Attitudes of biostatisticians toward implementation of the Bayesian paradigm have changed during the past decade due to the increased availability of computational tools for realistic problems. Empirical Bayes' methods, already widely used in the analysis of longitudinal data, promise to improve cancer incidence maps by accounting for overdispersion and spatial correlation. Hierarchical Bayes' methods offer a natural framework in which to demonstrate the bioequivalence of pharmacologic compounds. Their use for quantitative risk assessment and carcinogenesis bioassay is more controversial, however, due to uncertainty regarding specification of informative priors. Bayesian methods simplify the analysis of data from sequential clinical trials and avoid certain paradoxes of frequentist inference. They offer a natural setting for the synthesis of expert opinion in deciding policy matters. Both frequentist and Bayes' methods have a place in biostatistical practice.

Most medical trials are monitored for early evidence of treatment differences or harmful side effects. In this paper we review and critique various statistical approaches that have been proposed for the design and analysis of sequential experiments in medical applications. We discuss group sequential tests, stochastic curtailment, repeated confidence intervals, and Bayesian procedures. The role that a statistical stopping rule should play in the final analysis is examined.

William Playfair is a key figure in the history of quantitative graphics. He was a popularizer and propagandist, a prolific designer of charts, and a developer of economic and business graphics. He established the line graph (especially the simple surface chart) as an important alternative to the table for the nonspecialist reader. Although his charts were not quite so original as some have supposed, they do contain graphic design ideas of great interest and occasional brilliance. His Commercial and Political Atlas of 1786 was a notable venture; it began his 36-year career as a graphic communicator. Playfair's understanding of the psychology of the graphic method was remarkable; there was an inner coherence to his picture of the reader's psychological needs which is strikingly modern (the paper contains excerpts from his many publications).

Visual display based on data deserves careful attention to a long list of ideas and questions (19 are discussed below). While classical views of graphical display need to be re-examined and selectively used, the computer--mainly as display maker, but significantly as number cruncher--has so greatly enhanced our potentialities that we have much to explore and many important steps to take. In particular, we need to pay serious and continuing attention to securing: (a) immediate and strong impact, (b) easy flow of attention across parallel elements, (c) planning to show phenomena, not numbers, (d) attention to both prospecting for what the data might show and transfer (to others) of what we have learned from it, (e) partnership with computation, and (f) putting disproportionate response to work. The next decade or two should see major advances.

This paper discusses the similarities and differences in Playfair and Tukey's visions of what graphically displaying quantitative phenomena can do now, and might do in the future. As part of this discussion we examine: (1) how fundamental graphic tools have become to the scientist, (2) three instances where modern views of graphics are unchanged since Playfair's time, and (3) one area where there has been a change. The paper concludes with a discussion of five important areas of current and future graphic concern.