Myths of the Chicago School

4. The quantitative tradition at Chicago

4.4 Ogburn and the nurturing of quantitative techniques

Even if Park was opposed to, or at least sceptical of, statistical approaches other Chicago sociologists, notably Ogburn, were not ignoring statistical developments during the 1920s and 1930s.

Ogburn was doing empirical work, which is not what we would call experimental work today, he was just manipulating countings that had already been made for other purposes by the government. But he was squeezing theoretical material out of the census and other government methods of counting. (Dollard, 1972)

Ogburn represented a counterweight to the anti-statistical orientation Park represented and set the tone in the department for quite a while of being anti-statistician. Yet Faris insisted that it wasn't a good idea to have a sociologist falling behind because he didn't have good statistical training, so they brought Ogburn on and Ogburn made quite a splash and attracted some very good students. (Cottrell, 1972)

The development of quantification in sociology at Chicago received a boost from the employment of Ogburn who moved from Columbia with a well-established reputation. Indeed, Ogburn contributed a chapter on 'Statistical Studies on Marriage and the Family' to the text that resulted from the work of the American Statistical Association's Committee on Social Statistics (Rice, 1930) and had been editor of the Journal of the American Statistical Association. He turned down the chairmanships of sociology departments at Michigan and Minnesota (Ogburn 1929b) in order to come to Chicago where he took responsibility for the development of quantitative techniques and of quantitatively based research in the department. Nonetheless, the development of a statistical perspective, he recalled, was not easy.

On coming to the University of Chicago I found a much more hostile attitude toward statistics than I ever had at Columbia. Yet I fought the battle, taught all the statistics in the Sociology Department, and participated generally in the statistical work of University Committees. (Ogburn journal, 13th June 1952)

The arrival of Ogburn, while clearly a boost for quantification in Chicago sociology, was not simply the arrival of a statistician. Ogburn had rejected the idea of restricting his intellectual development by being nothing but a statistician as early as 1912 (Ogburn journal, 13th June 1952). With the depression Ogburn spent more time on substantive sociology and became particularly interested in technology and social change. Eventually he lost track of developments in statistical theory and gave up teaching the subject altogether. Some commentators, have suggested that Ogburn's development of statistics represented a split in the department. As Ogburn's influence began to be felt towards the end of the 'golden era'

statistics gained favour... and life histories began to fade as a primary means of study, to be replaced by the collection of comparable data from a number of cases that were subjected to statistical analysis.... Ogburn's arrival gave impetus to the quantitative analysis of social data and thereby contributed to the decline of case studies. (Cavan, 1983, p. 414)

 
Cottrell (1972) also suggests that the impact of Ogburn was to lead to a methodological split with him developing his own group of quantitatively oriented students. Such a view, which implicitly prevails as part of many designations of the 'Chicago School', gives credence to the view of the school as retaining 'anachronistic ethnographic techniques' for 'unscientific reformist' purposes (see chapter seven). However, such views are misleading. Small had wanted a statistician in the department for more than a decade.

Ogburn was recruited because sociology did not have a course in statistics... the realisation came up that it was an important subject and the realisation was there that Chicago ought to bring in some outsiders.... Columbia was the other strong department in the country... probably my father heard it from somewhere that Ogburn was dissatisfied at Columbia and it might be a good chance to raid... Ogburn really wanted to stay at Columbia but wasn't really decided. There was feeling against him there.... It was my father's initiative that got Ogburn here... he probably talked the others into it. (Faris, 1972)

Blumer (1972) thought that Burgess had pushed for Ogburn's appointment more than anyone else but that it had been a consensus of the Department that such an appointment was necessary as their was a gap to be filled. Ogburn moved rapidly to promote a more positive image of quantification in sociology, but was not acting either to reverse a trend, or in isolation. If Park and Blumer had reservations, in the main Ogburn's arrival was seen as advantageous for sociology at Chicago. The Society for Social Research, through its Bulletin, welcomed Ogburn and advertised and promoted his courses.
 
