work meeting
Editorial

What Real-Time Collaboration Data Can Tell Us About Gender Diversity

5 minute read
Laurence Lock Lee avatar
Studies of social interactions in the workplace did not reveal evidence of gender bias. Further research involving sentiment analysis may be needed.

Do men show an unconscious bias against women at work? Are men slower to respond to women in the workplace? Are women quick to respond because they feel they need to advance their own agendas, given that men still dominate senior management levels?

The answers to those questions could reveal a dynamic in the workplace that could disadvantage women.

Looking for Gender Bias in Enterprise Social Networks

Following up on my earlier CMSWire article on gender bias in the workplace, this article presents some results of analysis of enterprise social networking data in an effort to explore the possibility of unconscious gender bias suggested in the earlier article.

Prompted by our research partners at the Peace Innovation Lab at Stanford, we wanted to explore the time it took for people to respond to posts made on an enterprise social networking (ESN) platform. We used the SWOOP social networking analysis platform to gather collaboration data from a large financial institution with more than 30,000 users and a nearly 50/50 gender split.

The hypothesis we started with was that perhaps men might show an unconscious bias against women by responding to women less frequently, or at least more slowly, than they respond to men. Moreover, it could also be the case that women respond more quickly to men in an effort to advance their agendas.

Related Article: Addressing Gender Bias in the Workplace: A New Approach

A Deep Dive Into ESN Engagement 

In a nutshell, we found no such bias. In fact, the significant finding was that women received responses significantly faster than men did. This study has the advantage of a large data set, some 33,000 participants, nearly perfectly divided by gender (50.1 percent female and 49.9 percent male). We analyzed about 50,000 posts, looking for the time that the first written reply was received, if at all. We disregarded “likes” because we wanted to measure the substantive response a written reply represents. We identified the gender of both the poster and the person who first responded with a reply, along with the time it took for that first reply to arrive. Here are the results:

 ESN MembershipPosts MadeFirst Reply Response Made
Female16,598 (50.1%)25,310 (50.3%)10,629 (57.46%)
Male16,534 (49.9%)25,007 (49.7%)7,870 (42.54%)
Totals33,13250,31718,499

About 37 percent of the 50,3117 posts received written replies. While the posting frequency did not differ between males and females, women were significantly better at replying to posts.

In terms of time to respond, there was not a significant difference between the time it took for a male to reply to a post and the time it took a female to reply. The differences occurred when we looked at intergender time to respond, where we found the following results:

 Female Reply ReceivedMale Reply Received
Female Reply MadeAverage 35.28 hours
(5,645 first replies)
Average 43.39 hours
(3,959 first replies)
Male Reply MadeAverage 33.95 hours
(3,518 first replies)
Average 42.86 hours
(3,572 first replies)

It is clear that females drew more and faster responses than males. Females received, in total, 9,163 written replies, with males replying faster than females on average (33.95 hours versus 35.28 hours). In contrast, males received only 7,531 written replies to a roughly equal number of posts. And those responses came significantly slower (from males 42.86 hours and females 43.39 hours). The sisterhood is strong in that women respond more frequently and more quickly to other women than they do to men. And as for males, they may respond nearly the same to females as they do to males, but they respond faster to females than they respond to other males. If anything, the gender bias is the reverse to what we had hypothesized.

Learning Opportunities

Related Article: Looking Back: An Odd Year for Women in Tech

Implications for Gender Bias in Workplace

A recent article in the Harvard Business Review titled “A Study Used Sensors to Show that Men and Women are Treated Differently at Work” came to this conclusion when sociometric badges, worn by some 100 staff of a particular company, resulted in no perceptible behavioral differences between the genders:

“... we found almost no perceptible differences in the behavior of men and women. Women had the same number of contacts as men, they spent as much time with senior leadership, and they allocated their time similarly to men in the same role.”

In other words, if social interactions could not explain the imbalance of genders at senior levels, then what is left is the unconscious bias in the way women are treated when being considered for promotions.

In our study, we did find a difference, but the bias was more against men than women. So if gender bias cannot be detected through social interactions online (this study) and offline (as in the study published in the Harvard Business Review), what other data-driven approaches are open to us? Neither study was able to provide operational evidence supporting explicit gender bias against women. 

The HBR study still promotes the idea of using hard data, rather than tenuous survey data, to address the gender bias issue. Specifically, the researchers suggest more rigorous data-driven approaches to assessing behavior and advancement. For my part, I think there is more we can do with sentiment analysis by looking at the negative and positive sentiments employed by the genders in interacting with each other. That’s something for a future study.

The journey has just begun in using interaction data to combat gender bias in the workplace. But in the meantime, if you have a burning question that needs a quick response, get a woman to ask it.

About the author

Laurence Lock Lee

Laurence Lock Lee is the co-founder and chief scientist at Swoop Analytics, a firm specializing in online social networking analytics. He previously held senior positions in research, management and technology consulting at BHP Billiton, Computer Sciences Corporation and Optimice.