Back in December, there was a study that appeared in The British Journal of Psychology that got a fair amount of buzz. The paper (Horvath, M.A.H., Hegarty, P., Tyler, S. & Mansfield, S., " 'Lights on at the end of the party': Are lads' mags mainstreaming dangerous sexism?" British Journal of Psychology. DOI:10.1111/j.2044-8295.2011.02086.x) looked the influence that magazines aimed at young men ("lads' mags") might have on how the young people who read them perceive their social reality. Among other things, the researchers found that the subjects in the study found the descriptions of women given by convicted sex offenders and lads' mags are well nigh indistinguishable, and that when a quote was identified as from a lads' mag (no matter what its actual source), subjects were more likely to say that they identified with the view it expressed than if the same quote was identified as coming from a rapist.
I wrote about the details of this research in a post on my other blog.
One of the authors of the study, Dr. Peter Hegarty, is someone I know a little from graduate school (as we were in an anthropology of science seminar together one term). He was gracious enough to agree to an interview about this research, and to answer some of my broader questions (as a physical scientist turned philosopher) about what doing good science looks like to a psychologist. Owing to its length, I'm presenting the interview in two posts, this one and one that will follow it tomorrow.
Janet Stemwedel: Is there something specific that prompted this piece of research -- a particular event, or the Nth repetition of a piece of "common wisdom" that made it seem like it was time to interrogate it? Or is this research best understood as part of a broader project (perhaps of identifying pieces of our social world that shape our beliefs and attitudes)?
Peter Hegarty: We came to this research for different reasons. Miranda [Horvath] had been working more consistently on the role of lads’ mags in popular culture than I had been (see Coy & Horvath, 2011). Prompted by another students’ interests, I had published a very short piece earlier this year on the question of representations of ‘heteroflexible’ women in lads' mags (Hegarty & Buechel, 2011). The two studies reported in Horvath, Hegarty, Tyler & Mansfield (2011) were conducted as Suzannah Tyler and Sophie Mansfield’s M.Sc. Dissertations in Forensic Psychology, a course provided jointly by the University of Surrey and Broadmoor Hospital. Miranda and I took the lead on writing up the research after Miranda moved to Middlesex University in 2010.
JS: When this study was reported in the news, as the Twitters were lighting up with discussion about this research, some expressed concern that the point of the research was to identify lads' mags as particularly bad (compared to other types of media), or as actually contributing to rapes. Working from the information in the press release (because the research paper wasn't quite out yet), there seemed to be some unclarity about precisely what inferences were being drawn from the results and (on the basis of what inferences people thought you *might* be drawing) about whether the research included appropriate controls -- for example, quotes about women from The Guardian, or from ordinary-men-who-are-not-rapists. Can you set us straight on what the research was trying to find out and on what inferences it does or does not support? And, in light of the hypotheses you were actually testing, can you discuss the issue of experimental controls?
PH: Our research was focused on lads’ mags –- rather than other media –- because content analysis research had shown that those magazines were routinely sexist, operated in an advice-giving mode, and often dismissed their social influence. This is not the case –- as far as I know -- with regard to The Guardian. So there was a rationale to focus on lads’ mags that was not based on prior research. We hoped to test our hypothesis that lads’ mags might be normalizing hostile sexism. This idea hung on two matters; is there an overlap in the discourse of lads’ mags and something that most people would accept as hostile sexism? Does that content appear more acceptable to young men when it appears to come from a lads’ mag? The two studies mapped onto these goals. In one, we found that young women and men couldn’t detect the source of a quote as coming from a convicted rapist’s interview or a lads' mag. In another, young men identified more with quotes that they believed to have come from lads’ mags rather than convicted rapists.
JS: While we're on the subject of controls, it strikes me that good experimental design in psychological research is probably different in some interesting ways from good experimental design in, say, chemistry. What are some misconceptions those of us who have more familiarity with the so-called "hard sciences" have about social science research? What kind of experimental rigor can you achieve without abandoning questions about actual humans-in-the-world?
PH: You are right that these sciences might have different ontologies, because psychology is a human science. There are a variety of perspectives on this, with scholars such as Ian Hacking arguing for a separate ontology of the human sciences and more postmodern authors such as Bruno Latour arguing against distinctions between humans and things. Generally, I would be loath do describe differences between the sciences in terms of the metaphor of “hardness,” because the term is loaded with implicature. First, psychology is a potentially reflexive science about people, conducted by people and is characterized by what the philosopher Ian Hacking calls “looping effects;” people’s thoughts, feelings and behaviours are themselves influenced by psychological theories about them. Second, measurement in psychology is more often dependent on normalization and relative judgment (as in an IQ test, or a 7-point Likert item on a questionnaire, for example). Third, there is a lot of validity to the Foucaultian argument that the “psy- disciplines” have often been used in the service of the state, to divide people into categories of “normal” and “abnormal” people, so that different people might be treated very differently without offending egalitarian ideologies. Much of clinical psychology and testing takes this form.
