Otis Dudley Duncan Leo A. Goodman, Robert M. Hauser, Yu Xie, and Mustafa Emirbayer Click for larger view View full resolution Otis Dudley Duncan 2 december 1921 . 16 november 2004 [End Page 177] Otis Dudley Duncan died at the age of 82 on November 16, 2004. It is our view, and we expect that it is also the view of all others who are able to judge quantitative sociology, that Otis Dudley Duncan was the most important quantitative sociologist in the world in the latter half of the 20th century. One of Dudley’s special characteristics was the breadth of his interests: scholarly as well as non-scholarly interests, sociological as well as non-sociological interests, scientific as well as non-scientific interests; and other kinds of interests, as well: classical and contemporary music, electronic music, musicology, poetry, and graphic design. Another of his special characteristics was the fact that he knew when to wrap up his work in one area of study and turn to another—a characteristic that his many successful students did not always share. In rough chronological order, several of the main themes of his work covered the following areas of study: demography and human ecology, social stratification, the social psychology of social mobility, social indicators, statistical methods and models, quantitative reasoning in sociology and demography and social science more generally, measurement and scientific method, and the sociological analysis of music. A complete bibliography of Duncan’s published work is appended, as prepared by Professor Mustafa Emirbayer of the University of Wisconsin–Madison. We mention several highlights of Otis Dudley Duncan’s scientific contributions. Throughout his career, Duncan took interest in a subject, studied it exhaustively, developed new concepts and methods appropriate to it, and, then, after making a major contribution, took on another topic. Two constancies in his work were a focus on populations and an avoidance of typological thinking. Ultimately, he recognized that the fundamental problem vexing sociological inquiry was population heterogeneity. In opening up new areas of inquiry and tools appropriate to them, Duncan was often followed by others who hoped to extend and improve upon his methods and findings. But he never looked back, except with a fresh perspective. Defining Demography One early indication of Duncan’s concern with populations was his collaborative editorial work, with Philip M. Hauser, on The Study of Population, a classic volume that not only covered every aspect of the field as then understood, but also introduced what has become the standard definition of demography: “Demography is the study of the size, territorial distribution, and composition of population, changes therein, and the components of such changes, which may be identified [End Page 178] as natality, mortality, territorial movement (migration), and social mobility (change of status)” (Hauser and Duncan 1959, 2). The Warner School of Social Stratification In collaboration with Harold Pfautz, a fellow graduate student at the University of Chicago, Duncan wrote a devastating critique of the then-popular works on social stratification of the anthropologist, W. Lloyd Warner (Pfautz and Duncan 1950). Pfautz and Duncan leveled three major arguments against the validity of the work of Warner and his associates. One was that Warner, a social anthropologist, had failed to consider the differences between the small towns that he studied (e.g., “Jonesville”) and the larger and largely urban society that America had become—or to acknowledge others’ work on social stratification and inequality in that society. A second was that Warner’s work was ideological in character by espousing maintenance of and accommodation to existing stratification regimes. The third, two-part criticism was methodological. Warner’s methods relied on a scheme called evaluated participation (EP) to determine the class positions of individuals in a community. In the words of Pfautz and Duncan, “It involves the content analysis of ‘indirect interviews’ to ‘discover’ the social class system, as well as to get at the ‘social participation’ and ‘status reputation’ of community members to whom references are made in the body of the interviews. Data from this source are summarized on ‘status personality cards’ and form the basis for an ‘estimate of class position’” (1950, 206–207). However, Warner...

