Research for Applied Statistics

Syndicate content
Effect of the Medium of Administration on ASVAB Item Response Curves
April 1, 1986
Within a few years the Department of Defense may begin administering the Armed Services Vocational Aptitude Battery (ASVAB) using Computerized Adaptive Testing (CAT). Each test item is characterized by an Item Response Curve (IRC), which describes how the probability of correctly answering the item increases with ability. One important question in the CAT project is whether the IRCs are the same in paper-pencil and CAT administrations. This paper examines this issue.
Read More | Download Report
Applied Mathematicians and Naval Operators
March 1, 1982
This paper provides a personal view of applied mathematics in a nonacademic setting based on experiences as a field representative. It describes the principles necessary for effective analysis in this type of setting, and the traits a mathematician needs to effect changes as the results of his analysis.
Read More
ESTIMATION OF SOME LIMITED DEPENDENT VARIABLE MODELS WITH APPLICATION TO HOUSING DEMAND
March 1, 1982
This paper proposes a complete model which allows for the simultaneous determination of whether or not to own, and how much to spend.
Read More | Download Report
THE ANALYSIS OF DYNAMICALLY INTERACTIVE SYSTEMS (AIR COMBAT BY THE NUMBERS)
January 1, 1979
This paper analyzes the dynamic interactions involved in air combat maneuvering using a continuous analysis technique. An appendix summarizes U.S. and foreign fighter performance comparisons.
Read More
STRUCTURAL ESTIMATION BY K-CLASS METHODS AND REDUCED FORM FORECASTING
June 1, 1974
In this study, small-disturbance asymptotic approximation methods are employed to study the properties of forecasts based upon directly estimated reduced forms and upon k-class induced reduced forms. The actual errors in forecasts are approximated by random variables close in probability to the actual forecast errors, and the moments of the approximate forecast errors are studied in order to gain insight about the behavior of the actual forecast errors. Theorems are presented giving the small-disturbance asymptotic bias and mean squared error of forecasts based upon reduced forms estimated directly or derived from the k-class estimates of structural parameters.
Read More | Download Report
INTRODUCTION TO FOURIER ANALYSIS OF DATA
June 1, 1974
This is an introduction to the basic elements of Fourier analysis of a time series that is a sum of deterministic components and a stationary random process. The various mathematical concepts relating to Fourier analysis are presented in a basically intuitive manner.
Read More | Download Report
USERS GUIDE FOR GENERALIZED FACTOR ANALYSIS PROGRAM (FACTAN)
June 1, 1974
This paper discusses the procedure of factor analysis and provides details on the use of a computer program which performs a principal components analysis, tests of significance, and rotation of factors determined from a matrix of intercorrelations of up to 99 variables.
Read More | Download Report
RELATING FACTOR ANALYTICALLY DERIVED MEASURES TO EXOGENOUS VARIABLES
June 1, 1974
This paper presents a method by which factors and exogenous variables can be correlated without computing factor scores.
Read More | Download Report
COMPARISON OF FORECAST ACCURACY WHEN THE DISTURBANCES ARE SMALL: DIRECTLY ESTIMATED REDUCED FORMS VS. K-CLASS INDUCED REDUCED FORMS
June 1, 1974
This paper considers forecasts based upon reduced forms estimated directly, and, reduced forms derived from structural equations estimated by k-class methods. Using the method of small-o asymptotics, the asymptotic expected mean squared error of any linear combination of forecasts is derived in both cases. Except in special cases, we show that neither of the two methods of estimation dominates in the sense of having a lower expected mean squared error, regardless of which k-class method is used. This result confirms earlier findings of the same nature based upon a large sample asymptotic efficiency criterion.
Read More | Download Report