Showing papers in "Journal of Applied Statistics in 1987"
TL;DR: In this paper, maximum entropy sampling is used to sample the maximum entropy of a set of maximum entropy samples from a single maximum entropy sample set, and the sample set is used for maximum entropy estimation.
Abstract: (1987). Maximum entropy sampling. Journal of Applied Statistics: Vol. 14, No. 2, pp. 165-170.
553 citations
TL;DR: In this paper, a maximum likelihood (ml) procedure was proposed to adjust for bias due to non-random missingness, where nonrandomness is specified by a logistic distribution.
Abstract: SUMMARY Y is a linear regression on a variable X; X is fixed and all its sample values are observed. Y, on the other hand, has some sample values missing. This work outlines a maximum likelihood (ml) procedure that tries to adjust for bias due to non-random missingness; here non-randomness is specified by a logistic distribution. The ml procedure is implemented via two iterative technologies, namely the EM algorithm (of Dempster, Laird & Rubin, 1977) and the Newton-Raphson method. Data from a dialysis study are used to illustrate our estimation procedure, and results show that the ml procedure is quite effective in adjusting for bias.
290 citations
TL;DR: A Bayesian approach is presented for influential observations and predictive discordancy diagnostics and tests of significance based on predictive distributions are devised to test the discordancy of an observation.
Abstract: A Bayesian approach is presented for influential observations and predictive discordancy diagnostics. Tests of significance based on predictive distributions are devised to test the discordancy of an observation. Examples illustrating the methods are provided.
32 citations
TL;DR: In this paper, five different models of the Lorenz curve are estimated using the Australian grouped income distribution data and evaluated in regard to their adequacy in modelling the curve based on empirical results.
Abstract: Five different models of the Lorenz curve are estimated using the Australian grouped income distribution data. Performances of these models are evaluated in regard to their adequacy in modelling the Lorenz curve based on empirical results.
24 citations
TL;DR: In this article, the variance of mean fraction nonconformance under the exact lot quality distribution with first order dependence is examined and the extent of the validity of binomial assumption with respect to the exact distribution in constructing the p-charts for dependent production processes is investigated and it is shown that neglect of dependence for small lot sizes may cause bias in the width of the area between the control limits.
Abstract: In this study, the variance of mean fraction nonconformance under the exact lot quality distribution with first order dependence is examined. The extent of the validity of binomial assumption with respect to the exact distribution in constructing the p-charts for dependent production processes is investigated and it is shown that neglect of dependence for small lot sizes may cause bias in the width of the area between the control limits. An approximate lot quality distribution with great versatility is proposed and is used for correcting the bias due to dependence in p-charts.
24 citations
TL;DR: In this paper, the authors adapted control charts to meet practical needs: a chemical processing application, and applied them in a chemical plant. But they did not consider the effects of the control chart on the chemical process.
Abstract: (1987). Adapting control charts to meet practical needs: a chemical processing application. Journal of Applied Statistics: Vol. 14, No. 1, pp. 35-52.
21 citations
TL;DR: In this paper, several automated prediction techniques are discussed and compared, including least squares, weighted least square, James-Stein, and Harville, and a more data-intensive approach is discussed.
Abstract: The rules of American football favor the strategic placement of the 11 players per team making the identification of statistical tendencies a particularly useful capability. Gambling on American football games is explained. Several automated prediction techniques are discussed and compared, including least squares, weighted least squares, James-Stein, and Harville. A more data-intensive approach is discussed. That approach has coaching implications as well as predictive ability.
21 citations
TL;DR: In this article, the influence of individual values can be influential in randomisation inference, either when they are included in the sample or when they were not in a sample, and in the former case diagnostics are useful, in the latter case only prior knowledge can save the inference.
Abstract: If surveys are analysed using the randomisation distribution then model-based influence diagnostics have no role. Individual values can be influential in randomisation inference either when they are included in the sample or when they are not in the sample. In the former case diagnostics are useful, in the latter only prior knowledge can save the inference. Post-stratification unifies a variety of ad hoc procedures. Some ideas deducible from the influence function provide a justification for the Hajek ratio estimator which is an alternative to the Horvitz-Thompson estimator.
