摘要:

A stepwise procedure is often used to estimate one coefficient in a multiple linear regression. The relation between this stepwise estimate and the estimate obtained by direct multiple regression is spelled out.

ISSN:0162-1459    

DOI:10.1080/01621459.1961.10482095

出版商:Taylor & Francis Group    

年代:1961

数据来源: Taylor

摘要:

This paper proposes a simplification of Greenwood's estimate of the standard error of the survival rate. The simplifying assumptions are (a) that cases enter the follow-up study at a uniform rate, (b) there are no losses to follow-up, and (c) the mortality risk for successive follow-up intervals is constant. Numerical examples drawn from cancer registry data indicate that the proposed estimate yields acceptable results under certain realistic circumstances despite some marked deviations from the simplifying assumptions. A table is given for rapid estimation of the standard error of the survival rate.

ISSN:0162-1459    

DOI:10.1080/01621459.1961.10482096

出版商:Taylor & Francis Group    

年代:1961

数据来源: Taylor

摘要:

A common disturbance in follow-up surveys is the loss due to movement, a loss which makes it impossible to estimate in a straightforward fashion the frequency of events of interest irrespective of movement.

ISSN:0162-1459    

DOI:10.1080/01621459.1961.10482097

出版商:Taylor & Francis Group    

年代:1961

数据来源: Taylor

摘要:

The main concern of this investigation involves the concept of “the probability of reversal.” That is, the determination of the conditional probability ofnot rejectingthe null-hypothesis,H0, if all the data (nfixed) are available, given thatH0isrejectedby the test of significance based on incomplete data (Type 1). Or, on the other hand, the conditional probability ofrejecting H0if all the data are available, given thatH0isnot rejectedby the test of significance based on incomplete data (Type 2). Therefore, “the probability of reversal” indicates the degree of stability in the statistical inference procedure when the originally designed experiment is disrupted by loss of observations.

ISSN:0162-1459    

DOI:10.1080/01621459.1961.10482098

出版商:Taylor & Francis Group    

年代:1961

数据来源: Taylor

摘要:

Formulas describing optimum sample allocation in stratified and two-stage sample designs are given ordinarily in terms of “best” estimates of a population mean or total. The best allocation for estimating the variance of the mean may depart from that which is optimum for the mean. For stratified sampling, such a departure depends upon variation among strata in functions of coefficients of kurtosis for the strata (fourth moment ÷ squared second moment). For two-stage designs optimum sample allocation for estimating the variance of the mean depends upon coefficients of kurtosis as well as costs and variance components. Examination of the expressions for optimum variance estimation and some calculations for specific cases indicate that losses in efficiency of variance estimates due to using optimum allocation for a mean are not extreme in a variety of particular cases.

ISSN:0162-1459    

DOI:10.1080/01621459.1961.10482099

出版商:Taylor & Francis Group    

年代:1961

数据来源: Taylor

摘要:

Transformations for the stabilization of variance of the binomial and negative binomial distributions are considered. In each case a new transformation is suggested. These transformations are two-term inverse sine and inverse-hyperbolic sine transformations involving six adjustable constants, thus generalizing previous work of Anscombe [1] and of Freeman and Tukey [3]. In the sequel some properties of these transformations are derived. The nature of the dependence of the variances of these transformations on some parameter is worked out and depicted graphically.

ISSN:0162-1459    

DOI:10.1080/01621459.1961.10482100

出版商:Taylor & Francis Group    

年代:1961

数据来源: Taylor

摘要:

The problem considered is estimation by sampling methods of the number of names common to two or more lists. The situation covered is that in which the lists are found to have been merged and must be sampled as one. A comparison is made with the case in which the lists can be sampled separately. Tables are included giving the sample sizes necessary to achieve a specified level of precision.

ISSN:0162-1459    

DOI:10.1080/01621459.1961.10482101

出版商:Taylor & Francis Group    

年代:1961

数据来源: Taylor

摘要:

This probability table is based on the number of runs of signs of first differences in ordered series, or expressed differently, the number of rises and falls in ordered series, that would occur under this nonparametric null hypothesis: the relative magnitude of an observation is independent of its position in the series. The runs test used in conjunction with the probability table in this article is probably less sensitive than rank order correlation techniques for detecting monotonic correlations, but much better for rejecting the above null hypothesis in dealing with series involving cycles or some other non-monotonic systematic rise and fall in the magnitude of the observations, such asU-shaped or invertedU-shaped series of observations.

ISSN:0162-1459    

DOI:10.1080/01621459.1961.10482102

出版商:Taylor & Francis Group    

年代:1961

数据来源: Taylor

摘要:

Testing Statistical Hypotheses.E. L. Lehmann. New York: John Wiley and Sons, Inc., 1959. Pp. xiii, 369. $11.00. Reviewed by John W. Pratt

ISSN:0162-1459    

DOI:10.1080/01621459.1961.10482103

出版商:Taylor & Francis Group    

年代:1961

数据来源: Taylor

摘要:

This article has no abstract

ISSN:0162-1459    

DOI:10.1080/01621459.1961.10482084

出版商:Taylor & Francis Group    

年代:1961

数据来源: Taylor