# The American Statistician

The American Statistician is a statistical journal for professionals. Since it was founded in 1947 it is published quarterly in February, May, August and November. The American Statitician is published by the American Statistical Association.Subjects for The American Statistician include statistics. The editor is John Stufken.

## Articles from Vol. 49, No. 3, August

A Critical Look at Some Analyses of Major League Baseball Salaries
1. INTRODUCTION Before the 1988 Annual Statistical Meetings, the Statistical Graphics Section of the American Statistical Association made available salary data for 439 major league baseball players, along with various career and 1986 performance statistics...
Correspondence Analysis in Statistical Package Programs
1. INTRODUCTION This review examines the technique termed correspondence analysis (CA), which is a multivariate data analysis method that uses the singular value decomposition (SVD) to analyze contingency tables from multinomial variables. The main...
Equating Poisson and Normal Probability Functions to Derive Stirling's Formula
1. INTRODUCTION Stirling's formula n! [similar to] [square root of 2[Pi]n][(n/e).sup.n] (1) is often presented without proof in introductory probability courses. An exception is Feller (1968), who uses the sequence [d.sub.n] = log n! - (n + 1/2)log...
Examples of Differing Weighted and Unweighted Estimates from a Sample Survey
1. INTRODUCTION An ongoing debate in the statistical literature concerns how best to incorporate the survey design into the models and analysis of complex survey data (Smith 1976, 1984; Hansen, Madow, and Tepping 1983; Fienberg 1989; Kalton 1989). One...
Imposing Views, Imposing Shoes: A Statistician as a Sole Model
1. INTRODUCTION To paraphrase the noted American philosopher of a few decades past, Vincent Lombardi, "Image isn't everything. It is the only thing." In a recent issue of The American Statistician, 23 pages were devoted to a debate among 12 statisticians...
Meta-Analysis Software: A Comparative Review
1. INTRODUCTION Meta-analysis is broadly defined as a quantitative method for combining information across related but independent studies (Hedges and Olkin 1985). This methodology has a long history (for example, as early as Pearson 1904; Birge 1932)...
Optimal Stopping in "The Showcase Showdown."
1. INTRODUCTION Even before the controversy over the Monte Hall problem (Morgan, Chaganty, Dahiya, and Doviak 1991), the first author routinely discussed it and other examples of probability applied to everyday life in several of his classes. (For a...
Sample Sizes and the Central Limit Theorem: The Poisson Distribution as an Illustration
In many introductory statistics courses, the importance of the central limit theorem is rightly stressed as a justification for the widespread occurrence and assumption of the normal distribution. It is easy to find convincing examples that the theorem...
Statistical Artifacts in the Ratio of Discrete Quantities
1. INTRODUCTION The ratio is one of the most frequently used statistics. It is also one of the most frequently misused statistics (Schor 1968; Westwick 1973; Tamari 1978). Common mistakes in the use and interpretation of the ratio include: naively comparing...
Statistics Education Fin De Siecle
1. INTRODUCTION Like the last fin de siecle, the end of the 20th century features rapid change and the dissolution of old ways. The U.S.S.R. is gone, and IBM is tottering. Research universities, which have dominated the U.S. academic scene since World...
Statistics in the Semiconductor Industry: A Competitive Necessity
1. INTRODUCTION The use of proper statistical techniques has played an important role in maintaining the competitive posture of the U.S. semiconductor manufacturing industry. This is a highly competitive industry where the costs of new manufacturing...
The Horvitz-Thompson Theorem as a Unifying Perspective for Probability Sampling: With Examples from Natural Resource Sampling
1. INTRODUCTION An observation made by Stuart (1964), "Sample survey theory seems, more than most branches of statistics, to suffer the lack of a unifying thread on which to string the various topics of which it is composed," expresses a view probably...