TAS Touches on Topics Spanning
Methodology to Applications
The May 2018 issue of The American Statistician features 13 articles that span a range of methodology and application areas. There is something for everyone.
The General section begins with an article about identifiability and estimation issues that arise when parametric families are extended with extra parameters for increased flexibility. A second article investigates the effect of standardization on multicollinearity measures, and a third article discusses the construction of joint distributions from marginal distributions in such a way that constraints on the random variables are satisfied.
The lineup for Teacher’s Corner includes an experience report on a curriculum design for a professional master’s program of statistical practice. A second paper investigates incomplete data inference methods for “shaved dice.” A third paper proposes a graphical display of type-2 errors when testing for a normal distribution, and the final paper in this section develops and compares methodologies when using regression analysis to detect aging trends.
You will find two Short Technical Notes. The first develops a fast algorithm for computing the expected value of sample central moments, and the second offers alternative proofs that a Laplace distribution can be represented as a Gaussian mixture.
There are also two papers in the Statistical Practice Section. The first is a discussion about Cochran’s rule-of-thumb on the adequacy of the chi-square test for independence in a contingency table, and the second is an investigation of the effect population skew can have on sample size formulas.
Finally, there are two papers contributing to the Interdisciplinary and Statistical Computing and Graphics sections of the journal. The first studies high-school dropout rate and proposes correspondence analysis as a way to obtain additional insight. The second paper advocates for the use of symbolic computing tools and uses a context in which the efficient score test is of interest.
Have something to contribute? Submit your paper using our online submission system.