Continuing our series on boundary-crossing science, professor Douglas Massey, former president of the American Sociological Association and current chair of the sociology department at the University of Pennsylvania, writes about the challenges of combining micro and macro. His research on how segregation affects the lives of black and Hispanic Americans-best known in his 1993 volume American Apartheid-has continued with his current work on the experiences of minority students at elite colleges and universities, documented in The Source of the River (2003).
Susan T. FiskeAPS President
Although a card-carrying sociologist, I actually started out in psychology. Indeed, as an undergraduate I completed all the courses required for a psychology major before I ever took my first course in sociology; and if the truth be known, I took very few of them. One of the first sociology courses I did take was social psychology. I had already taken all the relevant social psych courses in the psychology department, and I was curious to see what passed for social psychology within the field of sociology.
What I found was rather a shock. In psychology I had learned the importance of the well-designed experiment grounded in theory, and the content of the psych courses I took consisted primarily of results from cleverly planned and well-executed experiments conducted inside and outside of the lab: the usual portfolio of work by Zimbardo, Darley, Asch, Milgram, and others. In the sociology course I took, however, it was as if none of these people existed. Instead we learned about Mead, Cooley, Goffman, and Humphries and none of these luminaries did experiments. Indeed, in class I learned that experiments were “positivist” and thus somehow suspicious in epistemological if not political terms.
Although some of the readings were interesting (I liked Goffman) and the lectures were often quite funny and insightful, I didn’t see where the science was. The explanations all seemed to be “stories” spun to account for social behavior without really being linked to a broader theory, and they were certainly not “proved” according to the standards I had by then assimilated through years of pre-med and psychology training. I thus fled sociology to become a demographer, only returning to the discipline of sociology in later years.
When I did return to sociology, I was pleased to discover that there was an experimental tradition in sociological social psychology after all, centered at Stanford University and a few other outposts. I also learned that considerable sociological work in social psychology was being done in conjunction with survey research, notably at Michigan’s Institute for Social Research. Indeed, the American Sociological Association currently publishes the journal Social Psychology Quarterly, edited by experimental sociologist Cecilia Ridgeway, Stanford University, and including on its board other experimentalists such as Karen Cook, Stanford University, as well as survey researchers Lawrence Bobo, Harvard University; Merilee Taylor, Pennsylvania State University; and David Williams, University of Michigan.
Although there is apparently more overlap between the psychological and sociological domains of social psychology now than in times past, it still seems to be insufficient and rather one-sided. Although most of the sociological social psychologists of whom I have spoken seem to follow the psychology literature quite closely, my admittedly cursory survey of psychological social psychologists suggests the reverse is less true. Such an asymmetry is perhaps unsurprising given sociology’s “soft” legacy in the field and the huge difference in the size between the two disciplines. Nonetheless there is room for improvement on both sides.
I believe that social psychology should concern itself with the interface between individual social expression and broader social structures. That expression may be behavioral or cognitive and may or may not occur in the presence of other actors; but the social contexts in which people grow up and learn, and within which they ultimately come to act and interact, are inevitably embedded within broader social structures. As these structures change over time, so will the specific social contexts within which people exist on a day-to-day basis and, hence, so will social learning, cognition, and behavior. A comprehensive social psychology must simultaneously work from the top down, studying how larger structures create specific social settings, as well as from the bottom up, studying how individual social cognition and behavior interact to generate specific settings.
It’s easy, of course, to stand loftily above the gritty realities of disciplinary life and make noble pronouncements about what one’s colleagues should be doing. Recently, however, I had the opportunity to put my views into practice in helping to organize and direct the National Longitudinal Survey of Freshmen with my colleague Camille Charles. This project was initiated with support from the Andrew W. Mellon Foundation to provide data on minority under-achievement in higher education. In the Fall of 1999, we surveyed the cohort of freshmen entering 28 selective colleges and universities around the country and gathered equal-sized samples of whites, Asians, Latinos, and African Americans (roughly 1,000 in each group). The baseline survey asked a detailed set of questions to determine respondents’ respective social, economic, and demographic origins. We have been surveying them every year since then to chart their progress as they move through higher education, or not (the fourth-year survey, when most respondents are seniors, is just going into the field).
In studying minority under-achievement, of course, one needs to measure and account for all kinds of background characteristics that tied to one’s position in the social structure, not the least of which is access to educational resources. But even a cursory review of the literature reveals that measuring structural variables is not enough: subtle and complex psychological processes also enter into the process of minority achievement, notably through the mechanism of stereotype threat hypothesized by Claude Steele and colleagues. In designing the NLSF, we deliberately sought to assess individual differences in the susceptibility to stereotype threat, attempting to measure the degree to which minority respondents subscribed to negative stereotypes about their group (internalization) and the degree to which they expected whites to use stereotypes when evaluating them (externalization).
Our results indicate that minority academic achievement is both sociological and psychological. It is sociological because a student’s position in the social structure in particular, whether he or she grew up under conditions of integration or segregation acts strongly to determine the quality of education received and the degree of childhood exposure to violence and disorder, both of which have strong effects on grades earned in college. However, achievement is also psychological because minority students who had internalized negative group stereotypes put in significantly few hours of study than those who had not, indicating disidentification with college achievement, and those with higher levels of externalization, expecting to be judged stereotypically, performed significantly worse in terms of grades. Both the “sociological” and “psychological” effects proved to be statistically significant while holding the other constant.
The simultaneous measurement of psychological and sociological determinants of academic achievement also produced another benefit: we were able to study the social structural roots of stereotype threat. Specifically, we found that having a large number of same-race friends in childhood and well-developed racial identity reduced a person’s internalization of negative stereotypes, and that greater childhood segregation and stronger racial identity decreased the degree of externalization. These early results from the NLSF only confirm my view that sociology and psychology stand to profit handsomely from a closer integration. The explanations and mechanisms hypothesized by each discipline are rarely mutually exclusive. On the contrary, they usually shed light on one another. When it comes to explaining social behavior, psychology without sociology is seriously incomplete and vice versa.