16 citations
TL;DR: In this paper, a theoretical and experimental study of the Principal Component Analysis of a stochastic process (PCAP) is presented, which aims to clarify the theoretical framework and provide a simulation experimental basis for evaluating potential errors when using PCAP as a mapping technique.
Abstract: This work proposes a theoretical and experimental study of the Principal Component Analysis of a stochastic Process (PCAP). It aims to clarify the theoretical framework and provides a simulation experimental basis for evaluating potential errors when using the PCAP as a mapping technique.
11 citations
TL;DR: In this paper, an empirical study of applying seven commonly used methods for identifying outliers and influential observations in linear regression models to six well-known data sets in the literature is presented.
Abstract: This paper reports the results of an empirical study of applying seven commonly used methods for identifying outliers and influential observations in linear regression models to six well-known data sets in the literature. The results show that the methods do not always agree. Observations which are detected as ‘significantly influential’ may only moderately affect the parameter estimates when they are removed from the regression. Combinations of procedures should be judiciously used in empirical studies.
9 citations
TL;DR: The authors describe their experiences with a teaching approach in which students generate and analyse their own data by conducting a factorial experiment, and note their expectations, successes, and disappointments, as well as their recommendations.
Abstract: A major challenge in teaching statistics is to convey how statistics is used in solving real problems. Introductory courses often fail to provide students with a sense of how to apply statistical ideas. This paper will describe our experiences with a teaching approach in which students generate and analyse their own data by conducting a factorial experiment. We note our expectations, successes, and disappointments, as well as our recommendations.
TL;DR: In this paper, the authors analyzed factorial experiments and least squares polynomial fitting by the method of orthogonal polynomials for any spacing of the levels of the independent variables.
Abstract: (1987). Analysis of factorial experiments and least squares polynomial fitting by the method of orthogonal polynomials for any spacing of the levels of the independent variables. Journal of Applied Statistics: Vol. 14, No. 1, pp. 83-89.
TL;DR: In this article, the practical merits of Neyman-Pearson confidence intervals versus hypothesis tests (especially p-values) are discussed, especially for the one-parameter case.
Abstract: The practical merits of Neyman-Pearson confidence intervals versus hypothesis tests (especially p-values) are discussed. In the one-parameter case, confidence intervals convey much more comprehensible information, especially for the ultimate users of statistics. For more complicated inferences, p-values may be simpler and more routine.
TL;DR: The distribution of CTV in citrus groves in Israel is considered using the data gathered from the CTV survey during the period 1979–1985, and a theoretical curve is fitted to the distribution of the number of infected trees in infected plots.
Abstract: In this study, the distribution of CTV in citrus groves in Israel is considered using the data gathered from the CTV survey during the period 1979–1985. A theoretical curve is fitted to the distribution of the number of infected trees in infected plots which contain at least a single infected tree. Some recommendations for improving the sampling method for locating infected plots with a low incidence of infection are presented.
TL;DR: In this article, three simple algorithms are presented that facilitate calculation of the analysis of variance for data from any block design and are intended for use with experiments that have suffered damage and therefore no longer conform to any recognised design.
Abstract: Three very simple algorithms are presented that facilitate calculation of the analysis of variance for data from any block design. The first provides the sums of squares for treatments and error. The second calculates the treatment means with adjustment for any non-orthogonality and also gives the degrees of freedom. The third gives the constant needed to derive the variance of an unconfounded contrast from the error mean square. All depend upon sweeping and are therefore iterative. They are intended for use with experiments that have suffered damage and therefore no longer conform to any recognised design. They could also be useful when an experiment is conducted under such physical constraints that it has been necessary to adopt some unconventional design. Another use could arise when assessing students (treatments) who opt for different papers (blocks) in an examination, the papers being possibly of varying difficulty.
TL;DR: In this article, a method for detecting outlier components in two-way contingency tables, using the hypothesis of homogeneity, based on the adjusted residuals is proposed, and the method is shown to be effective in detecting outliers.