Ogburn's impact was substantial and went beyond the mere importation of statistical expertise. His involvement in the department and outside was notable and he tended to adopt wide horizons. 'His concepts were primarily national whereas the others were local' (Dollard, 1972). Ogburn was directly involved in ongoing research in the department as well as his own heavy commitments that included the President's Committee on Social Trends. He wrote to Ruth Newcomb on October 17th 1929:

I am working this Fall on the planning of a survey of recent social changes to be conducted under the auspices of President Hoover. I'm just finishing up the study on the comparative strength of the various forces operating in the Presidential election of 1928. I'm also hoping to finish up a study on the business cycles and politics. I have also worked during the past month of the quarter and during two and one half months of the Summer quarter on outlining the report for a plan for a nationwide cost of living study with particular reference to its scope and method. This was done for President Hoover at the request of Secretary Wilbur. (Ogburn, 1929)

One avenue of his involvement in research at Chicago was through the Local Community Research Committee, which he joined as soon as he settled in. Ironically, in view of its general tendency, the Local Community Research Committee actually provided a focus for the development of quantitative research techniques. In the period from his appointment in 1927, to the demise of the Committee in 1929, Ogburn had begun three studies under the auspices of the Committee and two of the more quantitative students, Stephan and Tibbits, were involved in two other projects. The Committee was also responsible for establishing research professorships in the social sciences, two of whom, Thurstone and Schultz, were predominantly involved in quantitative research. The Local Community Research Committee was a major force in the development of quantitative research methods at Chicago (Bulmer, 1981a, 1984). The committee gave considerable financial support to quantitative research in the social sciences at Chicago. The funding of large-scale politics surveys, the purchase of census tract data, the appointment of support staff (crucial for the time consuming process of attempting factor analytic and multivariate techniques by hand), and the provision of physical facilities for research, notably computational machines and the space to house them, were important contributions of the Local Community Research Committee
 
This involvement of the Local Community Research Committee in the development of quantitative techniques towards the end of the 1920s is reflected in White's (1929a, p.25) summary. He noted that the social sciences have not ordinarily been thought of as using or needing laboratory equipment and that

social scientists regularly treat the community as a clinic, diagnosing on the basis of existing knowledge and insight, and prescribing with what wisdom they may possess for social ills.
In a more precise sense, however, the social sciences have now reached the point where it is open to use laboratory methods. Mr. Gosnell's experiment with a section of the Chicago electorate, applying a known stimulus under controlled conditions, reveals the social scientist at work in an out-of-door laboratory; the various analyses of personality, including the application of the technique of the psychologist and psychiatrist, involve laboratory technique and equipment provided in the new Social Science Building.
It may be more accurate to refer to this building…as a workshop. (White, 1929a, p. 31)

A further indicator of Ogburn's involvement in the inter-disciplinary nature of the development of quantitative techniques at Chicago can be seen in his collaboration with the economics department in reconstructing their written quantitative methods paper for the doctorate. What had once been a combined accountancy and statistics paper was restructured into two self contained papers with the statistics element greatly enhanced and including time series analysis, index number construction, multiple and partial correlation and probability distributions.
 
Ogburn was also quick in requesting additional full-time support in the quantitative area and circulated the department with a draft memorandum (c. 1928) entitled 'Preliminary Draft of the Statistics Proposal'

It is highly important that a man in statistical mathematics be added to the mathematics department, this is the first wish of all of us. If, however, it is not possible to arrange for this a man should be added to the economics business staff who is very strong on the mathematics side. There should be centralised operations of all elementary and intermediate statistics work in economics, sociology, psychology, commerce and administration and social service administration. The really advanced work should be given in a new building in connection with research projects. (Ogburn, 1928)

In the event, the Social Science Research Building was erected for such advanced social scientific work in 1929 and contained an advanced statistics laboratory. Ogburn, who had responsibility for the 'image' of the Social Science Research Building, managed to persuade his colleagues that a suitable motto would be an annotated quote from Lord Kelvin, 'When you cannot measure * your knowledge is * meagre * and * unsatisfactory *', which is engraved on the outside of the building. Ogburn's interdisciplinary perspective did not, however, have priority over his sociological concerns. Apart from taking on R. E. L. Faris as a statistics assistant in October 1929, he developed various new courses in the department. Ogburn's arrival, having been announced and welcomed in the Bulletin of the Society for Social Research in 1927, was followed up in January with an article that noted that

The University has announced a new course in methods of research for the first term of the Summer Quarter, 1928, which it is hoped will be of interest to members of the Society for Social Research, and to others who expect to attend the sessions of the Institute next August. (Bulletin of the Society for Social Research, January, 1928)

The course, which was taught by Ogburn, was aimed at graduate students who had some experience in, and who were carrying on, independent research work and who wanted to meet others engaged in the same or related lines of enquiry. The course consisted of lectures, demonstrations and a twice-weekly clinic, plus individual counselling from instructors. The course had been prompted by

The steady accumulation in recent years of maps, statistics and local studies at the Social Research Laboratory of the University; the interesting investigations now in progress at the Institute for Juvenile Research... and the increasing number of social investigations carried on by other local institutions.
The purpose of the new course in methods of research is to extend and complete the program of the Institute of the Society for Social Research; to make it a center where students may meet; a clearing house where methods of research may be compared and criticized. (Bulletin of the Society for Social Research, January, 1928)

The April 1928 issue of the Bulletin of the Society for Social Research called attention once again to the new course in 'Methods of research in the social sciences'. This was but one of several new courses in quantitative methodology that were to blossom at Chicago. In 1927 Ogburn offered an introductory course in statistics, which he taught until 1929. Stouffer, one of Ogburn's graduate students, took over the teaching in 1930. When, in 1932, Rice spent a year at Chicago, the course title was changed to 'Statistical sociology'. Ogburn continued teaching the new course until 1934. During his year at Chicago, Rice also ran a course entitled 'Measurement in social politics'. Running parallel to the introductory course was a course entitled 'Statistical methods' ('Methods of quantitative sociology', after 1932) which either Ogburn or Stouffer taught from 1927 to 1945 (with the exception of 1931, 1943).
 
Ogburn was supposedly to provide a course on the 'Statistics of social maladjustment' when he first arrived at Chicago but this was never given, although he did teach a course on 'Sociology and the social sciences' in which he demanded a more 'exact' approach to social scientific data and enquiry. In 1929, he began an advanced course in 'Research in quantitative sociology' that was available most years until 1940 and, in 1936, Ogburn gave a course on 'Partial correlation analysis', which he repeated each year he was available until 1945.
 
In 1931 Ogburn first offered a course on 'Statistical problems', which he taught between 1932 and 1935, before it became a more specific course in the 'Measurement of relationships' in 1937 and was taught until 1942. Stouffer taught this course in 1938 and had earlier taken the 'Statistical sociology' course in 1935. These were but a part of Stouffer's prolific involvement in teaching quantitative methodology courses in the mid-to-late 1930s. In 1935 he started a courses on 'Sampling in social research' (which was later taught by Williams in 1948 and 1949), 'Quantitative studies in the family', 'Quantitative criminology', as well as a course on the 'Applications of probability to sociology' which changed to 'Quantitative aspects of social problems' the following year. In 1938 he inaugurated 'Quantitative studies in social psychology' and in 1939 he took over and revamped Park's old course in 'Human migrations' before, a year later, focussing upon 'Quantitative studies in social organisation' and 'Statistical problems of governmental research'. Teaching developments in this field, however, were thwarted by the Second World War and Stouffer's involvement in government research culminating in his leave of absence for government service in 1944 and 1945.
 
