Cracking the Code: Survey Takes A 'Deep Look' at Science Video Audience and Gender Disparity

Earlier this fall Deep Look released its 100th video, about ants that kidnap other ants’ babies and enslave them. The episode has everything that makes our series of short videos about small animals and plants popular: stunning photography, a gripping story and great music.

By nearly any measure, Deep Look is a success. The series, which is produced by KQED public media and presented by PBS Digital Studios (PBSDS) on YouTube, has 1.2 million subscribers and reaches millennials, a younger audience that PBS is eager to serve. Our episodes receive hundreds of thousands of views every week and the series has won numerous awards across traditional and new media fields.

But Deep Look has a problem. For almost every one of our episodes, the percentage of women who watch is considerably lower than the percentage of men, a disparity that also happens on other science shows distributed by PBSDS. On average, about 70% of Deep Look’s YouTube audience is male and only 30% is female. Our audience’s disparity is even more pronounced than that of YouTube’s average audience, which is 60% male. Naturally, this is concerning to our team.

“Part of our mission at public media is to make media for everyone,” says Lauren Sommer, Deep Look host and co-writer and a radio reporter covering science and the environment at KQED. “We really want to make sure we’re reaching all audiences.”

So why aren’t more women watching?

We’re trying to figure this out with the help of a team of researchers from Yale and Texas Tech universities, and funding from the National Science Foundation. Our first research results, from a national survey of 2,500 people conducted in June by Yale professor of psychology Dan Kahan, are encouraging. They show that once women click on our videos they’re just as engaged with them as men. In other words, the women watch for as long and are just as likely to share them with their friends.

But something is keeping some women from clicking on our episodes.

One of our hypotheses going into the survey was that perhaps the YouTube algorithm — that mysterious recommendation engine that serves different videos to different audiences — is to blame. We thought that YouTube might be offering our videos to more men than women. But the survey showed that a similar gender disparity occurs when people watch our videos outside of YouTube, as they did for our survey.

So if the YouTube algorithm and the substance of our stories aren’t keeping women away, what are some other factors to investigate? Currently, we’re taking a look at the word choices in the titles of our videos and planning a second survey to explore this and other ways in which we might encourage more women to watch.

We’re hoping that what we discover about viewing patterns for Deep Look might be of use to other media organizations, along with educators, scientists and others who would like to see gains in the public’s understanding of science and environmental issues. So stay tuned as we try to get more women to take a Deep Look.

Below is a summary of the survey’s findings. You can read the full report, called “A Deep Look at Gender Disparity” here. 

1. The Deep Look audience gender disparity can be reproduced experimentally. This finding implies, among other things, that the disparity is not a consequence of the YouTube algorithm or related online and social media platforms.
2.  The disparity occurs because high-science-curiosity women are less likely to choose to view certain Deep Look episodes than are high-science-curiosity men. “Science curiosity” is a measure of the propensity of individuals to voluntarily consume science-related material for personal satisfaction. The disparity in viewing is not a natural or inevitable consequence of any difference in the satisfaction that men and women take, respectively, in being exposed to scientific insights into the workings of nature. Indeed, when high-science-curiosity women do view Deep Look episodes, they are just as engaged by them as high-science-curiosity men. Some other influence thus appears to impede women from electing to view episodes.
3. The difference in viewing rates among men and women is concentrated in high-science-curiosity women with modest levels of science comprehension — and disappears among high-science-curiosity women with higher levels of science comprehension. This surprising finding suggests that some unobserved disposition associated with science comprehension inhibits women (but not men) from availing themselves of an opportunity to satisfy their interest in science by choosing to view certain Deep Look videos. The subjects’ level of “science comprehension” was measured using standardized assessment questions incorporated into the survey.
4. Aversions to the subject matter associated with disgust do not appear to be responsible for the Deep Look gender disparity. Study measures that predict aversion to disgusting stimuli, such as lice crawling around human hair as seen in this Deep Look episode, were not correlated with viewing decisions for men or women. Self-reported disgust aversions did vary among men and women but not in patterns that corresponded to gender differentials in viewing decisions.

For more information about this study, see the attached report. To learn more visit kqed.org/crackingthecode.