The myth of gender cultures: similarities outweigh differences in men's and women's provision of and responses to supportive communication.

Erina L. MacGeorge, Angela Graves, S Gillihan B. Feng and B. Burleson

Research output: Contribution to journalArticle


Michaud and Warner (1997) and Basow and Rubenfeld (2003) recently reported studies of gender differences in “troubles talk” that allegedly provide support for the different cultures thesis, that is, the notion that men and women communicate in such different ways that they should be regarded as members of different communication cultures or speech communities communities. In this article, we identify several limitations in these two studies that, collectively, have the effect of casting doubt on their conclusions. We then report three studies that show that men and women provide and respond to supportive messages (“troubles talk”) in ways that are much more similar than different. The current findings, in conjunction with other recent findings, suggest that the different cultures thesis is a myth that should be discarded. The authors suggest that previous studies finding support for the different cultures thesis should be discarded because of small effect sizes. The authors suggest that an effect size of r2 > .16 and d>.87 should be used to support the different cultures thesis. This may be a high bar. r2 for gender and height is .33 and d = 1.41 so the author's bar is about half of that effect size. Likewise Cohen (1980) proposes that effect sizes in social science research should be small (d = .2), medium (d = .5) and large (d = .8) effect sizes. The greatest effect sizes found in the studies by Michaud & Warner are r2 = .12 and in Basow & Rubenfeld, r2 = .09. We don't know the Cohen's ds. McGraw and Wong (1992, Psychological Bulletin, 111: 361-365) proposed an effect size statistic called the common language effect size statistic (or CL) which for a two group design with a continuous dependent variable is the probability that a randomly selected score from the one population will be greater than a randomly sampled score from the other distribution. As an example they use sexual dimorphism in height among young adult humans. National statistics are mean = 69.7 inches (SD= 2.8 ) for men, mean = 64.3 inches (SD = 2.6) for women. If we assume that the distributions are normal, then the probability that a randomly selected man will be taller than a randomly selected woman is 92%, thus the CL is 92%. They argue that the CL is a statistic more likely to be understood by statistically naive individuals than are the other available effect size statistics. So, yes, the effect sizes by Michaud & Warner are small. MacGeorge et al also conduct their own studies which compensate for some deficiencies in the methodologies of the previously mentioned studies. They did find an effect for gender with men giving more advice and women more likely to offer help and affirm the other, but these differences accounted for only 2% of the variance (as a comparison, sex accounts for 33% of the variance in height). There are more similarities than differences, however, with men and women both expressing sympathy and giving advice in larger frequencies than other types of responses. The researchers do propose that there may be differences in the quality of sympathy provided by women (or, my cavet, the *perceived* quality). In other study looking at responses to comforting messages, MacGeorge et al again find gender differences, but more similarities than differences. Author's conclude that while the different cultures thesis is overstated, there are differences in the SKILLS of men and women in providing support, which is why both genders like to get their support from women. They also conclude that there is no reason to expect that an appreciation for gendered communication styles will improve the interactions of men and women. My response: First, small differences are worth paying attention to because we have thousands of communicative acts in a day. These differences, even if small, are continually reinforced. There are multiple opportunities in a day for men to perform behaviors that alienate women or women to perform behaviors that raise doubt in men. Moreover, skill doesn't fully explain the differences because men and women are perceived differently based on the communication acts that they perform. There are 3 aspects to consider with any communicative act: (1) Speaker Production: men and women produce different Speech acts (2) Audience Reception: men and women perceive the appropriateness of SAs differently (3) Perception of Speaker: male and female speakers are perceived differently for the same SA These 3 factors reinforce each other. Second it is important to look at gender differences in communication because communication style can be CHANGED. While women may not be able to change their confidence, they can change how they PERFORM confidence in their everyday interactions with others. Such performances may have a reinforcing/reciprocal/cumulative effect: if women are perceived as more confident, others are less likely to question their competence and treat them with respect. This trust and respect can, in turn, boost actual confidence. Likewise, while men may not be able to consciously change cultural stereotypes that have been deeply engrained, they can change highly masculine and aggressive patterns of interaction that have been shown to alienate and exclude women. In fact, there is some research to indicate that reducing stereotypically masculine patterns of self-performance and communication can substantially improve both workplace and learning environments ("Masculine Men", survey of NASA engineers, Rudman--negative things w/ self-promotion). Macgeorge et al also don't compare their generalizatins about cultural differences to accepted cultural differences. For instance, they don't state what the differences are between american and Japanese forms of support. Here are presumably two different cultures. Matsumudo (1999) found effect sizes of r2=.15 and r2=.07 for American vs. Japanese ratings of facial expressions and conclude strong cultural differences. Kityama et al don't report their effect sizes for American vs. Japanese self-esteem and self-criticism, but an eyeball estimate suggests that they would probably be small. Ambady, Koo, Lee et al 1996 "More than words: Linguistic and nonlinguistic politeness in two cultures" found more similarities than differences between American and Korean ratings f politeness strategies. Park & Kim found d = .31 in American vs. Asian preferences for open style and d = .64 in preferences for indirect style. Third, there is reason to suspect that MacGeorge et al also look at supportive communication, but the biggest differences have been found in interruptions. Anderson & Leaper's meta-analysis of interruption studies found an overall effect size of d=.15 for male interruptions and a size of d=.33 for intrusive interruptions. This is still small.
Original languageEnglish (US)
Pages (from-to)143-175
Number of pages33
JournalSex Roles
Issue number3-4
StatePublished - 2004

Fingerprint Dive into the research topics of 'The myth of gender cultures: similarities outweigh differences in men's and women's provision of and responses to supportive communication.'. Together they form a unique fingerprint.

  • Cite this