Showing papers on "Rand index published in 1989"

Assessment of children's health status. Field test of new approaches.

Abstract: The assessment of children's health status presents unique difficulties. These include parent-child differences in reports of functioning, knowledge of what constitutes age-appropriate functioning, obtainment of accurate information for child, and demonstration of the predictivity of health status measures. Recent measures (the Functional Status II-R and instruments from the RAND Health Insurance Experiment) address physical, social, and psychologic domains of children's health. The authors modified these instruments to develop short (7 and 14 items) questionnaires (RAND, FSQ) to assess child health. Scoring on these questionnaires was compared with traditional measures of illness severity and medical service utilization. The authors also evaluated coding illness-specific and general health limitations (FSQ-S and FSQ-G, respectively). Patients included the parents of 113 children with chronic illness (100 asthmatics). Measure stability was evaluated over a 6-month period in a subset of patients. Internal consistency (Cronbach's alpha) of the seven-item RAND measure was .78, the FSQ-S .78, and the FSQ-G .73 to .89 during repeated samplings over 6 months. The FSQ-S and Rand seven-item measure were moderately correlated (.47, P less than .001). The authors observed significant correlations among alternate codings of the FSQ and RAND and between the FSQ-S, FSQ-G, RAND, and severity measure with traditional indices of medical service utilization. Parents were more likely to attribute certain functional status problems (e.g., being tired) to illness than they were other problems (e.g., moodiness or interest in things). The findings demonstrate that these measures have acceptable psychometric properties and provide preliminary evidence of construct validity in a group of young children with asthma. Using general and specific measures will provide differing pictures of a child's functioning. No single measure completely taps the impact of illness as measured by a panel of traditional indicators of illness burden and medical service utilization.

Six clustering algorithms applied to the WAIS-R: the problem of dissimilar cluster results.

Abstract: Clusterings of the Wechsler Adult Intelligence Scale-Revised subtests were obtained from the application of six hierarchical clustering methods (N = 113). These sets of clusters were compared for similarities using the Rand index. The calculated indices suggested similarities of cluster group membership between the Complete Linkage and Centroid methods; Complete Linkage and Ward's methods; Centroid and Ward's methods; and Single Linkage and Average Linkage Between Groups methods. Cautious use of single clustering methods is implied, though the authors suggest some advantages of knowing specific similarities and differences. If between-method comparisons consistently reveal similar cluster membership, a choice could be made from those algorithms that tend to produce similar partitions, thereby enhancing cluster interpretation. Language: en

Conceptual and numerical analysis of data : proceedings of the 13th Conference of the Gesellschaft für Klassifikation e.V., University of Augsburg, April 10-12, 1989

Abstract: Section 1: Data Analysis and Classification: Basic Concepts and Methods.- Robustness of Estimation Methods against Small Sample Sizes and Nonnormality in Confirmatory Factor Analysis Models.- Probabilistic Aspects in Cluster Analysis.- Symbolic Cluster Analysis.- Analysis of Nonnormal Longitudinal Data with Generalized Linear Models.- Cluster Methods for Qualitative Data.- On Testing for and against Inequality Restrictions.- Evaluation of a Survey with Methods of Formal Concept Analysis.- A New Modification of the Rand Index for Comparing Partitions.- Relations between Models for Three-Way Factor Analysis and Multidimensional Scaling.- Algorithms in Multidimensional Scaling.- Extensions of Correspondence Analysis for the Statistical Exploration of Multidimensional Contingency Tables.- Numerical Classification of Biased Estimators.- An Agglomerative Algorithm of Overlapping Clustering.- Isotonic Regression - For a Monotone Index on a Hierarchy.- On the Interpretation of Median Relations.- Least Squares Approximation of Additive Trees.- Geometric Representation of Concept Lattices.- On Properties of Additive Tree Algorithms.- Section 2: Applications in Library Sciences, Documentation and Information Sciences.- Knowledge Structures and Knowledge Representation: Psychological Models of Conceptual Order.- Qualitative and Numerical Data in a Three-dimensional System.- New Concepts and Terms During the French Revolution. A Classification of the Neologisms According to their Origin.- Einige sprachliche Probleme bei der Arbeit an einer Klassifikation und deren Registern.- Data Analysis in Literary Studies.- Have Very Large Data Bases Methodological Relevance?.- Priority-Based Classification of Available Information - An Important Aspect of Future User Interfaces.- Section 3: Applications in Economics and Social Sciences.- Computer-Based and Quantitative Methods in Market Research.- Sample Techniques Used in Marketing Research.- Generalized Latent Class Analysis: A New Methodology for Market Structure Analysis.- The Use of the Logical Programming Language PROLOG as a Classification Method.- Classification and Selection of Consumer Purchase Behaviour Models.- Towards a New Socioeconomic Classification Scheme for Farm Households Using a Cluster Analysis Technique.- Identification of Multiple Criteria Decision Making.- Explorative Data Analysis and Macroeconomics.- A Microeconometric Study of Travelling Behaviour.- Inference Techniques in Decision Support Systems - Comparison and Example from Data Analysis.- Section 4: Applications in Natural Sciences and Computer Sciences.- The Classification of Organisms - The Hurdle of Homology -.- Limits in the Reconstruction of Phylogenetic Trees Exemplified with 5S rRNA Sequences.- The Concept of Information in Computer Science.- Exploring Homologous tRNA Sequence Data: Positional Mutation Rates and Genetic Distance.- Nucleotide Sequence Analysis of Conserved Genes from Bacteria.- An Analysis of Throughput Measurements on a Computer Network.- AC-Characteristics and the Pass/Fail Performance of a Memory Chip.- Authors.

A New Modification of the Rand Index for Comparing Partitions

Abstract: The comparison of different partitions is an essential problem in evaluating numerical classifications. Measures of the similarity between partitions can be used not only to compare the results obtained from different cluster analyses, but also to compare the known structure of a data set with the result of a cluster analysis. A large number of such indices have been proposed. Reviews and comparisons are, for example, given in (1, 5, 7, 11, 12, 15). Especially (11) appears to reveal the fact that a modification of the Rand index (14) as described in (8) is particularly suitable for measuring the similarity between partitions.