilities for tests and confidence intervals are bounded above by the nominal level. This article discusses ways of reducing the conservatism. Fuzzy inference is a recent innovation that enables one to achieve the error probability exactly. We present a simple way of conducting fuzzy inference for discrete one-parameter exponential family distributions. In practice, most scientists would find this approach unsuitable yet might be disappointed by the conservatism of ordinary exact methods. Thus, to use exact small-sample distributions, we recommend inferences based on the mid-P value. This approach can be motivated by fuzzy inference, it is less conservative than standard exact methods, yet usually it does well in terms of achieving desired error probabilities. We illustrate this and other small-sample methods for the case of inferences about the binomial parameter.
2006
COMPSTAT 2006: Proceedings in Computational Statistics
COMPSTAT 2006
Roma
28/8 - 1/9 2006
A.Agresti;A.Gottard
File in questo prodotto:
File Dimensione Formato  
compstat2006.pdf

Accesso chiuso

Tipologia: Pdf editoriale (Version of record)
Licenza: Tutti i diritti riservati
Dimensione 111.22 kB
Formato Adobe PDF
111.22 kB Adobe PDF   Richiedi una copia

I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificatore per citare o creare un link a questa risorsa: https://hdl.handle.net/2158/342931
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? 3
social impact