No face reveals: the struggles of YouTube’s female scientists

By Joshua Buchi

Title: Exploring the YouTube science communication gender gap: A sentiment analysis

Author(s) and Year: Inoka Amarasekara and Will J Grant (2018)

Journal: Public Understanding of Science (open access)

​TL;DR: Female-hosted science communication channels receive higher proportions of appearance-based, hostile, critical/negative, and sexist/sexual commentary, which the authors theorize may be one of multiple reasons women are underrepresented in YouTube’s science communication community. Why I chose this paper: I believe that it’s important to understand how gender can affect the way people use YouTube if we want to increase representation to more equitable levels and lessen the barriers that make it hard for women to create.

Background

YouTube as a platform has redefined modern content creation and established pathways for people to produce media that were unheard of just twenty years ago. However, despite lacking the barriers to entry inherent in most forms of media, YouTube’s most popular channels have been found to significantly lack in gender diversity when compared to traditional television. When considering YouTube’s popularity–and therefore the number of young people regularly consuming its content–a lack of diversity might be concerning. Parts of the population that don’t see themselves represented in these spaces–such as non-men–may feel excluded by the platform and choose not to engage in it, amplifying the disparities in representation and creating a new, biased barrier to entry.

This disparity in representation also exists across STEM-focused channels on YouTube, leading to the ‘the gender gap’ – in this context meaning the disproportion between the number of male and non-male people with STEM-focused channels on YouTube. While this paper focuses specifically on women, the authors plan to conduct similar research on other marginalized groups, such as nonbinary people, in the future. Present theories about STEM’s gender gap argue that beliefs that scientists are supposed to be ‘objective’ and ‘assertive’, traits typically attributed to masculinity, create implicit bias against women in STEM. This can lead to women not feeling as though they belong in STEM, women being systematically under-recognized in STEM fields, and people being hostile towards women in STEM fields. In an effort to understand the ways this might affect the gender gap on YouTube, Amarasekara and Grant sought to investigate the way people on the platform responded to female STEM creators. They theorized that the anonymity of YouTube’s online environment may enable increased hostility, which might contribute to the lack of female participation in science communication on YouTube.

The Study

The authors selected the top STEM-themed channels with at least five videos that were included on ‘Education’ and ‘Science & Technology’ top channel charts. These channels were then categorized based on their gender presences, the gender associated with either the hosts or the relevant voiceovers. These categories were determined based on the presenters’ self-identification, designations from a video by theBrainScoop used to gather some of the female-led channels, and the linguistic performances of the presenters. (The authors note here that their categorization fails to reflect the reality of non-cisgendered creators and people who fall outside the gender binary, and hope to do future research exploring other identities and the intersections therein.)

As this selection resulted in a lack of female-hosted and narrated channels, the authors included 21 additional channels from a recommended list of female-led STEM channels.

The channels were sorted into six broad gender presence categories: Continuous Male Host, Continuous Female Host, Male Voice-Over, Female Voice-Over, Neutral/Discontinuous Hosts, and Team of Hosts. Continuous channels had visible hosts of the correlating gender presence that were consistent across the channel’s videos, while discontinuous channels invited new speakers regularly and therefore constantly varied.

From these categories, 15 channels were selected for analysis on channel subscribers, channel views, video likes, video dislikes, video views, and video comments. Up to 100 comments from each were also analyzed to determine their general content and sentiment and then categorized into one of six categories: ‘appearance,’ ‘sexual and sexist,’ ‘hostile,’ ‘positive,’ ‘critiques and negative,’ and ‘general discussion’.

The Results

391 channels were analyzed in the final sample. Female presences made up under a fifth of the sample, but less than 5% of the views and less than 5% of the overall subscribers. Male presences made up the highest share of channel views and subscribers (with well over 50% in both categories) and represented almost half the overall sample.

The study found that gender presence affected the amount of dislikes, likes, comments, and subscribers each channel received per view. Interestingly, female-hosted channels were actually found to have more engagement per view (receiving more likes, comments, and subscribers proportionally) compared to other gender presence categories. While they also were found to receive more dislikes, the difference was not significant. In general, hosted channels seemed to regularly outperform voiced-over channels, though the effect was not universal.

The negative effects rear their head in the types of comments each category received.

Female-hosted accounts received higher proportions of appearance-based, sexual, hostile, and negative comments compared to the other categories and received fewer general comments–effects that were not seen in their male-hosted counterparts or female voiced-over channels. Interestingly, team-hosted channels also shared some of these trends, receiving more appearance-based and critique comments and fewer positive comments than the other presences.

Limitations and Takeaway

The data paints an ill picture for female science communicators on YouTube. Despite their high levels of numerical engagement compared to other groups, they also experience consistently worse direct engagement from their audiences through their comment sections. The fact that these effects were stronger across the board for channels that had a visible female host shows the unfortunate reality of presenting as female on YouTube, to the point where just being seen is enough to worsen your experience.

However, knowledge of these challenges could empower creators to counter them. The authors discuss how some female creators have opted towards restricting their comments, transitioning their content to narration, or taking their communication efforts to other platforms–and while it’s unfortunate this is required, the information still puts power into these creators’ hands. Other plans could also be devised with this in mind to help these creators best optimize their science communication approach and experience.

Though this study is limited by its use of a gender binary for analysis, alongside the limited set of characteristics used to sort these channels into the binary categories (such as manners of speaking and secondhand identification), the authors plan to conduct future research exploring the unique experiences of other groups on the platform, such as people that fall outside the gender binary, people of various ethnicities, and more. The partial insight developed in this study is intended as a first step towards future work on the experiences of various communities on YouTube.

Overall, understanding the way that minority groups across YouTube engage with science communication is vital to help empower people despite the challenges that persist as part of YouTube content creation. Studying this helps empower communicators to communicate more effectively, protect themselves, and to protect people in their communities.

Edited by Kay McCallum

Cover picture by ThisIsEngineering on Pexels