摘要:

AbstractA new estimation procedure for mixed regression models is introduced. It is a development of Henderson's best linear unbiased prediction procedure which uses the joint distribution of the observed dependent random variables and the unknown realisations of the random components of the model. It is proposed to replace the likelihood of the observations given the random components by the asymptotic likelihood of the maximum likelihood estimators and the prior distribution of the random components by a restricted prior distribution which is consistent with the usual restrictions placed on the random components when they are considered conditionally fixed.

ISSN:0323-3847    

DOI:10.1002/bimj.4710330202

出版商:WILEY‐VCH Verlag    

年代:1991

数据来源: WILEY

ISSN:0323-3847    

DOI:10.1002/bimj.4710330203

出版商:WILEY‐VCH Verlag    

年代:1991

数据来源: WILEY

摘要:

AbstractIn this paper we lay the foundation of life table construction by unifying the existing life table methods. We also present a new method of constructing current (period) abridged life tables. The proposed method produces more life talbe functions than other existing methods do. The development includes (1) a careful formulation and computation of age‐specific death rates, (2) derivation of a new set of formulas for computing the survivorship function from the observed age‐specific death rates and populations, (3) estimation of the main life table functions by spline interpolation, integration and differentiation, and (4) use of a quadratic and a Gompertz function to close the life table. The new abridged life table method features (i) improved accuracy, (ii) tolerance to observed age distributions with bulges and dents and (iii), together with the infant life table method of HSIEH(1985), coverage of the whole life span. The method is illustrated with construction of abridged life tables using Canadian d

ISSN:0323-3847    

DOI:10.1002/bimj.4710330204

出版商:WILEY‐VCH Verlag    

年代:1991

数据来源: WILEY

摘要:

AbstractSurvival times of patients can be compared using rank tests in various experimental setups, including the two‐sample case and the case of paired data. Attention is focussed on two frequently occurring complications in medical applications: censoring and tail alternatives. A review is given of the author's recent work on a new and simple class of censored rank tests. Various models for tail alternatives are discussed and the relation to censoring is demonstrate

ISSN:0323-3847    

DOI:10.1002/bimj.4710330205

出版商:WILEY‐VCH Verlag    

年代:1991

数据来源: WILEY

摘要:

AbstractConfidence envelopes for Ripley'sLfunction forNrandom points in the unit square were obtained by simulation. The 90, 95 and 99th percentiles of 1000 simulations are given in tabular form forNin the range 10 to 350. Results for a number of choices of the maximum interpoint distance are reported.

ISSN:0323-3847    

DOI:10.1002/bimj.4710330206

出版商:WILEY‐VCH Verlag    

年代:1991

数据来源: WILEY

ISSN:0323-3847    

DOI:10.1002/bimj.4710330207

出版商:WILEY‐VCH Verlag    

年代:1991

数据来源: WILEY

摘要:

AbstractA simple testing procedure “control versusktreatments” for one‐sided ordered alternatives for univariate, continuous variables is given. With a simulation study both the first kind riskaand the power behaviour under several distributions, expected value profiles, sample sizes andalevels are

ISSN:0323-3847    

DOI:10.1002/bimj.4710330208

出版商:WILEY‐VCH Verlag    

年代:1991

数据来源: WILEY

ISSN:0323-3847    

DOI:10.1002/bimj.4710330209

出版商:WILEY‐VCH Verlag    

年代:1991

数据来源: WILEY

摘要:

AbstractKernel estimators are used to obtain new estimators for the tumour onset distribution and its density. Asymptotical properties of the estimators are studied and confidence tests are constructed.

ISSN:0323-3847    

DOI:10.1002/bimj.4710330210

出版商:WILEY‐VCH Verlag    

年代:1991

数据来源: WILEY

摘要:

AbstractEfficiency of linked cross‐sectional scheme has been compared with pure longitudinal schemes for generalised estimation of norm relaxing the condition of constancy of variances. The relevant theory of estimation has been developed and the expression for the optimum estimators alongwith their variances have been derived. The percentage gain in efficiency has been calculated. It has been observed that for higher values ofQ, percentage gain in efficiency of proposed scheme over cross‐sectional is very h

ISSN:0323-3847    

DOI:10.1002/bimj.4710330211

出版商:WILEY‐VCH Verlag    

年代:1991

数据来源: WILEY