Abstract: We propose a method for detecting outlier components in two-way contingency tables, using the hypothesis of homogeneity, based on the adjusted residuals
TL;DR: In this article, a non-normal class of distribution (Edgeworth Series distribution) function in three and four parameters has been considered for dose-binary response relationship, which accounts for the nonnormality (expressed in terms of skewness and kurtosis) present in the relationship in addition to the usual location and scale parameters.
Abstract: A non-normal class of distribution (Edgeworth Series distribution) function in three and four parameters has been considered for dose-binary response relationship. This class accounts for the non-normality (expressed in terms of skewness and kurtosis) present in the relationship in addition to the usual location and scale parameters (generally considered by two parameter models). We present the maximum likelihood method of estimation of the parameters and test of probit (normal distribution) hypothesis. Edgeworth Series distribution when fitted to the data of Milicer & Szczotka (1966) showed an excellent closeness to the observed values, significant improvement over probit and logit fit (Aranda-Ordaz, 1981), and better fit compared to Prentice (1976) model.
TL;DR: In this paper, a linear multiple regression model is fitted to a tumour volume time series and the slope parameters are used to estimate the expected lifetime extension/reduction as an unambiguous index of treatment effect.
Abstract: A linear multiple regression model is fitted to a tumour volume time series. The slope parameters are used to estimate the expected lifetime extension/reduction as an unambiguous index of treatment effect.
TL;DR: In this article, the analysis of the total expenses of medical services of an individual during a given year can be split up into two parts: the probability of having nonzero expenses, and the amount of the non-zero expenses.
Abstract: The analysis of the total expenses of medical services of an individual during a given year can be split up into two parts: the analysis of the probability of having non-zero expenses, and the analysis of the amount of the non-zero expenses. In this paper we construct a logic model for the first component, to trace which factors influence this probability. We apply three procedures to estimate the coefficients of the model, namely: maximum likelihood, logit minimum chi-square and non-linear least squares. The model is applied to Dutch health insurance data, for different sets of explanatory factors. The results suggest that the third procedure is inferior to the other two, and that the first two differ only in required CPU time.
TL;DR: Some foundational issues in statistical inference, having a bearing on influence, are discussed in this article, where the authors discuss the influence of statistical inference on influence and its influence on influence.
Abstract: Some foundational issues in statistical inference, having a bearing on ‘influence’ are discussed.
TL;DR: A measure of association remains a useful concept for two-way tables, and examples still arise where none of the published measures seems entirely appropriate.
Abstract: A large number of measures of association for two-way contingency tables has been suggested over the years. More recently, research into the analysis of contingency tables has concentrated on log-linear models and on higher-order tables, and measures of association have largely been neglected. However, a measure of association remains a useful concept for two-way tables, and examples still arise where none of the published measures seems entirely appropriate. Two such examples, one from weather forecasting and one from mouse chronic toxicity trials are described and attempts are made to construct suitable measures of association.
TL;DR: Fractional differencing is considered and it is applied to model a hospital attendance series to study the persistence phenomenon in seasonal time series.
Abstract: This paper studies the persistence phenomenon in seasonal time series. Fractional differencing is considered and it is applied to model a hospital attendance series.
TL;DR: An approach is described using correspondence analysis and cluster analysis to reveal the patterns of covariation in a large data set.
Abstract: Surveys of human populations invariably yield data sets in a near-standard cases-by-variables format, where the variables divide into a set of biographical questions and a set of questions relating to the survey's direct objective. Analysis consists of investigating the interrelationships between these two sets of variables as well as exploring variation within each set. An approach is described using correspondence analysis and cluster analysis to reveal the patterns of covariation in a large data set. Correspondence analysis elicits new continuous constructs from the data, whereas cluster analysis elicits discrete constructs, and the two techniques often complement each other. The descriptions which they provide can influence one's understanding of the data by revealing the true complexity of features that are not always apparent in a formal modelling approach. Examples are presented in the context of readership surveys.