Nonoptimal termination properties of quadratic interpolation univariate searches

A number of well-regarded multivariable nonlinear programming algorithms employ univariate searches to enhance progress toward an optimum. One of the preferred univariate searches has been Coggins' quadratic approximation-interpolation algorithm.^* In this paper, conditions are derived for which Coggins' algorithm may terminate at very poor estimates of the optimum. Simple but representative numerical examples affirm these results for a preferred coding. Interestingly, this algorithm nevertheless performs extremely well on a wide range of test problems. Similar conditions can be derived for higher order approximations, although this does not appear warranted based on the present knowledge.     

Understanding individual problem-solving style: A key to learning and applying creative problem solving

More than five decades of research and development have focused on making the Creative Problem Solving process and tools accessible across a wide range of ages and contexts. Recent evidence indicates that when individuals, in both school and corporate settings, understand their own style of problem solving, they are able to learn and apply process tools more effectively, and when teams appreciate the styles of their individual members, their problem solving efforts are enhanced. We summarize recent studies and report new data supporting the conclusion that individual style differences provide an important key to understanding the interaction of person, process, product, and press when managing change.     

Enhancement and Analysis of Science Question Level for Middle School Students

The effects of instruction and achievement on science question level for high and low science topic interests were investigated. Eight seventh‐grade classes were randomly assigned to two treatments: instruction and no instruction on researchable questioning. Each student completed the Middle School Students' Science Topic Interest Rating Scale (test‐retest reliability, r = .84); selected two topics in which she or he was least interested and two topics in which she or he was most interested; wrote questions for each topic; and took the Stanford Achievement Tests in reading, mathematics, and science. The questions were rated using the four levels described by the Middle School Students' Science Question Rating Scale (interrater reliability, r = .96). The scores for each question were averaged for two raters, then summed for each interest level for each student. The data were analyzed for main and interaction effects using general linear modeling procedures. Question level was modeled with one within‐subjects factor (science topic interest) and four between‐subjects factors (instruction and three achievement scores). The results indicate that students who received instruction outperformed those students who were not instructed; and high achievers in mathematics, reading, or science outperformed low achievers. There were no interaction effects. © 2000 John Wiley & Sons, Inc. J Res Sci Teach 37: 210–224, 2000     

Science and Values in Undergraduate Education

Science & Education - While a conception of science as value free has been dominant since Max Weber defended it in the nineteenth century, recent years have witnessed an emerging consensus that...     

Elite cricket coach education: a Bourdieusian analysis

The social structures within coach education have been largely unexplored, undiscussed, and treated as unproblematic in contributing to coach learning, both in research and practice. The study used semi-structured, in-depth interviews with 11 elite cricket coaches to gather their perceptions of an elite coach education programme. In particular, this research attempted a more nuanced critical inquiry into the impact of culture on coach learning, habitus on knowledge production and the extent to which capital structures practice within the field of cricket coach education. Data analysis followed abductive reasoning, combining inductive thematic analyses of the data, with a deductive abstraction of these themes within a Bourdieusian framework to provide a level of explanation to the data. The findings present coach education as a complex social field in which coaches were active social beings in the (re)production of coaching knowledge. The culture of cricket was found to perpetuate a powerful doxic system that highlighted the tensions and conflict between an accepted model of coach education with a singular and prescribed body of knowledge and a strong underlying sporting culture and individuals hierarchically placed within it. This data further highlighted how coach education contributes to the (re)production of power within the field of cricket coaching.     

Humor and anxiety: Effects on class test performance

Measures of trait anxiety, state anxiety, and achievement were obtained on a sample of undergraduate students, half of whom received additionaal humorous items in the achievement test. In a regression analysis, the trait anxiety × test version interaction was a significant predictor of achievement. Subsequent analyses revealed a disordinal interaction in which highly anxious students had lower achievement on the humorous test than on the nonhumorous test, and students with low anxiety had higher achievement on the humorous test than did students with high anxiety. The results do not support the popular assumption that humor is a positive factor in reducing high anxiety associated with academic evaluations.     

An Empirically Derived Index of High School Academic Rigor

We derived an index of high school academic rigor (HSAR) by optimizing the prediction of first‐year college GPA (FYGPA) based on high school courses taken, grades, and indicators of advanced coursework. Using a large data set and nominal parameterization of high school course outcomes, the HSAR index capitalizes on differential contributions across courses and nonlinear relationships between course grades and FYGPA. Test scores from eighth grade were incorporated in the model to isolate the contribution of HSAR. High school courses with the largest relationships with FYGPA were English 11, English 12, Chemistry, English 10, Calculus, and Algebra 2. Participation in Advanced Placement, accelerated, or honors courses increased HSAR. The correlation of the HSAR index and FYGPA was .52 and the HSAR index led to modest improvement in overall prediction when combined with high school GPA and ACT Composite score. HSAR index subgroup differences were smaller than subgroup differences in ACT Composite score. Implications for high school counselors, researchers, and postsecondary student service personnel are discussed.