摘要:

Originally a talk delivered at a conference on Bayesian statistics, this article attempts to answer the following question: why is most scientific data analysis carried out in a non-Bayesian framework? The argument consists mainly of some practical examples of data analysis, in which the Bayesian approach is difficult but Fisherian/frequentist solutions are relatively easy. There is a brief discussion of objectivity in statistical analyses and of the difficulties of achieving objectivity within a Bayesian framework. The article ends with a list of practical advantages of Fisherian/frequentist methods, which so far seem to have outweighed the philosophical superiority of Bayesianism.

ISSN:0003-1305    

DOI:10.1080/00031305.1986.10475342

出版商:Taylor & Francis Group    

年代:1986

数据来源: Taylor

ISSN:0003-1305    

DOI:10.1080/00031305.1986.10475343

出版商:Taylor & Francis Group    

年代:1986

数据来源: Taylor

ISSN:0003-1305    

DOI:10.1080/00031305.1986.10475344

出版商:Taylor & Francis Group    

年代:1986

数据来源: Taylor

ISSN:0003-1305    

DOI:10.1080/00031305.1986.10475345

出版商:Taylor & Francis Group    

年代:1986

数据来源: Taylor

ISSN:0003-1305    

DOI:10.1080/00031305.1986.10475346

出版商:Taylor & Francis Group    

年代:1986

数据来源: Taylor

ISSN:0003-1305    

DOI:10.1080/00031305.1986.10475347

出版商:Taylor & Francis Group    

年代:1986

数据来源: Taylor

ISSN:0003-1305    

DOI:10.1080/00031305.1986.10475348

出版商:Taylor & Francis Group    

年代:1986

数据来源: Taylor

摘要:

The decision on whether or not to include a unit root in an autoregressive operator has profound implications. Formal tests for the presence of unit roots give analysts objective guidance in this decision. This article is a practical guide to the use of these tests.

ISSN:0003-1305    

DOI:10.1080/00031305.1986.10475349

出版商:Taylor & Francis Group    

年代:1986

数据来源: Taylor

摘要:

Approximate cut-off rules for stratifying a population into a take-all and take-some universe have been given by Dalenius (1952) and Glasser (1962). Their cut-off value (the value that delineates the take-all and take-some boundary) was derived on the assumption that all units from the take-all universe were included in the sample and that a simple random sample of fixed size was to be drawn without replacement from the population of take-some units. In the present context, exact and approximate cut-off rules are worked out in terms of required precision expected from the overall sample.

ISSN:0003-1305    

DOI:10.1080/00031305.1986.10475350

出版商:Taylor & Francis Group    

年代:1986

数据来源: Taylor

摘要:

The crude death rate of countryAmay be less than that of countryBeven if every age-specific death rate of countryAis greater than each corresponding one of countryB.This is an example of what statisticians (unjustly) call Simpson's paradox. What holds for death rates holds equally for all other demographic rates. Simpson's paradox can recur, reversing an inequality of rates, whenever an additional variable is introduced into a stratification. Repeated stratification of a finite population (e.g., by age, sex, education, income, region) may eventually produce comparison groups that are too small for a given difference in mortality to be detected. The trade-off between the increased homogeneity of highly stratified comparison groups and the decreased ability to detect small differences in probabilities of death is described here quantitatively by an uncertainty principle, which takes the form of an inequality. The possibility of encountering Simpson's paradox suggests that since sex is only one of many possible stratifying variables that appear to affect mortality, the use of mortality tables distinguished by sex and by no other variables is, in the absence of information about the importance of other variables, demographically arbitrary.

ISSN:0003-1305    

DOI:10.1080/00031305.1986.10475351

出版商:Taylor & Francis Group    

年代:1986

数据来源: Taylor