Women in the technology sector are using Twitter to share fear and vent, according to analysis by the TechPORTFOLIO team with the IBM Watson Tone Analyzer.

As part of our month-long series highlighting women in tech and their unique experiences and challenges, we looked at a list of prominent women in the sector who were active on Twitter and used cognitive analysis to determine the tone of the tweets sent to them in public. We wanted to find out whether there was a difference in the way that people in public and their community spoke to them on Twitter, compared with the technology sector at large.

Women on Twitter

Control Group

The figures above seem to indicate that there is much more anger and fear present in conversation aimed at women in technology, and marginally more emotional range and openness.

A large number of the tweets with a degree of fear or anger were people seeking or offering mutual emotional support, for matters large and small.

The tech community in the control group had more conversations, but they were more one-sided. There were a far higher level of “drive by” tweets without any real attempt to engage in conversation; such as requests to retweet, or feature suggestions.

In our analysis, we didn’t see that many examples of outright abuse directed at our women in tech list. The control group — which was filled with several high-tier individuals such as Apple CEO Tim Cook, philanthropist Melinda Gates, and “inventor of the web” Tim Berners-Lee — had several tweets that been marked as deleted. It’s possible that any retroactive reporting on Twitter’s language and possible tendency towards abuse might be affected by reporting of offensive content and moderation.

This shows that despite Twitter’s ongoing issues with harassment against women, the platform is still a valuable community space. Settings that allow for trusted conversation, like these from UX researcher Caroline Sinders, would support this while allowing for some protection against drive-by abuse.

https://twitter.com/dinosaurrparty/status/740227885772898308

Try the Tone Analyzer now. If you want to incorporate IBM’s Watson into your own application hosted in the cloud, click here.

Method

For our group, we selected the Anita Borg institute’s Twitter list of 500 most important women to follow on Twitter. Of those, we found the 50 most active, and took one week of tweets: June 8 to June 15, with a maximum of two @-replies each day.

There are no “men in tech” Twitter lists, of course, at least ones that are from comparatively reliable sources. Our control group, therefore, was Robert Scoble’s popular list of most influential people in technology. Several of the people on the list are, of course, women; a straight comparison with the online population at large works better for our purposes, anyway. Once again, we took the fifty most recently active users.

The control group contained several high-tier individuals on Twitter, making the volume of @-replies incredibly high. The TechPORTFOLIO team had to cut the time period down to 24 hours only, for June 8–and even then, we had to remove an unusually high spike of tweets aimed at @davidplouffe, an Uber board member and Barack Obama election campaign strategist. (Sorry, Mr. Plouffe, you were already having a bad day.)

For a free trial of IBM Watson Analytics click here.