Paul Felix Lazarsfeld Research Paper

As antisemitism was intensifying in Austria, Lazarsfeld arrived in the US in 1933 with a Rockefeller Foundation fellowship and decided to remain when it was extended for a second year. From the late 1930s, in a succession of positions at the universities of Newark, Princeton, and Columbia, and in research institutes and projects he led and/or founded, the Office of Radio Research and the Columbia Bureau of Applied Social Research, Lazarsfeld and his associates virtually created the field of mass communications research together with new methodologies and techniques in empirical social research (1940). In 1940, he was appointed to the faculty at Columbia University where he served until his retirement in 1969. At Columbia he initiated a program in political sociology marked by two pioneering longitudinal studies of the 1940 (The People’s Choice 1944) and the 1948 US presidential elections (Voting 1954), of opinion leadership (1955), and of the effects of McCarthyism on higher education (1959). Methodological innovations such as panel (longitudinal) analysis, snowball sampling, and contextual analysis were pioneered and developed in this research. During this time the Bureau of Applied Social Research became the premier academic social research institute in the US.

In the 1950s Lazarsfeld and his collaborators turned to the codification and systematization of these methods which resulted in many journal articles and edited volumes (Language of Social Research 1955; Continuities in the Language of Social Research 1972). Later still he focused on the uses and applications of social research (Introduction to Applied Sociology 1975). He was convinced that training, methodology, implementation, and utilization of social research in a professional manner were best carried in organizational settings such as those he had created at the Bureau rather than as solo pursuits by individual scholars, and that these organizations should be aided financially and administratively attached to universities (1962). He himself participated in creating such institutes elsewhere in Washington DC, Oslo, and Vienna.

Although known primarily as a methodologist, Lazarsfeld had a theoretical vision and program he called the empirical analysis of action (1955). It was inspired by Max Weber’s action theory, methodological individualism, and the operation of ‘ erstehen,’ framed in a realistic, empirical social psychology (Max Weber 1965). Its foundation was the choice studies of Lazarsfeld and his associates, on buying, voting, occupational choice, choice of friends, moving residence, and others (1959). Lazarsfeld’s studies in the history of social research found precedents for a sociological theory built from the ground up on empirical foundations rather than top down abstract philosophyzing (1961).

Throughout his life, he pursued other interests and lines of work, notably mathematical sociology (Mathematical Thinking in the Social Sciences 1954), scaling techniques, in particular latent structure analysis which he invented (Latent Structure Analysis 1967), and multivariate statistical analysis with attribute variates. He had a close professional relationship with Samuel Stouffer. At Columbia, he became in 1962 the first Quetelet Professor of Social Sciences. He had a unique 30-year working and professional relationship with his colleague the theorist Robert K. Merton (Merton 1998). Many leading sociologists in successive cohorts of students at Columbia were educated, trained by, and collaborated with these two outstanding scholars. Lazarsfeld’s personal style of doing social science was to recruit capable students and colleagues to work with him and to co-author books and articles. During his career, he received many honors, including honorary degrees from the universi- ties of Columbia, Chicago, Vienna, and the Sorbonne. He was president of the American Sociological Association and the American Association for Public Opinion Research. He was a member of the National Academy of Sciences and the National Academy of Education. He was an institution builder who forged transatlantic ties by inviting European scholars to Columbia, and who served frequently as a visiting professor in Europe, helping create research institutes and programs in France, Poland, Norway, and Austria.

1. Lazarsfeld’s Methodology

He distinguished philosophizing about social knowledge and how-to-do-it techniques from methodology, which is a systematic analysis and codification of what social scientists were doing on data collection, measurement, and drawing inferences from findings, i.e., explication and critique of the work of social scientists. Methodology was a creative act: ‘intuition reconsidered in tranquility.’ He identified four steps in measurement:

(a) imagery, or the sifting through intuition and disparate findings for an articulate idea. The goal was ‘depth of conceptualization’ (1959);

(b) concept specification and dimensionalizing;

(c) selection of indicators for each dimension; and

(d) the combination of indicators into an overall measure or index.

