ISSN:0162-1459    

DOI:10.1080/01621459.1964.10480698

出版商:Taylor & Francis Group    

年代:1964

数据来源: Taylor

摘要:

In the collection of consumer expenditures data through household interviews, a major source of error arises from respondents' failure to recall expenditures, or to recall exactly when the expenditures were made. In an experimental study conducted by the U. S. Bureau of the Census, the measurement of these two types of response errors in the reporting of alterations and repairs expenditures in household interviews was a major objective.

ISSN:0162-1459    

DOI:10.1080/01621459.1964.10480699

出版商:Taylor & Francis Group    

年代:1964

数据来源: Taylor

摘要:

The termspredictiveandpostdictiveare introduced for two basic and implicitly well-known modes of interpreting probabilities, and the roles of these two modes in statistical inference are discussed with emphasis on the confidence argument of Neyman and the fiducial argument of Fisher. The conflicts between the confidence approach and the fiducial approach are analyzed in a way which generally supports Fisher's position. Finally, difficulties internal to the fiducial approach are discussed. It is concluded that the general form of the fiducial argument is appealing but that many of the restrictions imposed by Fisher are awkward or ambiguous and ought, perhaps, to be replaced.

ISSN:0162-1459    

DOI:10.1080/01621459.1964.10480700

出版商:Taylor & Francis Group    

年代:1964

数据来源: Taylor

摘要:

Let θ denote the unknown proportion of defectives in an indefinitely large population. A sampling plan is constructed which approximately minimizes the maximum expected sample size among all plans satisfying the conditions

ISSN:0162-1459    

DOI:10.1080/01621459.1964.10480701

出版商:Taylor & Francis Group    

年代:1964

数据来源: Taylor

摘要:

Two theorems are presented which are useful in making inferences about the univariate normal distribution. Some applications of the theorems are given. It is thought that the theorems will prove useful in other contexts.

ISSN:0162-1459    

DOI:10.1080/01621459.1964.10480702

出版商:Taylor & Francis Group    

年代:1964

数据来源: Taylor

摘要:

A description is given of the interacting effects which exist between quality control stations associated with the various stages of manufacture of a product, and how these interactions influence the optimal choice of sampling plan for a given station. The applicability of the mathematical theory of dynamic programming to the design of an optimal sequence of such interrelated plans is demonstrated, along with a general description of this theory. In order to illustrate the dynamic programming solution procedure, a digital computer program is discussed which solves a simplified version of the above design problem. Extension of this program to the solution of more general multistage decision problems is also considered.

ISSN:0162-1459    

DOI:10.1080/01621459.1964.10480703

出版商:Taylor & Francis Group    

年代:1964

数据来源: Taylor

摘要:

In this paper the consistency of the two-sample tests of Sukhatme [6], Ansari and Bradley [1], Siegel and Tukey [5], and Mood [3] is investigated. Each of these tests is a distribution-free analogue of theF-test for testing the equality of the variances of two normal distributions. If the two samples are taken from continuous distributions with the same median (say 0) and distribution functionsF(x) andF(x/a) respectively, the hypothesis tested states thata= 1 and the two-sided tests are consistent fora≠ 1. If, however, the samples are from continuous distributions with the same median 0 and arbitrary distribution functionsFandG, little is known about their asymptotic properties. In this paper the conditions for consistency for this case are given; further it is investigated if the corresponding estimates of differences in dispersion satisfy the usual properties of measures of dispersion, namely 1) thatXand −Xhave the same dispersion and 2) that, ifXhas a larger (or equal or smaller) dispersion thanY, − Xhas a larger (or equal or smaller) dispersion thanY.

ISSN:0162-1459    

DOI:10.1080/01621459.1964.10480704

出版商:Taylor & Francis Group    

年代:1964

数据来源: Taylor

摘要:

Coding in the 1960 Censuses had two characteristics which complicated the control of quality. They were: (1) the codes to be verified were in the form of positioned marks on the Census schedule, and (2) coding was a difficult operation frequently requiring the use of judgment on the part of the coder. This paper describes the techniques used in controlling the quality of the operation.

ISSN:0162-1459    

DOI:10.1080/01621459.1964.10480705

出版商:Taylor & Francis Group    

年代:1964

数据来源: Taylor

摘要:

A two-parameter generalization of the Poisson distributions, called the “hyper-Poisson” family, is derived and discussed. It is in turn a special case of the three-parameter confluent hypergeometric series distributions. When one parameter is fixed, the hyper-Poisson distributions are also power series distributions. The distributions are classified as “sub-Poisson” or “super-Poisson” according as the variance is less than or greater than the mean. The latter category includes the truncated Poisson distributions. Certain properties of the hyper-Poisson distributions are considered, including moment, modified moment, and maximum likelihood estimators of the parameters. Estimation using moments is illustrated on Student's haemacytometer data.

ISSN:0162-1459    

DOI:10.1080/01621459.1964.10480706

出版商:Taylor & Francis Group    

年代:1964

数据来源: Taylor

摘要:

The power function of the one sided, one-sample sign test is studied for populations which deviate from exact normality, either by skewness, kurtosis, or both. The terms of the Edgeworth asymptotic expansion of order more thanN−3/2are used to represent the population density. Three sets of hypotheses and alternatives, concerning the location of (1) the median, (2) the mean, and (3) the median as approximated by the mean and coefficient of skewness, are considered in an attempt to make valid comparisons between the power of the sign test and Student'sttest under the same conditions. Numerical results are given for samples of size 10, significance level .05, and for several combinations of the coefficients of skewness and kurtosis. The computations reveal that in general the sign test performs well in comparison with thettest, and in the case of leptokurtic distributions, there is almost no difference in power.

ISSN:0162-1459    

DOI:10.1080/01621459.1964.10480707

出版商:Taylor & Francis Group    

年代:1964

数据来源: Taylor