Presents a single framework for conducting topic analysis that performs both topic identification and text segmentation. Key characteristics of the framework are: representing a topic by means of a cluster of words closely related to the topic; and employing a stochastic model, called a finite mixture model, to represent a word distribution within…

Discusses the interdependencies between parameter estimation and properties of probabilistic models, such as dependency assumptions, binary vs. nonbinary features, and estimation sample selection. An optimum estimation for binary features applicable to information retrieval is defined, a method for computing this estimation using empirical data is…

Argues that a major difficulty in using Lotka's law in information science arises from the misuse of goodness of fit tests in parameter estimation. Three approaches for studying Lotka's law are presented: an index approach, a time series approach, and a generating mechanism incorporating these two influential variables to derive an equilibrium…

Describes a five-number summary which is a display of the minimum value, lower quartile, median, upper quartile, and maximum value. Indicates how to draw box plots as graphical representations of a five-number summary. (YP)

The perception of an "explosion" of education litigation expressed by federal officials and by the relevant literature is incorrect. Summarizes the findings of previous research and of the present study to answer the frequency question, and reveals findings as to United States Supreme Court decisions. (MLF)

Discusses the effect of including relevance data on the calculation of term discrimination values in bibliographic databases. Algorithms that calculate the ability of index terms to discriminate between relevant and non-relevant documents are described and tested. The results are discussed in terms of the relationship between term frequency and…

Eight experiments examined whether certain human problem-solving mechanisms should be expected to represent probability information in terms of frequency. Findings are consistent with literature indicating that frequentist representations eliminate various cognitive biases, including overconfidence, the conjunction fallacy, and base rate neglect.…

Proposes and explains a modification of the mixed Poisson model for library circulations which takes into account the periods when a book is out on loan and therefore unavailable for borrowing. Highlights include frequency of circulation distributions; negative binomial distribution; and examples of the model at two universities. (Contains 34…

This paper focuses on similarities of optimal design of fixed-form tests, adaptive tests, and testlets within the framework of the general theory of optimal designs. A sequential design procedure is proposed that uses these similarities to obtain consistent estimates for the trait level distribution. (SLD)

Develops and evaluates the linkage of the Armed Services Vocational Aptitude Battery to the mathematics scale of the National Assessment of Educational Progress. The accuracy of the proficiency distribution estimated from the projection was close to the accuracy of the distribution estimated from the large scale assessment. (SLD)

Simulated data were used to investigate the impact of characteristics of threshold values (number, symmetry, and distance between adjacent threshold values) and delta values on the distribution of item information in the successive intervals Rasch model. Implications for computerized adaptive attitude measurement are discussed. (SLD)

Investigates the BILOG computer program's ability to recover known item parameters for different numbers of items, examinees, and variances of the prior distributions of discrimination parameters for the two-parameter logistic item-response theory model. For samples of at least 250 examinees and 15 items, simulation results support using BILOG.…

Examines the statistical distribution of relevance judgments by reviewing and comparing the results of several studies over the last three decades. A characteristic distribution is examined, and a hypothesis is offered to explain it based on human judging characteristics and the nature of relevance experimentation. (Contains 13 references.) (LRW)

Methods of testing relatively complete (distributional) models of internal psychological processes are described. It is shown that there is a sufficient condition for additive models to imply this property of the likelihood ratio. Also discussed are the examination of hazard rate functions of component processes and change in cumulative…

The problem of monitoring trends for changes at unknown times is considered. Statistics that permit one to focus high power on a segment of the monitored period are studied. Numerical procedures are developed to compute the null distribution of these statistics. (Author)