Indicators have a probabilistic relationship to an underlying dimension. With several indicators, however, the probability of describing a respondent correctly increases; thus, multiple indicators should be used (1958). Though the choice of indicators from a large pool is somewhat arbitrary, and though some respondents’ answers will change depending on specific indicators, findings in the aggregate will remain the same as will the correlation between variables. This principle of interchangeability of indices removed an important challenge to internal validity in survey research.

Other advances in method included the ‘accounting scheme’ and ‘reason analysis’ for getting at an assessment of why people make certain choices by measuring dispositions and attributes, exposure to various influences, and the respondents assessment of their importance (Analysis of Reasons 1942).

In Lazarsfeld’s approach to survey research much attention was expended on getting these four steps right and on creating a questionnaire that captured the process accurately. During the research that led to The Academic Mind (1958), the questionnaire went though eight versions over six months as the central ‘apprehension’ variable was sorted into a ‘worry’ and a ‘caution’ dimension in pretests, and indicators and wording appropriate for a faculty population ranging from small religious colleges to major research universities were carefully crafted. Lazarsfeld liked to end up with a ‘pivotal’ variable that he said was ‘less than a theory but more than a label for just one or two measurements’ (1959). It organized a range of findings into a key concept or typology that provided deep insight into the topic and linked the research to other social science research. In Jahoda et al. (1971), the pivotal variable was apathy or resignation.

Although identified with quantitative techniques, Lazarsfeld favored combining information from a variety of sources. In Jahoda et al. (1971), the researchers made use of participant observation, census statistics, election results, library circulation records, newspaper circulation, school essays, family records, life histories, psychological tests, expert reports, even measurements of the walking speed of villagers. Mindful of critics who argued that survey research forced an individualist straightjacket on sociological topics, Lazarsfeld created contextual analysis. Groups, social units, organizations, and social milieus were characterized by global, relational, and aggregated variables called collective properties. These enabled hypotheses to be researched on social influences, climates of opinion, and other sociological processes (1961, Continuities in the Language of Social Research, 1972). Other methodological innovations that came out of the Lazarsfeld–Columbia tradition were deviant case analysis (Language of Social Research 1955), concept of property space and substruction for systematizing typologies, and snowball sampling (Personal Influence 1955).

2. Data Analysis And Statistical Inference

Lazarsfeld presented findings in a simple, understandable form to a wide public: two and three variable contingency tables and bar charts. He favored data collected in natural settings to controlled experiments carried out in a laboratory. Discrepancies between findings in survey research and in controlled experiments in communications research he attributed to selection effects, i.e., in natural settings people exposed themselves selectively to information and influences they tended to agree with. To compensate for the lack of control groups in natural settings, he developed a multivariate analysis derived from the English mathematical statistician G. Udny Yule, a pioneer in the analysis of covariance for categorical data (Continuities in Social Research 1950). The complex effect of three variables on one another is decomposed into a weighted sum of two variable effects. Starting with a hypothetical two variable relationship, a third ‘test’ variable is introduced to check whether the original relationship is spurious, contingent on the level of the test variable, or exhibits other patterns called specification and interpretation (Language of Social Research 1955, Sect. 2, pp. 111–66). Survey analysis is the art of probing hypothesis with test variables for ascertaining the conditions, contexts and domains, and extent to which they are true. In the 1940s and 1950s, computers were not available and statistical operations were performed on IBM countersorters using IBM data cards. For quantitative variables, desk calculators took much time and concentration to grind out even the simplest correlations. Thus, survey analysis seldom went beyond the simultaneous study of three variables. Although analysis of findings for causal patterns is nowadays done with multivariate regression techniques with statistical packages in computers in a matter of minutes, the logic of explanation in quasi-experimental design has remained the same.

In the Lazarsfeld’s quasi-experimental designs, the time order of the variables was crucial because it ruled out some patterns of causation. Repeated observation on the same units, called ‘panel’ analysis and known today as longitudinal sample design pinned down a specific time order and became central in major Lazarsfeld studies. The 1940 election campaign study (People’s Choice 1944) was based on five waves of interviews with the same respondents over several months. The method allowed a deep understanding of how various groups and types of voters formed political opinions and choices as a result of campaign, media and personal influences. Lazarsfeld (1955) realized that the study of long term effects, e.g., the cumulative impact of television viewing on children’s attitudes and behavior, was beyond both the experimental and quasi-experimental methods then current.

