Cohpasisons of improved bonferroni and sidak/slepian bounds with applications to normal markov processes

The recent literature contains theorems improving on both the standard Bonferroni inequality (Hoover (1990)) and the Sidak/Slepian inequalities (Glaz and Johnson (1984)), The application of these improved theorems to upper bounds for non coverage of simultaneous confidence intervals on multivariate normal variables is explored. The improved Bonferroni upper bounds always hold, while improved Sidak/Slepian bounds only apply to special cases. It is shown that improved Sidak/Slepian bounds will always hold for Normal Markov Processes, a commonly occuring and easily identifiable class of multivariate normal variables. The improved Sidak/Slepian upper bound, if it applies, is proven to be superior to the computationally equivalent improved Bonferroni bound. This improvement, however, is not great when both methods are used to determine upper bounds for Type I error in the range of .01 to .10.     

Unit-Root Tests and the Statistical Pitfalls of Seasonal Adjustment: The Case of U.S. Postwar Real Gross National Product

An often-used scenario in marketing is that of individuals purchasing in a Poisson manner with their purchasing rates distributed gamma across the population of customers. Ehrenberg (1959) introduced the marketing community to this story and the resulting negative binomial distribution (NBD), and during the past 30 years the NBD model has been shown to work quite well. But the basic gamma/Poisson assumptions lack some face validity. In many product categories, customers purchase more regularly than the exponential. There are some individuals who will never purchase. The purpose of this article is to review briefly the literature that addresses these and other issues. The tractable results presented arise when the basic gamma/Poisson assumptions are relaxed one issue at a time. Some conjectures will be made about the robustness of the NBD when multiple deviations occur together. The NBD may work, but there are still opportunities for working on variations of the NBD theme.     

Projecting From Advance Data Using Propensity Modeling: An Application to Income and Tax Statistics

This article proposes and evaluates two new methods of reweighting preliminary data to obtain estimates more closely approximating those derived from the final data set. In our motivating example, the preliminary data are an early sample of tax returns, and the final data set is the sample after all tax returns have been processed. The new methods estimate a predicted propensity for late filing for each return in the advance sample and then poststratify based on these propensity scores. Using advance and complete sample data for 1982, we demonstrate that the new methods produce advance estimates generally much closer to the final estimates than those derived from the current advance estimation techniques. The results demonstrate the value of propensity modeling, a general-purpose methodology that can be applied to a wide range of problems, including adjustment for unit nonresponse and frame undercoverage as well as statistical matching.     

On Jointly Estimating Parameters and Missing Data by Maximizing the Complete-Data Likelihood

One approach to handling incomplete data occasionally encountered in the literature is to treat the missing data as parameters and to maximize the complete-data likelihood over the missing data and parameters. This article points out that although this approach can be useful in particular problems, it is not a generally reliable approach to the analysis of incomplete data. In particular, it does not share the optimal properties of maximum likelihood estimation, except under the trivial asymptotics in which the proportion of missing data goes to zero as the sample size increases.     

On Regression Analysis with Data Cleaning via Trimming,Winsorization, and Dichotomization

ABSTRACT

In a two-variable regression model setup, we examine in this paper the effects of trimming and winsorization and dichotomization of the independent variable on the regression estimates and the efficiency of the regression estimation. These results have direct implications in the accounting literature when dealing with the effects of analyst forecast errors on stock returns.     

Minicommunications

This section of the periodical is reserved for Conclusions (Results), Comments, Conjectures, and Microcommunications. Contributions will be welcomed and might even be printed!     

Military Triumph, Racial Transcendence, and Colin Powell

A "black kid of no early promise," Colin Powell became the youngestgeneral in the U.S. Army, and then in short order national securityadvisor, chairman of the Joint Chiefs of Staff, and the nation’ssecretary of state. What does this extraordinary story revealabout contemporary American political life? Analyzing surveyssupplied by the National Election Study and the National BlackElection Study, we first establish that Powell’s rapidrise to prominence is matched by his remarkable popularity amongthe American public. Next, we develop and test two possibleexplanations for Powell’s popularity. One supposes thatthe secret to Powell’s high standing with the public liesin his association with success on the battlefield: Powell asthe victorious general. The other explanation invokes racialprogress, the disappearance of racism among whites, and thedecline of identity politics among blacks: Powell as raciallytranscendent. In the final section of the article, informedby our results, we offer some speculations about American politicstoday—about the political implications of military accomplishmentand about the multiplicity of conditions that are required forAmericans to "see through" race.     

AN ANALYSIS OF PORTFOLIO REVISION STRATEGIES UTILIZING VARIABLE REVISION INTERVALS*

This paper investigates portfolio revision with an emphasis on the decision of when to revise. A statistical technique based on the sequential analysis of the time series of portfolio return relatives determines when revision is to occur. The technique detects changes in the time series which are an indication that the underlying generating process of the time series has changed and that the portfolio should be, if necessary, revised. Thus, the length of the revision interval is variable and a function of the data. The statistical technique is utilized in conjunction with three portfolio revision strategies. These three revision strategies are compared to a buy and hold policy over three nonoverlapping, 12-year investment horizons. The basis of comparison is the net terminal values which include adjustments for transaction costs and taxes. The sensitivity of the statistical technique to its parameters is also analyzed.     

A DYNAMIC OPTIMIZATION MODEL FOR PLANNING IN A MULTI-PRODUCT ENVIRONMENT

This paper describes an extension of the product mix problem and explicitly considers situations where demand can fluctuate over time. The additional decision variables and constraints that must be considered are described, and a revised objective function is discussed. The solution to the proposed linear programming formulation is compared to the traditional simulation and illustrates the potential for increased profits.     

Management Policies to Improve the Effectiveness of Multi-Station Service Organizations*

A SLAM based simulation model of a multi-station, tandem queuing structure characteristic of a variety of service systems is employed to test various design options for the system. The model is based on an extensive study of the State of Florida driver licensing offices. The multiple objectives of low time in the system for customers and the efficient use of personnel resources are employed to measure the benefits of policy options. The use of simulation analysis permits the incorporation of complex system characteristics, therefore providing a realistic representation of the effects of possible management actions. Effective methods to control labor in such systems are suggested.