Search

An assumption central to many tests of statistical association between two variables is the null expectation of zero association. Here, we draw attention to the fact that in many published tests of mass-dependent mass loss in breeding birds, this assumption has been violated. We show that a correct null hypothesis can be derived by using resampling methods, and analyse three data sets (two previously published) from passerine birds to illustrate the approach. Our results show, that under a correct null hypothesis, the biological interpretation of the previously published results is reversed-initially heavy birds do actually lose less mass (relative to their weight) than the initially light birds.

This paper generalizes the work of Blomqvist (1977, Journal of the American Statistical Association 72, 746-749) on inference for the relationship between the individual-specific slope and the individual-specific intercept in a linear growth curve model. The paper deals with longitudinal data involving one or more response variables and irregular follow-up times, with each response variable postulated to follow a linear growth curve model. The problem considered is inference concerning the association between one growth curve coefficient and another--for example, the slope and intercept for a selected response variable, or the two slopes for two different response variables--after adjusting for all remaining coefficients among all of the response variables. An inferential approach based on the method of moments and an inferential approach based on maximum likelihood are described, and the asymptotic properties of these procedures are presented. Extensions of the methodology to allow polynomial growth curves and baseline covariates are outlined. The methodology is illustrated with a practical example arising from a clinical trial in lung disease.

This paper is concerned with the sampling behaviour of raw and partial measures of association between categorical variables. It summarizes the asymptotic results established for raw measures and extends them in a systematic way for the derived partial associations. The validity of the asymptotic results is then stressed by means of a simulation study. Three proportional reduction in error of prediction measures are considered for nominal variables and three concordance-discordance indices for ordinal variables.

the Section wishes to emphasize the special role that social statistics has played in the history of ASA. For most of its 150 years, social statistics was the dominant force in the association, as indicated by the topical activities of its members, by the range of topics featured in the Journal of the Americani Statistical Association throughout most of this period, and by the emphasis on government data bases and statistics in the service of public policy. (p. 18)

In the context of a practitioner's view of statistics, this paper attempts to revive and clarify G. Udny Yule's concern with the effects of marginal distributions upon summary measures of association. Ignoring these effects may give you the right answer to the wrong question. In particular, when the hypothesis is causal, varying independentvariable distributions can artifactually distort association measures. The distortions of some common measures are shown in three kinds of tables. A cure, equiweighting, is explained. Reasons are given for not also equiweighting the dependent variable. A socialization example illustrates the use of equiweighting in a complex problem.

On introduit un nouveau type de mosaique pour des tables de contingence 2×2 a partir d'une modification due a Hartigan et Kleiner (1981)

The alternative ratemaking methodology proposed here is based on the use of an exponential utility function to describe an insurer's attitude toward risk. Justification for a utility theory approach is provided by considering the relationship between risk, uncertainty, and ratemaking. The specific choice of the exponential utility function results from the fact that this utility function satisfies certain axioms about ratemaking. Several of the properties of an exponential utility theory ratemaking system are Leonard R. Freifelder is assistant professor of operations research, Temple University. He holds a Ph.D. in operations research from the University of Pennsylvania. Professor Freifelder is an Associate Fellow, Society of Actuaries. This paper is based on his book A Decision Theoretic Approach to Insurance Ratemaking published by Richard D. Irwin, Inc., 1976, for the S. S. Huebner Foundation for Insurance Education. Dr. Freifelder received the 1978 C. A. Kulp Award from the American Risk and Insurance Association for the outstanding original contribution made to the literature of insurance by this book. This paper was presented at the annual meeting of the American Statistical Association in San Diego, California, August 1978.

This research was supported by grants from the National Science Foundation. Clogg was supported in part by Grant SES-7823759 and Goodman by Grant SES-8303838. Previous versions of this chapter were presented at the annual meetings of the American Statistical Association, Detroit, August 1981, and at the annual meetings of the American Sociological Association, Detroit, September 1983. For helpful comments, the authors are indebted to William Dillon, Otis Dudley Duncan, Anton Formann, Rolf Langeheine, George Macready, Thomas Madden, Peter Marsden, James Shockey, and Marylee Taylor.

This paper is concerned with the statistical aspects of the phenomenon of disease occurring more frequently in individuals with some genotypes than in individuals with others. A correlation coefficient is defined to quantify association between disease and genotype. A distinction is made between the concepts of independence of allele and disease and independence of genotype and disease. This distinction is used to define two components of association which describe separate aspects of association of disease with genotype. One component is a measure of the association of disease with allele; the other a measure of the effect of allele interaction on association of disease and genotype. One aspect of the usefulness of the partition into components which is discussed is in expressing the recurrence risk of disease for a relative of an affected individual. A chi-squared analysis is provided to test hypotheses about the components of association and other hypotheses of genetic interest. This analysis is illustrated using a study done to determine the effect of the sex-linked dwarfing gene in male chickens on resistance to E. coli infection. This analysis shows a significant allele interaction effect on resistance to disease but no association of disease with alleles. In conclusion, some extensions and limitations of the proposed concepts and procedures are discussed.