Critics of psychology often stop there. By so doing, they overlook the rich tradition within psychology of generating knowledge that troubles forms of normalization, by suggesting that the distinction between the “normal” and the “abnormal” is not as firm as common sense suggests. Studies in this tradition might include Evelyn Hooker’s (1957) demonstration – from that dark era when homosexuality was considered a mental illness – that there are no differences in the responses of gay and straight men to personality tests. One might also include David Rosenhan’s (1973) study in which ordinary people managed to deceive psychiatrists that they were schizophrenic. A third example might be stereotype threat research (e.g., by Claude Steele and Joshua Aronson, 1995), which shows that the underperformance of African Americans on some standardized tests reflects not genuine ability, but a situational constraint introduced by testing conditions. Like these studies, we would hope ours would trouble’s people’s sense of what they take for granted about differences between people. In particular we hope that people will reconsider what they think they know about “extreme” sexism – that leads to incarceration – and “normal” sexism, that is now typical for young men to consume. I would urge academic critics of psychology – particularly those that focus on its complicity with Foucaultian disciplinary power, and the power of the state more generally - to develop more critiques that can account for such empirical work.
For the last half a century, “rigor” in empirical psychology has been organized by the language of validity and reliability of measurement (Cronbach & Meehl, 1955). Psychologists also tend to be Popperians, who construct “falsifiable” theories and use Fischerian inferential statistics to construct experiments that afford the possibility of falsification. However, inferential norms are changing in the discipline for three reasons. First, the rise of neuroscience has lead to a more inductive form of inference in which mapping and localization plays a greater role in scientific explanation. Second, social psychologists are increasingly engaging with structural equation modelling and offering confirmatory models of social processes. Third, there is “statistical reform” in psychology, away from the ritual of statistical significance testing toward making variability more transparent through the reporting of confidence intervals, effect sizes, and exact significance values. See Spellman (2012) for one very recent discussion of what’s happening within the genre of scientific writing in psychology around retaining rigor and realism in psychological science.
JS: One thing that struck me in reading the paper was that instruments have been developed to measure levels of sexism. Are these measures well-accepted within the community of research psychologists? (I am guessing that if the public even knew about them, they would be pretty controversial in some quarters ... maybe the very quarters whose denizens would get high scores on these measures!)
We used two well-established measures; the ambivalent sexism inventory and the AMMSA, and one measure of endorsement of lads’ mags that we developed ourselves for the study. We describe some of the previous findings of other researchers who have used these scales to examine individual differences in responses to vignettes about sexual violence in the article. We feel more confident of the measure we developed ourselves because it was highly correlated with all other measures of sexism and because it was highly correlated with men’s identification with quotes from rapists and from lads’ mags. In other words, we followed the logic of psychologists such as Lee Cronbach, Paul Meehl and Donald Campbell for establishing and developing the “construct validity” of the empirical scales.
* * * * *
Tomorrow, in the second part of my interview with Peter Hegarty, we discuss the WEIRD-ness of college students as subjects for psychological research, how to go from description to prescription, and what it's like for scientists to talk about their research with the media in the age of Twitter. Stay tuned!
Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52, 281-302.
Coy, M., & Horvath, M.A.H. (2011).‘Lads mags’, young men’s attitudes towards women and acceptance of myths about sexual aggression. Feminism & Psychology, 21, 144-150.
Foucault, M. (1978). Discipline and Punish: The Birth of the Prison. Trans. Alan Sheridan. New York, Random House.
Hacking, I. (1995). The looping effects of human kinds. In Dan Sperber, David Premack and Ann James Premack (Eds.), Causal Cognition: A Multi-Disciplinary Debate (pp. 351-383). Oxford, UK: Oxford University Press.
Hegarty, P., & Buechel C (2011) '"What Blokes Want Lesbians to be”: On FHM and the socialization of pro-lesbian attitudes among heterosexual-identified men'. Sage Publications Feminism & Psychology, 21, 240-247.
Hooker, E. (1957). The adjustment of the male overt homosexual. Journal of Projective Techniques, 21, 18-31.
Horvath, M.A.H., Hegarty, P., Tyler, S. & Mansfield, S. (2011).“Lights on at the end of the party”: Are Lads Mags’ Mainstreaming Dangerous Sexism? British Journal of Psychology. Available from http://onlinelibrary.wiley.com/doi/10.1111/j.2044-8295.2011.02086.x/abstract
Latour, B. (1993). We Have Never Been Modern. Cambridge, MA: Harvard University Press.
Rosenhan, D.L. (1973). On being sane in insane places. Science, 179, 250-258.
Spellman, B.A. (2012). Introduction to the special section: Data, data everywhere. . . especially in my file drawer. Perspectives on Psychological Science, 7, 58-59.
Steele, C., & Aronson, J. (1995). Stereotype threat and the intellectual test performance of African Americans.” Journal of Personality and Social Psychology 69, 797-811.