High-breakdown linear discriminant analysis

The classification rules of linear discriminant analysis are defined by the true mean vectors and the common covariance matrix of the populations from which the data come. Because these true parameters are generally unknown, they are commonly estimated by the sample mean vector and covariance matrix of the data in a training sample randomly drawn from each population. However, these sample statistics are notoriously susceptible to contamination by outliers, a problem compounded by the fact that the outliers may be invisible to conventional diagnostics. High-breakdown estimation is a procedure designed to remove this cause for concern by producing estimates that are immune to serious distortion by a minority of outliers, regardless of their severity. In this article we motivate and develop a high-breakdown criterion for linear discriminant analysis and give an algorithm for its implementation. The procedure is intended to supplement rather than replace the usual sample-moment methodology of discriminant analysis either by providing indications that the dataset is not seriously affected by outliers (supporting the usual analysis) or by identifying apparently aberrant points and giving resistant estimators that are not affected by them.

Breast cancer five-year survival, by New South Wales regions, 1980 to 1991

Breast cancer five-year relative survival was calculated for 16 urban and rural regions in New South Wales (NSW) for cases incident in 1980-1991. Survival analysis employed cancer registry data linked with the death register, and age- and period-matched regional mortality of NSW women, Proportional hazard regression analysis was used to compare excess mortality in breast cancer cases in each region. The effect of region was significant (P < 0.05) in the analysis, after age and the follow-up variable (and their intel action) were adjusted for, although no region was significantly different from the referent group (chosen because of average relative five-year survival). When degree of spread and its interactions were entered into che model, the effect of region became nonsignificant. A significant linear trend (P < 0.05) in the adjusted relative risk for excess mortality in breast cancer cases was noted when regions were divided into quartiles based on socioeconomic status, with higher relative risk in low-socioeconomic-status groups; this effect also disappeared with adjustment for degree of spread at diagnosis. There was no general effect of rurality versus capital city or other metropolitan centres. This study demonstrates a small effect of region of residence and implied socioeconomic status on breast cancer survival in NSW women, but this becomes nonsignificant when the data are adjusted for degree of spread at diagnosis, This suggests that earlier diagnosis would he of benefit in reducing minor inequalities in breast cancer survival in NSW women.

Applying linear time-varying constraints to econometric models: With an application to demand systems

When linear equality constraints are invariant through time they can be incorporated into estimation by restricted least squares. If, however, the constraints are time-varying, this standard methodology cannot be applied. In this paper we show how to incorporate linear time-varying constraints into the estimation of econometric models. The method involves the augmentation of the observation equation of a state-space model prior to estimation by the Kalman filter. Numerical optimisation routines are used for the estimation. A simple example drawn from demand analysis is used to illustrate the method and its application.

Evaluation of bioelectrical impedance for prospective nutritional assessment in cystic fibrosis

We have compared the use of bioelectrical impedance analysis (BIA) with anthropometry for the prediction of changes in total body potassium (TBK) in a group (n = 31) of children with cystic fibrosis. Linear regression analysis showed that TBK was highly correlated (r > 0.93) with height(2)/impedance, weight, height, and fat-free mass (FFM) estimated from skin-fold measurements. Changes in TBK were also correlated, but less well, with changes in height(2)/impedance, weight, height, and FFM (r = 0.69, 0.59, 0.44, and 0.40, respectively). The children were divided into two groups: those who had normal accretion of TBK (> 5%/y) and those who had suboptimal accretion of TBK (< 5%/y). Analysis of variance showed that the significant difference in the change in TBK between the groups was detectable by concomitant changes in impedance and weight but not by changes in height, FFM, or weight and height Z scores. The results of this study suggest that serial BIA measures may be useful as a predictor of progressive undernutrition and poor growth in children with cystic fibrosis. (C) Elsevier Science Inc. 1997.

Children's cooperative behavior and interactions in trained and untrained work groups in regular classrooms

This study examined group and individual factors that facilitate changes in cooperation and learning outcomes in trained and untrained work groups of elementary school-age children. The study had two foci. The first was to determine if the cooperative behaviors and interactions of children in classroom groups who were trained in cooperative learning skins were different from those of children who were given no training, and the second was to investigate small group interactions and achievement in these groups over time. The results showed that there were observable differences between student interactions in the two conditions and these differences were maintained over time. Compared with children in the untrained groups, those in the trained groups were consistently more cooperative and helpful to each other; they actively tried to involve each other in the learning task by using language which was more inclusive (e.g., frequent use of ''we''), and they gave more explanations to assist each other as they worked together. It appeared that as the children worked together over time, they became more responsive to the learning needs of each other. Furthermore, the children in the trained groups performed significantly better on the learning outcomes questionnaire than those in the untrained groups. (C) 1997 Society for the Study of School Psychology. Published by Elsevier Science Ltd.

A high-resolution genetic map of the familial Mediterranean fever candidate region allows identification of haplotype-sharing among ethnic groups

Familial Mediterranean fever (FMF) is a recessive disorder of inflammation caused by mutations in a gene (designated MEFV) on chromosome 16p13.3, We have recently constructed a 1-Mb cosmid contig that includes the FMF critical region. Here we show genotype data for 12 markers from our physical map, including 5 newly identified microsatellites, in FMF families. Intrafamilial recombinations placed MEFV in the similar to 285 kb between D16S468/D16S3070 and D16S3376. We observed significant linkage disequilibrium in the North African Jewish population, and historical recombinants in the founder haplotype placed MEFV between D16S3082 and D16S3373 (similar to 200 kb). In smaller panels of Iraqi Jewish, Arab, and Armenian families, there were significant allelic associations only for D16S3370 and D16S2617 among the Armenians. A sizable minority of Iraqi Jewish and Armenian carrier chromosomes appeared to be derived from the North African Jewish ancestral haplotype. We observed a unique FMF haplotype common to Iraqi Jews, Arabs, and Armenians and two other haplotypes restricted to either the Iraqi Jewish or the Armenian population. These data support the view that a few major mutations account for a large percentage of the cases of FMF and suggest that same of these mutations arose before the affected Middle Eastern populations diverged from one another. (C) 1997 Academic Press.

Student characteristics and academic success: An econometric study of the performance of "equity groups"

Like previous volumes in the Educational Innovation in Economics and Business Series, this book is genuinely international in terms of its coverage. With contributions from nine different countries and three continents, it reflects a global interest in, and commitment to, innovation in business education, with a view to enhancing the learning experience of both undergraduates and postgraduates. It should prove of value to anyone engaged directly in business education, defined broadly to embrace management, finance, marketing, economics, informational studies, and ethics, or who has responsibility for fostering the professional development of business educators. The contributions have been selected with the objective of encouraging and inspiring others as well as illustrating developments in the sphere of business education. This volume brings together a collection of articles describing different aspects of the developments taking place in today’s workplace and how they affect business education. It describes strategies for breaking boundaries for global learning. These target specific techniques regarding teams and collaborative learning, transitions from academic settings to the workplace, the role of IT in the learning process, and program-level innovation strategies. This volume addresses issues faced by professionals in higher and further education and also those involved in corporate training centers and industry.