3. Mass Communications And Political Sociology

Lazarsfeld and associates undertook two major election campaign studies, in 1940 (People’s Choice 1944) and in 1948 (Voting 1954), both using a longitudinal design of repeated interviews with potential voters over several months in typical towns, and a study of opinion leadership among 800 women about marketing, fashion, movies, and public affairs (Personal Influence 1955) which used snowball sampling enabling influentials repeatedly named by the initial sample to be interviewed in turn. The findings challenged the conventional wisdom on democratic theory and on the mass media.

Political opinions and voting are only in part based on informed, calculated interest reflecting the arguments of parties, platforms, and candidates. Voting and opinion are social validations of group membership by conformity to social milieus and family tradition. Campaigns reinforce existing views rather than change them, reactivate those who were disengaged from politics, and foster turnout and participation. Much of the public have their minds made up early in the campaign. The least informed and those under cross-pressures from multiple group memberships delay making choices, are the most likely nonvoters, and are the most likely to switch their vote between parties. Lazarsfeld’s discoveries challenged the common notion that democracy would only work with an informed and enlightened public.

The impact of the mass media on voting is modest. Voters expose themselves selectively to views they tend to agree with. Moreover, in what he called the twostep flow of communication, people seek political guidance from opinion leaders in proximate social milieus—at work, in neighborhoods, in families—who are a transmission belt for public affairs and attitudes from the political class to the citizenry. Interpersonal communication is flexible and casual, from a trusted source, which makes it more effective than mass communications. Opinion leaders are found in every group and social stratum, and reinforce the dominant views of these social milieus. Thus, there is no large political influence from elites and top strata down to the lower strata. The impact of the mass media is limited and interpersonal communication has a conserving and differentiating effect on political opinion, attitudes, and choices. These findings challenged the views of many intellectuals, culture critics, and mass society theorists.

Lazarsfeld was open to criticism and was his own critic as well. On political sociology, he knew that he had neglected the institutional side of politics—political parties as organizations, platforms, choice of agendas. In mass communications, he regretted not studying the ‘inside of the communications industry.’ There were also missed opportunities: the pioneering campaign studies were not followed with studies of national campaigns that the University of Michigan’s Survey Research Center undertook. And when television became the dominant mass medium, Lazarsfeld and the Bureau failed to develop a research program that was both acceptable to and supportable by the foundations and the networks.

4. Lazarsfeld’s Influence

His greatest influence and lasting legacy is on social science methodology, on survey analysis, mass communications theory and research, and political sociology and political science, in which he created entirely new fields of scholarship, as has been described above. Public opinion polling and research has diffused worldwide. In a broad sense his research highlighted the importance of small groups and proximate social milieus, and personal influences on attitudes, opinions, and behaviors which sheltered people from macrosociological forces. Much of what Lazarsfeld accomplished has been incorporated thoroughly into social science and is not even referred to or footnoted. This is true not just in academic disciplines and university settings but in journalism, business schools, communications departments, market research, policy studies, and practitioners and professionals in related fields. Though he thought of himself as a marginal man in academia, both in Vienna and in the US, which he overcame with an entrepreneurial style of action combining intellectual and administrative innovation, he became internationally influential and recognized. His legacy and thinking live on in the pursuits and accomplishments of his colleagues, collaborators, associates, and students. At one time or another, and sometimes over many years, he was working with Robert Merton, Marie Jahoda, Hans Zeisel, Bernard Berelson, Patricia Kendall, Samuel Stouffer, Alan Barton, Elihu Katz, Ernest Nagel, Mira Komarovsky, Robert Lynd, C. Wright Mills, Arthur Kornhauser, Theodor Adorno, Morris Rosenberg, William McPhee, Frank Stanton, David Sills, Edward Suchman, Herbert Menzel, Leo Lowenthal, Joseph Klapper, Marjorie Fisk, Neil Henry, Sam Sieber, Wagner Thielens, Jr., Seymour Lipset, James Coleman, Juan Linz, Herbert Hyman, Raymond Boudon, Bernard Lecuyer, Anthony Oberschall, and others. He was a man of broad learning and enormous energy who, in addition to social science was well-versed in history, the humanities, and the philosophy of science.

