why we could use more experiments

One thing that organizational and economic sociology could use more of is experimental methods. While sociologists are not completely averse to experiments (see its prominent use in exchange theory), the method seems to occupy a small niche. Some sociologists express a real distaste for experiments. Our love of context and history seems to bias us against experiments, which emphasize internal validity over external validity and random assignment over sampling from real populations.

My sense though is that a number of theoretical areas could be more fully developed by using experiments. The real value of experiments comes from being able to more precisely identify theoretical mechanisms, especially at the cognitive level. (If you have any doubt of the utility of experiments, check out Correll’s, Benard’s and Paik’s beautiful study of the motherhood penalty.) Given the calls to explore the micro/cognitive foundations of social theories, experiments could be very useful. Here are just a few conceptual areas that could benefit from experiments.

  • Networks and relationship formation – what cognitive dynamics explain homophily? How does framing affect relationships (see, for example, this paper in Psych Science). What sorts of social cues trigger relationship formation? What is the role of emotion in choosing friends?
  • Institutions and cultural persistence – Zucker (1977) broke ground in this area but since then experimental methods have been scantly used. What cognitive dynamics explain habituation? What role does social influence play in the transfer of cultural preferences? What situational dynamics lead to rule conformity?
  • Collective action frames – why are some frames more resonant than others? How important is shared identity to frame resonance?
  • Categories and legitimacy – to what extent does categorical contrast lead to perceptions about legitimacy? How different does something have to be from others in a category before individuals perceive a fit problem? What is the relationship between categorical fit and valuation?
  • Status and power – why are individuals so biased by status? How sensitive are individuals to status differences? What are the cognitive dimensions of status deference?

What else would you add to the list?

Written by brayden king

February 25, 2011 at 10:19 pm

4 Responses

Subscribe to comments with RSS.

  1. Much of what is usually called “Behavioral Economics” could easily qualify as experimental sociology. See for example Fehr and Gintis’ (2007) ARS paper on cooperation (



    February 26, 2011 at 1:10 am

  2. I think there’s a lot of great work starting to look (experimentally) at the microfoundations of inequality in organizations and labour markets.

    Bertrand and Mullainathan’s field experiment sending out resumes (“Emily and Greg vs. Lakisha and Jamal”) comes to mind, as does Hannah Riley Bowles’ work on gender and employment negotiations (“Sometimes it does hurt to ask”, with Linda Babcock).



    February 26, 2011 at 6:36 am

  3. Absolutely agreed. I am actually presently planning an experiment that is pretty close to that first bullet point (homophily dynamics). There is already a pretty large amount of experimental work on the 3rd and 5th bullet points in social psychology and in political science, though. This work involves a somewhat different definition of “frames” from the social movements research, but the cognitive dynamics it explores are largely the same ones that would be relevant to frame resonance. (For example, there is lots of really interesting work relating framing to identity — e.g., Cohen, Geoffrey, 2003. “Party over Policy”, JPSP; Druckman, 2001. “On the Limits of Framing effects: who can frame”, J of Pol.) And status is a big topic in soc psych. But I bet researchers from other areas of sociology would be interested in different questions, so I still agree with the call for more soc research experiments on all these topics.


    Andrei Boutyline

    February 26, 2011 at 11:57 pm

  4. Agree with the general call for more experimental work, but what about field experiments where establishments or organizations are studied before, during, and after exposure to an organizational change? The examples you gave seemed to point towards lab experiments.

    My comment is clearly self-serving. I’ve published 3 pieces recently utilizing a quasi-experimental design (Kelly et al 2010 in Gender & Society, Moen et al 2011 in Social Problems, and Kelly et al 2011 in April issue of ASR). We investigated how a “culture change” initiative focused on flexibility and work redesign affected the organizational culture, turnover, work-family conflict, and work-family fit.

    I’m also involved in an interdisciplinary, group-randomized field experiment in two industries — lots of great design issues to work through but also a great opportunity to see how employees’ perceptions, managers’ attitudes and behaviors, everyday work practices, and organizational systems respond to an effort to change organizational norms. I’d love to have more people who identify as organizational scholars involved in this work.


    Erin Kelly

    March 5, 2011 at 9:08 pm

Comments are closed.

%d bloggers like this: