SCROLLING TO THE FOSSILS: EXAMINING PROFILE AND VIDEO VIEWERSHIP IN PALEONTOLOGY TIKTOK

The advent of social media has provided novel ways to engage in public science education and outreach. Scientists and science educators are able to reach diverse audiences and share experiences and insights that might not otherwise be accessible. Regarding outreach in paleontology, past studies have examined how to optimize engagement on Facebook, Twitter and Instagram. Engagement metrics, such as number of likes, views and followers, can inform scientists and science educators if and how their message was received. Here, we look at TikTok, a social media app that allows users to watch and upload videos up to three minutes long in a vertical (portrait) format. I was interested in two questions: Can profile views be explained by video views for an inactive account? And can video views be explained by date since posted, number of likes or a combination of the two? To answer these questions, I performed single and multiple linear regressions using data from my own TikTok account, MacFossil, which contains only paleontological videos, 58 in total. Profile views could be explained by video views, although very weakly (R²=0.129, df=58, p=0.005). This is likely because viewership goes down during account inactivity which accounts for fewer people tapping on a profile. Video views could be weakly explained by days since posted (R²=0.153, df=56, p=0.002) and strongly by number of likes (R²=0.901, df=56, p=2.2e-16). When combined, number of days since posted did not significantly explain number of views (p=0.388) and number of likes did (p=2e-16) (R²=0.907, df=55). While newer videos tend to get more views (as preferred by the algorithm) and as an audience grows, older videos can still get views and may have been more successful upon posting. Likes, on the other hand, are directly tied to the algorithm. The algorithm will select videos for either a viewer’s “Following” or “For You” page partially based on the number of likes, thereby increasing the number of views. These findings may suggest that topics from older, poorer preforming videos could be redone to make them more engaging in order to increase viewership.