Reports on a research study in which college students used a statistical ordination method to reveal relationships among plant community structures and physical, disturbance, and spatial variables. Concludes that polar ordination helps students understand the methodology of plant geography and encourages further student research. (CFR)

The use of content validity as the primary assurance of the measurement accuracy for science assessment examinations is questioned. An alternative accuracy measure, item validity, is proposed. Item validity is based on research using qualitative comparisons between (1) student answers to objective items, (2) clinical interviews, and (3) student…

This study used survival analysis to overcome problems of principled generalization and individual variability in analysis of the conversational speech of 24 children with speech delay recorded over two years. The derived normalization probabilities were lagged according to the strong delay hypothesis and results converged with previous normative…

Explores the consistency and accuracy of faculty evaluation, peer evaluation and self-evaluation of applied end-of-semester performances. Finds that faculty inter-judge reliability was uneven, while student peer inter-judge reliability was more uniform; faculty-peer evaluations were highly correlated, but self-evaluations were poorly correlated…

Identifies four additional data patterns that discriminate among five theories of students' expected grades/teacher ratings correlation. The presence of all four of these markers in student ratings data is consistent with the theory that the correlation is due to an unwanted influence of instructors' grading leniency. (MMU)

Examines coverage of critical thinking in 37 full-length introductory psychology textbooks published from 1955 to 1997. Checks for coverage of scientific thinking, cognitive biases in thinking, and difficulties in statistical reasoning. Suggests that findings should facilitate the review and decision process for instructors interested in…

A quantitative study modeled the inherently subjective admissions process for 592 graduate school candidates and 72 raters. Logistic regression models were well-fitting and parsimonious, allowing analysis of each stage of the process. Extended committee discussion/deliberation phases were of limited productivity when inter-rater agreement was…

Evaluates "College & Research Libraries" and "Journal of Academic Librarianship" on the basis of types of articles published, structure of articles, types of statistics used, and data-collection methods. Results indicate that surveys are the primary form of data collection and descriptive statistics are used more frequently…

Assesses the introduction of a computer-laboratory experience using a popular statistical package into a behavioral-science statistics class required of all sociology, psychology, and business-administration majors at a liberal arts college. Indicates that the laboratory did not improve student comprehension or performance, possibly due to poor…

Demonstrates the extent to which school-level performance in the British General Certificate of Secondary Education (GCSE) program is underpinned by variations in the socioeconomic backgrounds of students. Concludes that uncontextualized performance statistics are flawed, and that policies for school improvement must address underlying constraints…

This study used Tobit analysis to estimate retention probabilities and credit hours at two universities. Tobit was judged as appropriate for this problem because it recognizes the lower bound of zero on credit hours and incorporates this bound into parameter estimates and forecasts. Models are estimated for credit hours in a single year and…

Uses evidence from seven surveys of the U.S. 12th-grade population and the National Assessment of Educational Progress to show that gender differences in mean and variance in academic achievement are small from 1960 to 1994 but that differences in extreme scores are often substantial. (SLD)

Demonstrates the use of Rasch models to assess the dimensionality of an item set across cohorts in longitudinal studies. Presents data from a study on adolescents' deductive reasoning ability showing that the same test items can be measuring qualitatively different latent traits in different subsamples or at different time points. (JPB)

Demonstrates how Rasch models can be used to study differences in item salience within and across time points, and that even if a test does not fulfill assumptions of the Rasch model, it can be used to obtain information on item functioning across participants and over time. Illustrates method using longitudinal data from Baltimore Prevention…

Examines current standards and guidelines relative to libraries in Catholic elementary schools, using a sampling of library literature and school district documents. Discusses the importance of data; quantitative data; qualitative data; the mission of the Catholic Church; and the role of the Catholic Library Association in formulating proposed…