摘要:

In various contexts of statistical problems one deals with the measurement of a “risk”—a risk of dying, a risk of becoming ill, and so forth. In applied vital statistics various indices are used to measure such risks, and typically these arenotprobabilities. In formal mathematical statistical analyses, typically theyaremeasured as probabilities. A special problem arises when more than one risk must be measured at the same time. Even if they are independent in a technical statistical sense, the presence of one risk will complicate the measurement of another. This situation has been referred to as “competing risks.” In this paper we apply an analysis of competing risks to some questions that arose in the study of the relation of smoking to lung cancer.

ISSN:0162-1459    

DOI:10.1080/01621459.1960.10482072

出版商:Taylor & Francis Group    

年代:1960

数据来源: Taylor

摘要:

A nonparametric procedure is presented to test the null hypothesis that two independent samples come from the same population against the alternative hypothesis that the samples come from populations differing in variability or “spread.” Extensive tables of critical values are included forn1≤n2≤20. Large sample procedures are presented which include a correction for tied observations. The test is entirely distribution-free under the usual randomization procedures against the null hypothesis that the two distributions are identical. The absence of any normality assumption is a particularly important feature of the test, because its parametric alternative, theFtest for variance differences, is quite sensitive to departures from normality. The test has the additional advantage of being directly applicable to non-numerical ordinal data.

ISSN:0162-1459    

DOI:10.1080/01621459.1960.10482073

出版商:Taylor & Francis Group    

年代:1960

数据来源: Taylor

摘要:

Modern commercial cloud seeding dates from the discovery in 1946 that silver iodide crystals are good nuclei for the formation of precipitation from cold clouds. Whether silver iodide seeding does actually increase the precipitation over what would occur naturally has been a controversial question.

ISSN:0162-1459    

DOI:10.1080/01621459.1960.10482074

出版商:Taylor & Francis Group    

年代:1960

数据来源: Taylor

摘要:

Social statistics are defined as the statistics required for social programmes, and a social programme as the scheme of public and private activities which have a direct bearing on a particular aspect of living conditions. It is assumed that there is a separate programme for each component of the levels of living, e.g. health, education, housing.

ISSN:0162-1459    

DOI:10.1080/01621459.1960.10482075

出版商:Taylor & Francis Group    

年代:1960

数据来源: Taylor

摘要:

There is wide variation in the need for and the ability to accumulate liquid assets. The model used for investigating liquid asset holdings attempts to isolate the precautionary motive from other motives for holding liquid reserves. The propensities for holding liquid assets for precautionary purposes are analyzed by classifying consumers according to their long-run income position. Within each class regression equations are used to test hypotheses concerning the relationship between liquid balances and age, size of consumer unit, level of current income, and home ownership.

ISSN:0162-1459    

DOI:10.1080/01621459.1960.10482076

出版商:Taylor & Francis Group    

年代:1960

数据来源: Taylor

摘要:

Data from certain life test experiments exhibit such an unusually high concentration of failures near time 0 that the assumption of an over-all exponential density is unwarranted. One hypothesis for this phenomenon is that due to faulty construction or defective parts, certain test items fail prematurely; these items are termed “early failures.” To handle this type of data, an early failure model is postulated in which one failure rate is assumed to be in effect for an initial time interval [0,T0) and another, lower failure rate is operative thereafter. Estimators for the two failure rates are given in the case whereT0is known and in the case whereT0is not known exactly but can be assumed to be within a specified interval. Methods for obtaining approximate large sample confidence regions are outlined and procedures for handling small samples are described.

ISSN:0162-1459    

DOI:10.1080/01621459.1960.10482077

出版商:Taylor & Francis Group    

年代:1960

数据来源: Taylor

摘要:

Brief mention is made of several models in each of two categories: (I) Each possible ranking of items is assumed to have a “utility” which depends on the expected scores of the items in paired comparisons. In particular, the “worth” of an item may bedefinedin terms of its expected scores in comparisons with others. (II) Each item isassumed to have an intrinsic worth;these intrinsic worths determine the expected scores.

ISSN:0162-1459    

DOI:10.1080/01621459.1960.10482078

出版商:Taylor & Francis Group    

年代:1960

数据来源: Taylor

摘要:

Values ofp(times 10,000) satisfying the equation

ISSN:0162-1459    

DOI:10.1080/01621459.1960.10482079

出版商:Taylor & Francis Group    

年代:1960

数据来源: Taylor

摘要:

A method is presented for analyzing survival data with incomplete follow-up when there is reason to suppose that the data have a translated log-normal distribution. The life table method is used to make allowances for those patients lost to observation and for tabulating the resulting survival frequencies. Modified maximum likelihood equations and short-cut procedures are presented for calculating the parameters of the general three parameter log-normal distribution from grouped data as represented by the life table.

ISSN:0162-1459    

DOI:10.1080/01621459.1960.10482080

出版商:Taylor & Francis Group    

年代:1960

数据来源: Taylor

摘要:

This article is an attempt to throw more light on the problem of choosing plotting positions for the frequencies related to ordered sample values. The problem is particularly pertinent where special-scale graph paper is used such that a straight line indicates conformance of data to type of distribution assumed to apply. Such graph paper is often briefly referred to as “probability paper.” The types of distributions here considered are the normal and the extreme-value distribution of Type I, with the accent on the extreme-value case.

ISSN:0162-1459    

DOI:10.1080/01621459.1960.10482081

出版商:Taylor & Francis Group    

年代:1960

数据来源: Taylor