Across 60 years, John W. Tukey contributed to the advancement of democracy, peace and industry via development, application and teaching of knowledge. In his nation's service, he contributed to the Nike missile defense, U-2 spy plane, surveillance satellites in space, hydrophones in the oceans, seismic data interpretation and communications code breaking. As computer and communication pioneer, Tukey collaborated with von Neumann, Shannon and Pierce; coined "bit'' and "software''; applied statistical time series methods to processing signals; and recognized the usefulness of fast Fourier transform algorithms to digital processing of correlated data. Practical problems inspired Tukey to invent new ways to analyze data. As teacher and author, he made these available to others. Tukey advised government and industry regarding environmental quality, educational testing, the census, pharmaceutical efficacy, manufacturing quality and technologies for gathering intelligence. This paper explores the civic career, influences and philosophies of a practicing data analyst, inventor and remarkable public servant.

From the time that John W. Tukey started to do serious work in statistics, he was interested in problems and techniques of data analysis. Some people know him best for exploratory data analysis, which he pioneered, but he also made key contributions in analysis of variance, in regression and through a wide range of applications.This paper reviews illustrative contributions in these areas.

John Tukey's impact on statistics, and on science in general, is broad and lasting. This article discusses some of these contributions, with a special emphasis on those that led to the development of robust methods and data exploration. In view of today's emphasis on data mining techniques, the recollection of Tukey's influence is especially timely.

John Tukey was a Bell Labs employee for 40 years. In that time, he influenced very many researchers and contributed significantly to the growth and luster of the Bell Labs Statistics Research Department. His counsel was valued by senior management, and his involvement in many applied problems led to important advances, including his work on spectral analysis. I will survey this and other contributions.

John Tukey was an inspiring teacher, encouraging his students to think by expressing new ideas in provocative ways, and by tolerating a wide range of questions and discussion in class.

Categorical---or qualitative---time series data with random time-dependent covariates are frequently encountered in diverse applications as the list of examples shows. As with "ordinary'' time series, the data analyst is faced with the same problems of modeling, estimation, model checking, diagnostics and prediction. The present work shows that these questions can be attacked by means of regression theory for categorical time series whose foundation is based on generalized linear models and partial likelihood inference. A variety of models are provided to illustrate the selection of the link function and recent large sample results are reviewed. The theory is developed without resorting to the Markov assumption and to the notion of stationarity. Moreover, regression methods for categorical time series allow for parsimonious modeling and incorporation of random time-dependent covariates as opposed to other procedures. In particular, nominal and ordinal time series are analyzed and compared empirically to Markov chains and mixture transition distribution models.

In the longitudinal regression setup, interest may be focused primarily on the regression parameters for the marginal expectations of the longitudinal responses, the longitudinal correlation parameters being of secondary interest. Second, interest may be focused on both the regression and the longitudinal correlation parameters. Under the first setup, there exists a "working'' correlation matrix based generalized estimating equation (GEE) approach for the estimation of the regression parameters. Under the second setup, there exist two approaches for the joint estimation of the regression and the longitudinal correlations. In one approach, true longitudinal correlations are modeled and the regression and the true correlation parameters are jointly estimated based on a GEE approach. The second approach avoids the specification of the true longitudinal correlation structure and deals with the joint estimation of the regression and a vector of "working'' correlation parameters. In this second approach under the second setup, there again exist two joint estimation methods, one requiring moments up to order 4 and the other somehow using moments up to order 2 for the construction of the estimating equations for the "working'' correlation parameters. In this paper, we first provide an outline of the desirable features and drawbacks of each of these four existing approaches. By using a general autocorrelation structure to model the true longitudinal correlations, we then provide an outline of the advantages of three new approaches. In the first new approach, the true longitudinal correlations are estimated by the method of moments, whereas the regression estimates are obtained based on a generalized quasi-likelihood (GQL) estimation approach. The other two new approaches simultaneously estimate the regression and the true longitudinal correlation parameters. It is shown through a simulation study that, among these three new approaches, the first approach performs the best in estimating both the regression and the true correlation parameters, even though the longitudinal correlations are estimated separately by the method of moments.

A general expository description is given of the use of quadratic score test statistics as inference functions. This methodology allows one to do efficient estimation and testing in a semiparametric model defined by a set of mean-zero estimating functions. The inference function is related to a quadratic minimum distance problem. The asymptotic chi-squared properties are shown to be the consequences of asymptotic projection properties. Shortcomings of the asymptotic theory are discussed and a bootstrap method is shown to correct for anticonservative testing behavior.