Bibliography:

1. Berelson B, Lazarsfeld P F, McPhee W 1954 Voting: A Study of Opinion Formation in a Presidential Campaign. Chicago University Press, Chicago
2. Jahoda M, Zeisel H, Lazarsfeld P F 1971 Die Arbeitslosen on Marienthal. Hirzel, Leipzig, 1933, Frankfurt, 1961 (new introduction by Lazarsfeld P English trans. Marienthal: The Sociography of an Unemployed Community). Aldine, Chicago
3. Katz E, Lazarsfeld P F 1955 Personal Influence: The Part Played by People in the Flow of Mass Communications. Free Press, Glencoe, IL
4. Lautman J, Lecuyer B P (eds.) 1998 Paul Lazarsfeld (1901–1976). L’Harmattan, Paris
5. Lazarsfeld P F 1940 Radio and the Printed Page. Arno Press, New York
6. Lazarsfeld P F 1942 Statistical analysis of reasons as a research operation. Sociometry 5: 29–47
7. Lazarsfeld P F 1955 Why is so little known about the effects of television on children and what can be done. Public Opinion Quarterly 19: 243–51
8. Lazarsfeld P F 1958 Evidence and inference in social research. Daedalus 87(4): 99–130
9. Lazarsfeld P F 1959 Methodological problems in empirical social research. Proceedings of the Fourth World Congress of Sociology. International Sociological Association, Louvain, Belgium, Vol. 2
10. Lazarsfeld P F 1959 Reflections on business. American Journal of Sociology 65(July): 1–31
11. Lazarsfeld P F 1961 Notes on the history of quantification in sociology—trends, sources and problems. ISIS 52(2): 147–203
12. Lazarsfeld P F 1962 The sociology of empirical social research. American Sociological Review 27(6): 757–768
13. Lazarsfeld P F 1970 Sociology Main Trends of Research in the Social and Human Sciences. UNESCO, Paris, Chap. 1, 61–165
14. Lazarsfeld P F, Barton A 1951 Qualitative measurement in the social sciences: classifications, typologies and indices. In: Lerner D, Lasswell H D (eds.) The Policy Sciences. Stanford University Press, Stanford, CA, pp. 155–92
15. Lazarsfeld P F, Berelson B, Gaudet H 1944 The People’s Choice. Columbia University Press, New York
16. Lazarsfeld P F, Fleming D, Bailyn B (eds.) 1969 An episode in the history of social research: A memoir. In The Intellectual Migration. Cambridge University Press, Cambridge, MA, Chap. 6
17. Lazarsfeld P F, Henry N 1968 Latent Structure Analysis. Houghton Mifflin, New York
18. Lazarsfeld P F, Menzel H 1961 On the relation between individual and collective properties. In: Etzioni A (ed.) Complex Organizations: A Sociological Reader. Henry Holt, New York, pp. 424–40
19. Lazarsfeld P F, Merton R 1950 Continuities in Social Research. Free Press, Glencoe, IL
20. Lazarsfeld P F, Oberschall A 1965 Max Weber and empirical social research. American Sociological Review 30(Spring): 185–99
21. Lazarsfeld P F, Pasanella A, Rosenberg M 1972 Continuities in the Language of Social Research. Free Press, New York
22. Lazarsfeld P F, Reitz J, Pasanella A 1975 An Introduction to Applied Sociology. Elsevier, New York
23. Lazarsfeld P F, Rosenberg M 1954 Mathematical Thinking in the Social Sciences. Free Press, Glencoe, IL
24. Lazarsfeld P F, Rosenberg M 1955 The Language of Social Research. Free Press, Glencoe, IL
25. Lazarsfeld P F, Sills D 1979 In International Encyclopedia of the Social Sciences. Free Press, New York, vol.18
26. Lazarsfeld P F, Thielens W Jr. 1958 The Academic Mind. Free Press, New York
27. Merton R K, Coleman J, Rossi P (eds.) 1979 Qualitative and Quantitative Social Research, Papers in Honor of Paul F. Lazarsfeld. Free Press, New York
28. Merton R K 1998 Working with Lazarsfeld: notes and contexts. In: Lautman J, Lecuyer B P (eds.) Paul Lazarsfeld (1901–1976). L’Harmattan, Paris