Describes a new effectiveness measure as an alternative to the traditional evaluation measures of recall and precision in information retrieval systems. The statistical approach--which compares two retrieval algorithms--is explained, the information needs of the user are considered, and an experiment with a test collection of abstracts is…

A marginal maximum likelihood estimation algorithm is presented for the ordered partition model of M. Wilson that does not require the set of available responses to be fully ordered. The model and its estimation algorithm are illustrated in a comparison of alternative scoring schemes for open-ended science items. (SLD)

Presents a probabilistic technique to retrieve passages from texts having a large size or heterogeneous semantic content. Results of experiments comparing the probabilistic technique to one based on a text segmentation algorithm revealed that the passage size affects passage retrieval performance; text organization and query generality may have an…

This study investigated and compared the properties of five methods of item exposure control within the purview of estimating examinees' abilities in a computerized adaptive testing (CAT) context. Each of the exposure control algorithms was incorporated into the item selection procedure and the adaptive testing progressed based on the CAT design…

This paper presents a study on the generation of mathematics test items using algorithmic methods. The history of this approach is briefly reviewed and is followed by a survey of the research to date on the statistical parallelism of algorithmically generated mathematics items. Results are presented for 8 parallel test forms generated using 16…

Complex conditional instructions ("if X, then do Y") are prevalent in public documents, where they typically appear in prose form. Results of two previous studies have shown that conditional instructions become very difficult to process as the structure becomes more complex. A study was designed to investigate whether this difficulty can…

Research on automated scoring is extended by comparing alternative automated systems for scoring a computer simulation of physicians' patient management skills. A regression-based system is more highly correlated with experts' evaluations than a system that uses complex rules to map performances into score levels, but both approaches are feasible.…

Discusses the design of multiversion information retrieval systems and provides a theoretical justification for the necessity of creating such systems to perform an optimal search for the user's information needs. Topics discussed include comparing query formulations; feedback algorithms; an experiment with a test collection; and the mechanism of…

Analyzes the properties of genetic sequences that cause the failure of classical algorithms used for data compression. A lossless algorithm, which compresses the information contained in DNA and RNA sequences by detecting regularities such as palindromes, is presented. This algorithm combines substitutional and statistical methods and appears to…

This project examines anaphora (the linguistic device of abbreviated subsequent reference to a concept) in information retrieval (IR) systems in order to develop procedures to recognize anaphors in text and distinguish between anaphoric and non-anaphoric uses of a given term, estimate the number of anaphors appearing in bibliographic records, and…

This study investigated the effects of the nature of the algorithm, reinforcement, and level of problem difficulty on the ability of fourth graders to solve addition problems as measured by the number of problems correctly solved and the number of columns attempted. Subjects were selected for high scores on a test of basic addition facts; only…

A study was conducted to investigate learner control of instruction in contrast to response sensitive branching algorithms with respect to two specific types of instructional decisions: (1) whether a student should enter and study a particular instructional module given his score on an associated diagnostic pretest; and (2) when a student should…

Using citations, papers and references as parameters a relatedness factor (RF) is computed for a series of journals. Sorting these journals by the RF produces a list of journals most closely related to a specified starting journal. The method appears to select a set of journals that are semantically most similar to the target journal. The…

A method for predicting freshman performance based on high school grades allows calculation of any student's likely grades in a similar university course. The method is contrasted with several more traditional predictive methods and examined in a study of 3,734 University of Western Australia students. (MSE)

Discusses the results of modeling the performance of searching large text databases via various parallel hardware architectures and search algorithms. The performance under load and the cost of each configuration are compared, and a common search workload used in the modeling is described. (Contains 26 references.) (LRW)