Free TAS Download
The featured article for this issue of TAS is “Strategies for Pulling the Goalie in Hockey,” by David Beaudoin and Tim B. Swartz. This article will be available to download free for a limited time.

The August issue opens with the Statistical Practice article, “Strategies for Pulling the Goalie in Hockey,” by David Beaudoin and Tim B. Swartz. The authors develop methodology to assess strategies for hockey teams to pull their goaltender. Does it pay to pull a goaltender when trailing in the game, and if so, when? Using data from the National Hockey League (NHL) and Bayesian methods, the authors develop a simulator that allows them to compare strategies.

The second article in this section is “Two Pitfalls in Survival Analyses of Time-Dependent Exposure: A Case Study in a Cohort of Oscar Nominees,” by Martin Wolkewitz, Arthur Allignol, Martin Schumacher, and Jan Beyersmann. The authors discuss the occurrence of length bias and time-dependent bias in the context of survival analysis. Using a data set with information about Oscar-nominated actors, the authors advocate and illustrate the use of multi-state models to avoid such forms of survival bias.

In the History Corner article, “Ronald Fisher and Gertrude Cox: Two Statistical Pioneers Cooperate and Collide,” Nancy S. Hall describes interactions between Fisher and Cox. After a brief introduction to each, Hall discusses Fisher and Cox’s often positive and synergistic interactions, but she also details a more difficult episode related to a paper Fisher submitted to Biometrics at a time when Cox was the editor.

The Teacher’s Corner starts with “When Can One Test an Explanation? Compare and Contrast Benford’s Law and the Fuzzy CLT,” by David Aldous and Tung Phan. The authors present ideas for the content of an upper-level undergraduate probability course. The course they have in mind is one in which probability is taught through student projects on real-world phenomena, offering the students a “lab experience” to develop their thinking about probability. In addition to suggestions for the content of such a course, the article offers notes for instructors.

The second article in this section, “A Moment-Generating Function Proof of the Lindeberg-Lévy Central Limit Theorem,” by Mark Inlow, is aimed at first-year graduate students and offers exactly what it promises in the title. Nitis Mukhopadhyay stresses the importance of being careful and precise when making statements about relationships between independence, zero covariance, and zero correlation in his article, “When Finiteness Matters: Counterexamples to Notions of Covariance, Correlation, and Independence.”

In the General section, Joseph Hirschberg and Jeanette Lye present a geometric approach for comparing the Delta and Fieller methods for the construction of confidence intervals of a ratio of parameters under normal distribution assumptions in “A Geometric Comparison of the Delta and Fieller Confidence Intervals.” In “Fixed-Width Sequential Confidence Intervals for a Proportion,” Jesse Frey develops sequential procedures for obtaining exact confidence intervals for a binomial proportion p that have a user-specified width. Frey demonstrates the use of the procedures through a simple simulation study.

In “Closed Form Prediction Intervals Applied for Disease Counts,” Hsiuying Wang proposes two new prediction intervals for the frequency with which a disease occurs. The methods are compared to existing methods, with special attention given when the disease proportion is small.

For a variety of scenarios, Russell Y. Webb, Peter J. Smith, and Abdulla Firag, in “On the Probability of Improved Accuracy with Increased Sample Size,” attempt to quantify the probability of improved inference as the result of an increase in sample size. They develop an expression for a required increase in sample size that, with a specified probability, ensures a desired improvement for inference.

In the final article of this section, “Consistency of Normal Distribution-Based Pseudo Maximum Likelihood Estimates When Data Are Missing at Random,” Ke-Hai Yuan and Peter M. Bentler show that, when variables with missing values are linearly related to observed variables, the normal distribution-based pseudo MLEs are still consistent.

Book Reviews

Applied Statistical Genetics with R: For Population-Based Association Studies
Andrea S. Foulkes

Dynamic Linear Models with R
Giovanni Petris, Sonia Petrone, and Patrizia Campagnoli

Introduction to Scientific Programming and Simulation Using R
Owen Jones, Robert Maillardet, and Andrew Robinson

An Introduction to Statistical Inference and Its Applications with R
Michael W. Trosset

Linear Programming with MATLAB
Michael C. Ferris, Olvi L. Mangasarian, and Stephen J. Wright

Making Sense of Data II: A Practical Guide to Data Visualization, Advanced Data Mining Methods, and Applications
Glenn J. Myatt and Wayne P. Johnson

Mathletics: How Gamblers, Managers, and Sports Enthusiasts Use Mathematics in Baseball, Basketball, and Football
Wayne L. Winston

A Modern Approach to Regression with R
Simon J. Sheather

Picturing the Uncertain World: How to Understand, Communicate, and Control Uncertainty Through Graphical Display
Howard Wainer

A Quantitative Tour of the Social Sciences
Andrew Gelman and Jeronimo Cortina (eds.)

Statistical Methods in e-Commerce Research
Wolfgang Jank and Galit Shmueli (Eds.)

Statistical Modelling in R
Murray Aitkin, Brian Francis, John Hinde, and Ross Darnell

Understanding Probability: Chance Rules in Everyday Life (2nd ed.)
Henk Tijms

The Manager’s Guide to Statistics
Erol A. Peköz

Multilevel and Longitudinal Modeling Using Stata (2nd ed.)
Sophia Rabe-Hesketh and Anders Skrondal

Probability and Statistical Inference (8th ed.)
Robert V. Hogg and Elliot A. Tanis