A family of coefficients for measuring monotone association is presented. These include measures of association of ordinal or interval variables such as gamma of Goodman and Kruskal, Somers's dyx , Kendall's tau, or Spearman's rho as special cases. The article shows how a large number of measures of association can be put into a single general form. These coefficients are used as a basis for defining a variety of data analysis techniques.

We present methods for the analysis of a K-variate binary measure for two independent groups where some observations may be incomplete, as in the case of K repeated measures in a comparative trial. For the K 2 X 2 tables, let theta = (theta 1,..., theta K) be a vector of association parameters where theta k is a measure of association that is a continuous function of the probabilities pi ik in each group (i = 1, 2; k = 1,..., K), such as the log odds ratio or log relative risk. The asymptotic distribution of the estimates theta = (theta 1,..., theta K) is derived. Under the assumption that theta k = theta for all k, we describe the maximally efficient linear estimator theta of the common parameter theta. Tests of contrasts on the theta are presented which provide a test of homogeneity Ha: theta k = theta l for all k not equal to l. We then present maximally efficient tests of aggregate association Hb: theta = theta 0, where theta 0 is a given value. It is shown that the test of aggregate association Hb is asymptotically independent of the preliminary test of homogeneity Ha. These methods generalize the efficient estimators of Gart (1962, Biometrics 18, 601-610), and the Cochran (1954, Biometrics 10, 417-451), Mantel-Haenszel (1959, Journal of the National Cancer Institute 22, 719-748), and Radhakrishna (1965, Biometrics 21, 86-98) tests to nonindependent tables. The methods are illustrated with an analysis of repeated morphologic evaluations of liver biopsies obtained in the National Cooperative Gallstone Study.

N THE first half of the nineteenth century the North Atlantic area saw an efflorescence of statistical societies. Over thirty of them, usually urban in name and character, sprang up in England, Scotland, Ireland, France, several German states, the United States, and Mexico. Nearly all soon withered but the Statistical Society of London and another in this country were exceptions. The former throve and half a century later became the Royal Statistical Society. The latter called itself the American Statistical Association but provided that its annual meetings should be held in Boston and notwithstanding its name was in fact through over half a century little more than the statistical society of Boston. During the same half century national organizations of natural scientists began to form. An early and important one was the Association of German Scientists founded in 1822 by a handful of scholars in that complex of countries. After a few dark years it won royal approval from Bavaria and Prussia and, more important, received the allegiance of scientists the world over after it had elected Humboldt as president. He was the recognized head of the naturalists of the world, as distinguished in his field as Napoleon or Goethe were in theirs. Babbage, the main founder of the Statistical Society of London which soon became the cradle of international statis

(1968). Association and Estimation in Contingency Tables. Journal of the American Statistical Association: Vol. 63, No. 321, pp. 1-28.

A study of the efficient allocation problem in production by the evaluation of the merits of private or corporate enterprise versus a centrally directed economy. Presented before a joint meeting of the American Statistical Association, the American Economic Association, and the Econometric Society in New York City, December 29, 1949.

I wish to report three major closely-linked developments in the American Statistical Association: (1) the appointment of a full-time executive director, (2) a dues increase, and (3) important changes in pu}lication policy in the Journal of the American Statistical Association. These developments follow many years of steady progress during which ASA has grown and advanced in all phases of our regular activities under the guidance of previous officers with great contributions by many members who conducted. national and chapter meetings, produced our publications, carried forward various committee activities, and performed the day by day work of our Association. Central in this progress during the past twelve years has been Don Riley, our devoted Executive Director. Don has served two full five-year terms in that office on a part-time basis while carrying his regular duties in the Office of Statistical Standards. For some time he has reminded the Board and Council of the possibility that an aggravation of a health problem might make it necessary for him to withdraw abruptly from the office. Don has won many friends in ASA and this warning was sad news. There has been a steady increase in the demands on the Executive Director, as on other officers, for leadership and high level services, reflecting the widening opportunities and ever more urgent need confronting ASA for fostering statistics in all its phases. From 1964 to 1966, the Council of the Association, and especially its younger members had vigorously encouraged the ASA to go further in the direction of professional activities than it has before. The financing was a roadblock. A first step was taken by Robert E. Johnson's planning committee on the future of the Association. Following up its work, last year's President, Frederick Stephan, appointed a committee to review the national office of the ASA from the point of view of the ASA's, responsibilities to the profession and its opportunities to serve its members and the nation.