In Trond Petersen’s “Multiplicative Models for Continuous Dependent Variables: Estimation on Unlogged versus Logged Form” (this volume, pp. 113–164), the following phenomenon is noted: With the arithmetic and geometric means of a nonnegative quantitative variable measured for two groups (say, groups 1 and 2), it is possible for the arithmetic mean [Formula: see text] for group 1 to be larger than the arithmetic mean [Formula: see text] for group 2 and the geometric mean [Formula: see text] for group 1 to be smaller than the geometric mean [Formula: see text] for group 2. In the present note, some new formulas are introduced that will help to make clear when this phenomenon will occur and when it will not occur. The phenomenon considered here is of interest in situations where, for example, as noted in Petersen’s paper, results obtained using the raw form for a continuous dependent variable are compared with the corresponding results obtained using the logged form for the dependent variable. In addition, the formulas that are introduced in the present note also provide a new way to view the magnitude of the difference between the arithmetic mean and the geometric mean.

Paul Felix Lazarsfeld was a major figure in the development of modern empirical methods in sociology and the social sciences. He laid many of the foundations for social survey techniques and other empirical methods for understanding key aspects of contemporary society, such as voting studies, opinion polling, panel studies, and mass media research. The American Sociological Association Methodology Section honors each year a distinguished sociologist with the Paul F. Lazarsfeld Award for a career of distinguished contributions to the field of sociological methodology. Paul Lazarsfeld’s methodological work that he was most proud of was his introduction of the concept of latent classes, latent class analysis, and latent structure analysis. We consider in this paper Lazarsfeld’s institutional impact through this ASA award though we focus our discussion on his work and his impact on the development of latent class/structure analysis.

We craft a historically sensitive model of entrepreneurship linking individual actors to the evolving social structures they must navigate to acquire resources and launch new ventures. Theories of entrepreneurship and industry evolution suggest two opposing hypotheses: as an industry develops, launching a new venture may become more difficult for all but industry insiders and the socially prominent because of competition from large incumbents, or it may become easier for all people because the legitimacy accorded to the industry simplifies the entrepreneurial task. To test these two conflicting claims, we study the American magazine industry from 1741 to 1860. We find that magazine publishing was originally restricted to publishing-industry insiders, professionals, and the highly educated, but most later founders came from outside publishing and more were of middling stature. Gains by entrepreneurs from the social periphery, however, were uneven: most were doctors and clergy without college degrees in small urban areas; magazines founded by industry insiders remained predominant in the industry centers. Our analysis demonstrates the importance of grounding studies of entrepreneurship in historical context. It also shows that entrepreneurship scholars must attend to temporal shifts within the focal industry and in society at large.

We craft a historically sensitive model of entrepreneurship linking individual actors to the evolving social structures they must navigate to acquire resources and launch new ventures. Theories of entrepreneurship and industry evolution suggest two opposing hypotheses: as an industry develops, launching a new venture may become more difficult for all but industry insiders and the socially prominent because of competition from large incumbents, or it may become easier for all people because the legitimacy accorded to the industry simplifies the entrepreneurial task. To test these two conflicting claims, we study the American magazine industry from 1741 to 1860. We find that magazine publishing was originally restricted to publishing-industry insiders, professionals, and the highly educated, but most later founders came from outside publishing and more were of middling stature. Gains by entrepreneurs from the social periphery, however, were uneven: most were doctors and clergy without college degrees in small urban areas; magazines founded by industry insiders remained predominant in the industry centers. Our analysis demonstrates the importance of grounding studies of entrepreneurship in historical context. It also shows that entrepreneurship scholars must attend to temporal shifts within the focal industry and in society at large.

We trace the social positions of the men and women who found new enterprises from the earliest years of one industry’s history to a time when the industry was well established. Sociological theory suggests two opposing hypotheses. First, pioneering entrepreneurs are socially prominent individuals from fields adjacent to the new industry and later entrepreneurs are from an increasingly broad swath of society. Second, the earliest entrepreneurs come from the social periphery while later entrepreneurs include more industry insiders and members of the social elite. To test these hypotheses, we study the magazine industry in America over the first 120 years of its history, from 1741 to 1860. We find that magazine publishing was originally restricted to industry insiders, elite professionals, and the highly educated, but by the time the industry became well established, most founders came from outside publishing and more were of middling stature – mostly small-town doctors and clergy without college degrees. We also find that magazines founded by industry insiders remained concentrated in the three biggest cities, while magazines founded by outsiders became geographically dispersed. Finally, we find that entrepreneurship evolved from the pursuit of a lone individual to a more organizationally-sponsored activity; this reflects the modernization of America during this time period. Our analysis demonstrates the importance of grounding studies of entrepreneurship in historical context. Our analysis of this “old†new media industry also offers hints about how the “new†new media industries are likely to evolve.

