The Segregated Storage Problem: Some Properties and an Effective Heuristic

The segregated storage problem is formulated as an integer programming problem. Some special cases are shown to reduce to ordinary assignment problems. A special property of the integer formulation is used to develop a new heuristic procedure which produces near optimal solutions and which can be used as a tight bound in branch and bound procedures. Computational comparisons with heuristics proposed by other authors indicate this method to be superior in solving large problems.     

Fairness of psychological tests: Implications of four definitions for selection utility and minority hiring.

Expected utility losses in moving from T. A. Cleary's (1968), R. L. Thorndike's (1971), and R. B. Darlington's (1971) Definition 3 selection fairness models to the quota model were assessed on an interval scale for various combinations of validity, minority base rate, and selection ratio. Expected changes in minority selection ratios across conditions were also determined. Utility losses were shown to be large enough to be of practical significance in many commonly occurring selection situations, although when considered as a percentage of maximum value, utilities remained quite high for the Thorndike and Darlington models. Increases in minority selection ratios across models were more striking than utility losses. Because no accepted method for converting minority selection ratios to utility units exists, and in light of the fact that the legal status of all 4 models is as yet unclear, it is concluded that each personnel researcher or organization must consider the trade-off between utility and the minority selection ratio subjectively and choose the model of selection fairness most consistent with his or its values. (18 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Relative Accuracy of Connecting Channel Discharge Data with Application to Great Lakes Studies

The flows in the Great Lakes connecting channels are a major component in the water balance of the Great Lakes Basin. The increased emphasis on Great Lakes water quality and quantity requires an assessment of the accuracy of both measured and computed connecting channel discharge data. In this study, the standard error of typical discharge measurements was found to be approximately 3 to 5 percent, depending upon the number of panels used in the cross section. Mesurement sets were found to have a practical limit of about 25 measurements. The standard error of a set of measurements was found to be on the order of 1 percent. The procedure used to compute the published flows of the Niagara River was found to have an apparent bias of about 2 percent on the high side. It is recommended that the published Niagara River flows be adjusted prior to use in detailed water balance studies.     

The Golden Flea Award.

Comments on the original article by R. C. Atkinson (see record 1977-28549-001), which discusses psychology's progress during the last quarter century and its potential for significant scientific advances in the future. In the interests of balanced discussion, this letter brings attention to a lesser known award-the Golden Flea Award-that is given annually for the year's most nit-picking criticism of American science. It is argued that carrying the game to our political critics may be an approach whose time is due. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Development of competitive behavior as a function of culture, age, and social comparison.

Observed the choice behavior of 868 dyads of 2nd-, 4th-, and 6th-grade children from 5 different cultural backgrounds (Mexican-American, Anglo-American, Japanese, Greek, and Belgian) in the maximizing difference game, an outcome interdependence task where individualistic and cooperative motives were placed in opposition to competitive ones. In 4 cultures, boys received either social comparison feedback or only information concerning their own cumulative scores. As predicted, the level of competitive responding increased as a function of age, social comparison information, and number of trials. Further, there were systematic cultural differences. A newly developed covariation analysis revealed that individual dyad variation accounted for more covariation than the main effects noted above. Hence, several types of within-dyad analyses were performed. Finally, cultures were classified in terms of the characteristic play by dyad members and the results compared with a classification developed by M. Mead (1937), in which she ordered the cooperative, competitive, and individualistic orientations of members of 13 "primitive" societies. (23 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A factor analysis of assertive behaviors.

Factor analyzed the data from 5 objective measures of interpersonal behavior, presumed to be tapping "assertiveness," administered to 55 male alcoholic inpatients. In addition, each behavioral measure was correlated with a subjective global rating of assertiveness. Four of the behavioral measures loaded highly on a general factor of assertiveness. The fifth behavioral measure loaded highly on a separate factor, Response Latency. All behavioral measures with the exception of response latency evidenced significant correlations with the subjective rating of global assertiveness. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Morphology neural networks: An introduction with applications

The area of artificial neural networks has recently seen an explosion of theoretical and practical results. In this paper, we present an artificial neural network that is algebraically distinct from the classical artificial neural networks, and several applications which are different from the typical ones. In fact, this new class of networks, calledmorphology neural networks, is a special case of a general theory of artificial neural nets, which includes the classical neural nets. The main difference between a classical neural net and a morphology neural net lies in the way each node algebraically combines the numerical information. Each node in a classical neural net combines information by multiplying output values and corresponding weights and summing, while in a morphology neural net, the combining operation consists of adding values and corresponding weights, and taking the maximum value. We lay a theoretical foundation for morphology neural nets, describe their roots, and give several applications in image processing. In addition, theoretical results on the convergence issues for two networks are presented.This research was supported in part by National Science Foundation, Contract No. ECS-9010403.     

Estimation of the magnitude and spatial distribution of combustible materials in urban areas. A case study of the San Jose area,California

This paper describes a methodology for estimating the magnitude and spatial distribution of urban fuel loadings using San Jose, California, as a case study. The study was undertaken to provide basic inputs into the modeling of large urban fires ignited by nuclear explosions, which has important implications for studies of the global climatic effects of nuclear war, or the so-called ‘nuclear winter’ phenomenon. The methodology is based on the measurement of the numbers and sizes of different building occupancy types from aerial photographs for approximately 6000 1000-ft (304.8-m) grid cells covering the study area. These measurements were combined with literature-derived average fuel loadings for each building type to yield the total fuel loading per cell. The total fuel loading thus calculated is approximately 4558.56 × 10⁶ kg, yielding an average of 8.4 kg m⁻² (1.7 lb ft⁻²) for the built-up area of the city and 9.4 kg m⁻² (1.9 lb ft⁻²) if unoccupied cells (i.e. cells containing no fuel) within the built-up area are excluded. These figures are lower than previously published average values, due to the low-density character of San Jose and the exclusion of certain components of total fuel loading by the methodology.