Mabel Abraham – Columbia Business School
Monique Alexandria Domingo – UConn School of Business
Hannah Weisman – Department of Management, LSE
Article link: https://doi.org/10.1177/0001839219832813
1. Background. Your research contributes to a growing body of research about the audience-based mechanisms of gender bias. In basic terms, why does the mere presence of an audience induce gender bias?
When we think of gender bias we often think of the following scenario: How does a person’s beliefs or opinions lead them to favor men over women? Researchers, including myself, have spent a lot of time thinking about this scenario because in many cases this is how biases manifest. However, the presence of audiences adds an important dimension. In some cases, we are incentivized to consider the beliefs or preferences of others, such as a focal audience, when making decisions. When we care about gaining the approval or acceptance of another person(s), we are likely to make decisions that we expect will garner that approval. What this means for studying gender bias is that we typically do not know for certain whether a given audience is biased. Instead, we are left to make inferences about their likely gender preferences. In making these inferences, we rely on generally held beliefs. Therefore, the mere fact that we know people tend to be biased can lead us to conclude a given audience is likely to be biased, irrespective of whether that is the case.
2. Method. Your study generates important insights about resource exchanges among entrepreneurs and their contacts. Why did you select this entrepreneurial research setting, and how do you think the results might compare in a non-entrepreneurial context?
My initial motivation for looking at networking among entrepreneurs was two-pronged. First, we know that networks are particularly important for entrepreneurs. Second, it is well-established that female entrepreneurs fare worse than male entrepreneurs across various stages of entrepreneurship, from founding to running businesses. Gender differences in networks are often posited as a key explanation, namely that because women tend to be in worse networks, they are less able to access necessary resources.
That being said, there is little reason to believe the patterns I observed here would be significantly different than what we would expect to see among men and women more broadly. Further unpacking gender differences in networking across context is certainly an open area for future research.
3. Method. Your research mainly draws on quantitative data, but you also collected qualitative data via in-person observations and semi-structured interviews. Could you share your perspectives on how supplementing quantitative work with some qualitative data enhances your research?
I think carefully about the appropriate data for answering a given research question in each of my projects. In my papers, I typically start with the research question, even if that question is not fully developed, as opposed to starting with an available data source. In this case, I identified my research setting (RefClubs) as a possible opportunity for examining gender differences in resource exchange among entrepreneurs, which was something I had long been interested in pursuing. Before embarking on the long data collection process necessary to gain access to the quantitative data, I wanted to be sure I fully understood my setting, and that is where the qualitative component was most critical for this study. For example, it was only through my observations of many different networking groups that I became confident RefClubs would be an appropriate context. I then interviewed members of these groups to understand their experiences and how they made decisions about sharing resources.
As a quantitative researcher, the nature of the qualitative work I do in my research, including in this paper, is not at all the same as engaging in a full ethnography or interview study. My main arguments and conclusions come from the quantitative data. But my observations and interviews certainly helped to inform this study, from the data collection, to the model specifications, to offering support for the inferences and claims I was able to make from my quantitative models. In this way, I would say developing a qualitative understanding of one’s research setting is always worth the extra investment.
4. Method. You acquired access to RefClubs, an organization that promotes entrepreneurs’ business growth by facilitating resource exchanges. How did you find and gain access to the organization? Do you have any advice for Ph.D. students who are trying to find and gain access to organizations for their dissertations?
Simply stated, gaining access to field data is hard! At the time that I was pursuing access to RefClubs, I was also working on a few other leads that fell through. In total, I believe it took me nearly 2 years to gain access and collect all of my data. It certainly is not a task for the faint of heart.
My advice to students is to set realistic expectations and recognize that most organizations say no, even when they initially seem interested. It is also always helpful if you can leverage your own networks to make connections. In this case, my advisor’s neighbor was a member of one of these networking groups and helped me schedule that first meeting. But often that is not the case and you simply need to figure out who you can contact and do some digging. The inherent uncertainty can be very stressful as a graduate student (and as a pre-tenured faculty member). One must do the calculation of whether accessing organizational data provides an ideal means for answering the research question, which relates to my previous point above. The main thing I would encourage students to avoid is investing the time and effort to gain access to organizational data based on the assumption that once they have the data, they will find an interesting question to answer.
5. Findings and Future Research. Your paper finds that there are certain conditions under which the presence of an audience is more or less likely to fuel gender bias. Can you please elaborate on the significance of these findings? What other boundary conditions do you think might be fruitful to examine in future research?
I find that people are more likely to advantage men in exchanges involving an audience when considering a female network contact who is in a male-typed job or field. This finding is consistent with other research showing that the activation of gender biases more generally is most pronounced in these male-typed contexts. Considering my earlier points about how audiences induce gender biases, when one is deciding whether to select a male versus a female network contact, they will consider the likely gender preferences of their audience. In making inferences about the audience’s gender preferences, it is reasonable to assume that an audience is more apt to prefer, or at least expect a man, in male-dominated areas, but less so to prefer a man in more gender-neutral or female-dominated areas.
One clear next step for research in this vein would be to examine how characteristics of the individual deciding whether or not to share resources, or of the audience, affect this observed pattern. We are really at the tip of the iceberg in understanding the mechanisms leading to gender differences in network benefits more broadly. I look forward to continuing to conduct research in this area and seeing what other scholars, including current doctoral students, are doing to help advance knowledge in this area!