Author(s): Haveman, Heather A.; Habinek, Jacob; Goodman, Leo A. | Abstract: We trace the social positions of the men and women who found new enterprises from the earliest years of one industry’s history to a time when the industry was well established. Sociological theory suggests two opposing hypotheses. First, pioneering entrepreneurs are socially prominent individuals from fields adjacent to the new industry and later entrepreneurs are from an increasingly broad swath of society. Second, the earliest entrepreneurs come from the social periphery while later entrepreneurs include more industry insiders and members of the social elite. To test these hypotheses, we study the magazine industry in America over the first 120 years of its history, from 1741 to 1860. We find that magazine publishing was originally restricted to industry insiders, elite professionals, and the highly educated, but by the time the industry became well established, most founders came from outside publishing and more were of middling stature – mostly small-town doctors and clergy without college degrees. We also find that magazines founded by industry insiders remained concentrated in the three biggest cities, while magazines founded by outsiders became geographically dispersed. Finally, we find that entrepreneurship evolved from the pursuit of a lone individual to a more organizationally-sponsored activity; this reflects the modernization of America during this time period. Our analysis demonstrates the importance of grounding studies of entrepreneurship in historical context. Our analysis of this “old” new media industry also offers hints about how the “new” new media industries are likely to evolve.

How has access to the public sphere been affected by the rise of mass media? We address this question by studying magazines in America from the eighteenth century, when all periodicals had small circulations, to the mid-nineteenth century, when many reached mass audiences. Specifically, we investigate how the social positions of those who founded new magazines changed over this period. Previous research is divided on whether the rise of mass media made it more difficult for non-elites and industry outsiders to launch new magazines by creating large and powerful publishing houses, or made it easier by fostering acceptance of magazines as legitimate cultural products and improving access to resources needed for publishing. Using Goodman’s (1972) modification of multiple regression for the analysis of categorical data, we examine whether magazine founders were increasingly drawn from social elite and from inside publishing, or from an increasingly broad swath of society. We find that magazine publishing was originally restricted to industry insiders, elite professionals, and the highly educated, but after the rise of mass media, most founders came from outside publishing and more were of middling stature – mostly small-town doctors and clergy without college degrees. We also find that magazines founded by industry insiders remained concentrated in the major publishing centers, while magazines founded by outsiders became geographically dispersed.

Discussion of ``The William Kruskal Legacy: 1919--2005'' by Stephen E. Fienberg, Stephen M. Stigler and Judith M. Tanur [arXiv:0710.5063]

Consider an m-way cross-classification table (for m = 3, 4, …) of m dichotomous variables that describes (1) the 2m possible response patterns to a set of m questions (where the response to each question is binary), and (2) the number of individuals whose responses to the m questions can be described by a particular response pattern, for each of the 2m possible response patterns. Consider the situation where the data in the cross-classification table are analyzed using a particular latent class model having T latent classes (for T = 2, 3, …), and where this model fits the data well. With this latent class model, it is possible to estimate, for an individual who has a particular response pattern, what is the conditional probability that this individual is in a particular latent class, for each of the T latent classes. In this article, the following question is considered: For an individual who has a particular response pattern, can we use the corresponding estimated conditional probabilities to assign this individual to one of the T latent classes? Two different assignment procedures are considered here, and for each of these procedures, two different criteria are introduced to help assess when the assignment procedure is satisfactory and when it is not. In addition, we describe here the particular framework and context in which the two assignment procedures, and the two criteria, are considered. For illustrative purposes, the latent class analysis of a classic set of data, a four-way cross-classification of some survey data, obtained in a two-wave panel study, is discussed; and the two different criteria introduced herein are applied in this analysis to each of the two assignment procedures.

