Gender Bias in Researcher-to-Researcher Communication

22 March 2013 by Matt Shipman, posted in Uncategorized

Image credit: Karsten W. Rohrbach

Science communication extends beyond the attempts of reporters and scientists to communicate with nonexpert audiences (i.e., the so-called “general public”). For example, science communication also describes the efforts of researchers to communicate with each other. And, as a recent study points out, that process includes challenges that researchers have little control over. Challenges like gender bias.

Among other findings, an Ohio State University study found that graduate students rated research abstracts as having greater “scientific quality” when they thought the abstracts were written by men. A paper describing the study was published online by Science Communication in February. The results are interesting, if saddening, and worth discussing. One qualifier should be mentioned up front: the study participants were all graduate students in communication research programs, and the abstracts they were reviewing were all of communication research. The paper itself notes: “Given that biases against women have been shown to vary by scientific discipline, it seems likely that examination of different academic fields would show biases of different extent.”

Researchers began with a collection of abstracts, and divided them into “types” according to whether the subject matter covered topics normally associated with male, female or gender-neutral norms. Abstracts the researchers listed as “female-typed” often dealt with children or gender issues; “male-typed” abstracts included political issues; and gender-neutral abstracts looked at health-related issues. The researchers then created a list of fictional author names that were clearly identifiable as male or female (e.g., Gary or Lisa).

The researchers then sent a collection of abstracts to the study participants: 243 graduate students from 20 different institutions across the United States, of whom 70 percent were female. Each study participant was presented with two gender-neutral abstracts, two male-typed abstracts and two female-typed abstracts. For each of those categories, one of the abstracts listed two female author names and one abstract listed two male author names. The researchers rotated the names and abstracts across the pool of participants.

The researchers found that not only were male authors associated with greater scientific quality, but that “the contributions from male authors were perceived as having particularly high scientific quality if they pertained to male-typed topics.” In fact, the abstracts with the highest ratings of scientific quality were those on male-typed topics that were associated with male authors.

The lowest ratings of perceived scientific quality went to abstracts where there was a disconnect between the sex of the author and the “type” of research. Female authors of abstracts focusing on male-typed research scored lowest, and male authors of abstracts on female-typed research did only slightly better.

When study participants were also asked to rate their interest in collaborating with the abstract authors, the results were similar. The highest rates of interest in collaboration were for male authors whose abstracts addressed male-typed research issues, followed by female authors whose abstracts addressed female-typed research issues.

Ultimately, the study found that – at least in the communication field – gender bias plays a role in how research is evaluated. That has implications for everything from how grant proposals are reviewed to promotion and tenure reviews. “The overall conclusion,” as the paper states, “is that male scholars will have a much smoother ride.”

I have no idea what can be done to change this or to mitigate the effects of gender bias. But I wanted to bring the subject up, because ignoring it won’t make it go away.

Note: The paper is not open access. Citation below.

The Matilda Effect in Science Communication: An Experiment on Gender Bias in Publication Quality Perceptions and Collaboration Interest,” Science Communication, Silvia Knobloch-Westerwick, Carroll J. Glynn and Michael Huge, DOI: 10.1177/1075547012472684


6 Responses to “Gender Bias in Researcher-to-Researcher Communication”

  1. Joanne Kamens Reply | Permalink

    Ignoring it really won't make it go away. Just about the only thing you can do about Implicit Biases (which these data clearly highlight) is to radically raise awareness. Thanks for this great summary

  2. Thomas Lumley Reply | Permalink

    Conference abstracts or paper abstracts? If the latter, perhaps the names just shouldn't be in the abstract -- it's traditional, but not really necessary.

  3. Ivana Gadjanski Reply | Permalink

    As a female scientist I can only say I'm horrified but not that surprised. It would be interesting to see if the journal editors and reviewers follow similar logic...

    • Bob O'H Reply | Permalink

      No they don't seem to. There have been several studies of this, and there was a meta-analysis that showed no effect of blinding (which I can't find, but this one study finds an effect of gender, but also heterogeneity).

      Damn, I really should find that paper. :-(

  4. Michelle Paulsen Reply | Permalink

    I have used the Implicit Association test (https://implicit.harvard.edu/implicit/demo/) to introduce just the idea that people could hold unconscious biases. It was a good springboard for discussions with graduate students - which is at least a start.

Leave a Reply


8 − = two