The courses Stouffer began in 1936 on 'Quantitative problems in population', 'Quantitative studies in population and human ecology' and 'Dynamics of population' initiated a series of courses on population in which Hauser and Duncan were heavily involved. In 1948 Hauser taught 'Comparative population structure and dynamics', 'Quantitative methods for population research' and 'Seminar in the analysis of census data' as well as the introductory course to the study of population human ecology. Hauser had already offered a course on statistical sources in 1947 and followed it up with one specifically on 'Sample surveys as a research method' in 1949. In 1950 he took over a course on the 'Design of research', which Goodman had started in 1947.

Other sociologists who took on the teaching of quantitative methodology courses [4] included Williams, Kittagawa and Hart. Williams taught 'Methods of quantitative sociology' in 1947, 'Introduction to statistical reasoning', and 'Mathematics essential to elementary statistics' in 1948. Kittagawa also taught these courses during the next three years. In 1948 the more advanced statistics courses were 'rationalised' into three successive courses 'Statistical methods of research' parts I to III, and taught initially by Williams and Kittagawa before becoming the province of Goodman. The latter also developed a course on statistical inference (in 1951) and on 'Recent advanced methods' in quantitative research a year later. In 1949, Lazarsfeld was a visiting lecturer from Columbia and offered a course on 'New developments in attitude measurement' and a year later Hart (director of the National Opinion Research Centre) gave a course on 'Research in public opinion'. (Appendix 4 provides full details of courses offered in the Sociology Department of the University of Chicago).

Notes

[ 4] Ogburn gave up teaching statistics some time before his retirement to concentrate on his substantive interests in sociology. In his journal he reflected on whether he had made the correct decision.

 Last evening, Rubyn, Harriet Welch and I had dinner at the Quadrangle Club. Nearby at a table for four sat R.A. Fisher, famed statistician with Allen Wallis and Thurstone, two U. of C. statisticians. Allen Wallis is editor of the American Statistical Journal and Executive Secretary of the U. of C. Committee on Statistics. Thurstone has developed several important techniques and published several books on statistical methods. At lunch I saw a group of some 20 persons in the private dining room at a table where R.A. Fisher sat. I judge we gave him an honorary degree. I had no part in any of this. I was not invited or consulted. It was not because I was an emeritus. Thurstone become an emeritus this July. I was away nearly all year. But that does not explain my not being in on this statistical gathering. The pang of not being invited is a feeling I have seldom experienced.

But twenty-five years ago, more or less, I was editor of the American Journal of the Statistical Association and also president of the American Statistical Association. Then, I think the nearest to an academic ambition I ever had was to be a social statistician, which is the most exact of the scientific activities of a scientist in the social field. Yet about 1912 or thereabouts, I recall definitely rejecting the idea of being nothing but a scientist and thus of restricting my intellectual, interests .... But then came the depression of the '30's, the War and there was competition for my interests. I became more and more interested in the significance of technological change for society. I spent lots of time on Recent Social Trends, a big undertaking. Then I gave up teaching statistics. But I envy those who stayed by statistics, and sometimes I think I wish I had.

Clearly this envy and this regret are strong emotions. But I wonder how much rationality there is to making this emotion the measure of my values, or the criterion of my action. My worship of statistics has a somewhat religious nature. If I wanted to worship, to be loyal, to be devoted, then statistics was the answer for me, my God. But a God only meets an emotional need, which has little to do with reason. I wonder would I have been content to have been only a very good statistician; and to be a good one, all one's effort and attention is needed. I doubt it. My work in technology and social change and social evolution give me much intellectual pleasure and many thrills. Yet I regret keenly that the march of statistics has passed me by. There was a vacant place at R. A. Fisher's table. (Ogburn Journal, 14th June, 1952). Return

 

Next 4.5 Burgess as the barometer of methodological tendencies