This article has two main purposes. The first main purpose is to take note of some of Dudley Duncan's major contributions as a quantitative sociologist, as the leader in introducing path analysis and related structural equation models to sociology, as a pioneer in applying loglinear models and related models in many different sociological contexts, and as the leader in introducing Rasch models to sociology. The second main purpose of this article is to show the reader something of Dudley Duncan, of the person and his character.

This essay describes in simple terms some of the major concepts of categorical data analysis (CDA) that have been and will continue to be useful in the analysis of sociological data, examples of which include data in the area of social stratification and mobility, and in many other areas that make use of survey data and/or panel studies data, and in empirical studies of latent types, latent variables, and latent structures. The exposition does not make use of any mathematical formulas, and the only arithmetic used is very simple multiplication, division, and addition. Simple numerical examples, constructed for expository purposes, are used as an aid in describing the concepts of categorical data analysis that are considered in the essay. These concepts include quasi-independence, quasi-symmetry, symmetric association, uniform association, and other related concepts useful in the analysis of mobility tables, and also other concepts that are useful in other areas of study.

Cet article commence par comparer le concept de quasi-symetrie du a Henri Caussinus avec un concept parent mais different que j'appellerai la quasi-symetrie de Karl Pearson; et nous trouverons que le concept de Caussinus est preferable et plus utile que celui de Pearson. J'introduirais alors un ensemble de modeles log-bilineaires quasi symetriques qui sont plus parcimonieux que le modele de quasi-symetrie de Caussinus, et un ensemble de modeles d'analyse des correspondances quasi symetriques qui sont plus parcimonieux que le modele correspondant de Pearson; et nous etablirons que les modeles log-bilineaires quasi symetriques sont preferables et plus utiles que leurs vis-a-vis, les modeles d'analyse des correspondances quasi symetriques. Notre attention est concentree dans cet article sur le modele de quasi-symetrie de Caussinus et sur les modeles qui lui sont apparentes. En addition aux modeles mentionnes ci-dessus, nous commenterons brievement les modeles de quasi-independance, et verrons comment les modeles de quasi-independance et de quasi-symetrie peuvent etre consideres comme des precurseurs directs des modeles log-lineaires et log-bilineaires. Finalement nous etudierons la relation entre quasi-independance et quasi-symetrie; et je citerai ici, pour les lecteurs qui pourraient etre interesses, une serie particuliere d'articles concernant la quasi-independance, publies pendant une periode de trente trois ans commencant en 1961.

The credibility of transportation engineers is essential to guide decisionmakers in making sound decisions about a constituency's mobility and safety. This article focuses on the area of technical inconsistency--why inconsistencies occur and what inconsistencies are the most common. The authors provide several examples of technical inconsistencies that have led transportation professionals to lose face and to have their recommendations questioned or ignored. The examples encompass a traffic impact study, traffic diversion, traffic safety, traffic calming, transit ridership estimates, and freeway improvements. The authors conclude with suggestions on how to avoid technical inconsistencies.

For the analysis of nonindependence between the row classification and column classification in an I x J contingency table, a general method is introduced (viz., generalized nonindependence analysis) that includes as special cases the usual correspondence analysis method and the two forms of association analysis (viz., unweighted and weighted association analysis). For each of these methods of analysis, both the saturated model and related unsaturated models are considered, and so are the related graphical displays. With respect to the graphical displays, we show, for example, that the usual correspondence analysis graphical display can be misleading, and that it should be replaced by a different graphical display. The advantages of using the graphical displays presented here rather than the usual correspondence analysis graphical display are discussed. AMS 1980 classification numbers: 62H17; 62H20; 62J10