A. Kimball Romney

Metric scaling : correspondence analysis

Introduction The Basic Structure of a Data Matrix Principal Components Analysis Multidimensional Preference Scaling Correspondence Analysis of Contingency Tables Correspondence Analysis of Non-Frequency Data Ordination, Seriation, and Guttman Scaling Multiple Correspondence Analysis

Research methods in social network analysis

Based on a conference held in Laguna Beach, CA, 1980. Comprises 14 essays on network representations and boundaries, relational structure, clustering and positioning of actors. No index. Annotation copyright Book News, Inc. Portland, Or.

Test theory without an answer key

A general model is presented for homogeneous, dichotomous items when the answer key is not known a priori. The model is structurally related to the two-class latent structure model with the roles of respondents and items interchanged. For very small sets of respondents, iterative maximum likelihood estimates of the parameters can be obtained by existing methods. For other situations, new estimation methods are developed and assessed with Monte Carlo data. The answer key can be accurately reconstructed with relatively small sets of respondents. The model is useful when a researcher wants to study objectively the knowledge possessed by members of a culturally coherent group that the researcher is not a member of.

Regions Based on Social Structure

Boas argued that anthropologists should make historical comparisons within well-defined regional contexts. A century later, we have many improvements in the statistical methodologies for comparative research, yet most of our regional constructs remain without a valid empirical basis. We present a new method for developing and testing regions. The method takes into account older anthropological concerns with relationships between culture history and the environment, embodied in the culture-area concept, as well as contemporary concerns with historical linkages of societies into world systems. We develop nine new regions based on social structural data and test them using data on 351 societies. We compare the new regions with Murdocks regional constructs and find that our regional classification is a strong improvement over Murdocks. In so doing we obtain evidence for the cross-cultural importance of gender and descent systems, for the importance of constraint relationships upon sociocultural systems, for the historical importance of two precapitalist world systems, and for strikingly different geographical alignments of cultural systems in the Old World and the Americas.

Predicting informant accuracy from patterns of recall among individuals

Abstract This paper examines the question of informant accuracy in reporting patterns of communication in face-to-face groups. We are attempting to establish the extent to which it is possible to predict individual differences in accuracy from the patterns of recall among informants. We use data from a series of studies by Bernard, Killworth, and Sailer ( Killworth and Bernard, 1976 , Killworth and Bernard, 1979 Bernard and Killworth 1977; Bernard et al., 1980 , Bernard et al., 1982 ) in which they collected observed behavior interaction frequencies and subsequently asked informants to recall and rate the degree of previous communication. In this paper we attempt to predict the accuracy of recall, i.e. how well each individuals ranking corresponds to the overall observed interactions for the group as a whole, by looking solely at the recall ranking. Using this method, we are able to account for a major share of the variance in accuracy among the subjects. We outline a theory and method for predicting accuracy based upon recall data that may be generalized to a variety of situations beyond social interaction data.

Predicting the structure of a communications network from recalled data

Abstract In a series of papers on informant accuracy in social network data, Bernard, Killworth, and more recently, Sailer, have concluded that “what people say, despite their presumed good intentions, bears no useful resemblance to their behavior” (Bernard, Killworth, and Sailer 1982: 63). In this paper we reanalyze one of the data sets (the technical group) utilized by Bernard, Killworth and Sailer in arriving at their conclusions. Unlike Bernard et al. we find that the observed behavior data corresponds closely to the recalled data. Using different methods of analysis we find that the verbal recall data can be used to predict structural aspects of the observed data. Two major findings emerge from our analysis: first, the more similarly two people judge the communication pattern of others, the more they interact with each other, and, second, the more two people share accurate knowledge of others, the more they interact with each other. Implications of our findings for the assertions of Bernard, Killworth and Sailer are discussed.

The implications of social structure for dominance hierarchies in red deer, Cervus elaphus L

Abstract A recent paper by Iverson & Sade (J. quant. Anthropol., 1990, 2, 61–83) suggests that red deer stags that are high in a dominance ranking are equally likely to defeat those lower in the ranking, regardless of how different they are in their relative positions. Their result contradicts the common observation that stags find it easier to defeat those far below them in the hierarchy than those close to them in the hierarchy. This paper demonstrates that the apparent contradiction can be eliminated by considering two questions about the structure of agonistic encounters: ‘Who fights with whom?’ and ‘Who fights at all?’.

Does structure find structure?: A critique of Burt's use of distance as a measure of structural equivalence

This paper examines some of the assumptions and consequences of the use of distance as a measure of structural equivalence, as implemented in Burts STRUCTURE program. We take the general perspective that for a measure to be useful it should not confound separate types of information which are theoretically and mathematically independent. The mathematical relationship between distance and the Pearson product moment correlation coefficient is presented. We show that use of distance as a measure of similarity without proper attention to appropriate standardization procedures confounds information on differences between means and differences between variances with information on the similarity of the patterns between pairs of individuals, e.g. correlation. A detailed examination of Burt and Bittners analysis of Bernard, Killworth and Sailers Ham radio operator group is presented, and it is demonstrated that use of distance as a measure of structural equivalence led to nonsensical results.

New Results in Test Theory Without an Answer Key

This paper discusses some recent developments in a new methodology for the social sciences being developed jointly by the authors. The methodology is called cultural consensus analysis, and it is designed to aggregate informant responses to test items about their common culture. The object is to determine correct knowledge about the nature of that culture. Our methods estimate informant competency parameters and “correct” answers to each item from the informant response data. The methodology is designed for anthropological field studies and related research designs. This paper studies the consequences of relaxing the axioms of our model for dichotomous items, and it presents new models for other testing formats.

Concepts of success and failure

In this paper, we present a method for eliciting and describing indigenous concepts of success and failure. The domains of success and failure are first identified by obtaining from respondents a large number of statements that describe people in terms of perceived “success” and “failure.” This pool of statements is systematically reduced and the final items are submitted to similarity judgments by additional respondents. Multidimensional scaling and cluster analysis techniques are used to analyze the similarity data. The final success and failure schemes are represented by three-dimensional models. The results are highly robust and reliable. Reliability is above .90. The methods for the collection and analysis of data are applied to four samples in the Los Angeles area. The results constitute a cultural description of the domains of success and failure that should be valid for comparable populations in the United States. The results have “external validity” since the methods are “respondent centered” and free from any researcher-imposed categories or structure. For this reason, the approach is suited for use in other cultures.