class: center, middle, inverse, title-slide .title[ # The Effect of Streaming Chats on Political Misperceptions ] .author[ ### Tiago Ventura | Center for Social Media and Politics | NYU ] .date[ ### <span style="font-size: 70%;"> Big Data for Development and Governance, 10/21/2022 </span> <br><br> <span style="font-size: 70%; font-style: italic"> Joint work with Kevin Munger, Katherine McCabe, Victory Asbury and Keng-Chi Chang </span> ] --- class:middle ### Summary of the Talk -- <svg viewBox="0 0 512 512" style="height:1em;position:relative;display:inline-block;top:.1em;" xmlns="http://www.w3.org/2000/svg"> <path d="M504 256C504 119 393 8 256 8S8 119 8 256s111 248 248 248 248-111 248-248zm-448 0c0-110.5 89.5-200 200-200s200 89.5 200 200-89.5 200-200 200S56 366.5 56 256zm72 20v-40c0-6.6 5.4-12 12-12h116v-67c0-10.7 12.9-16 20.5-8.5l99 99c4.7 4.7 4.7 12.3 0 17l-99 99c-7.6 7.6-20.5 2.2-20.5-8.5v-67H140c-6.6 0-12-5.4-12-12z"></path></svg> **Streaming chats (e.g Facebook Lives, Twitch...) accentuate misperceptions about mass public opinion** - Streaming chats are a representation of what social media is: - information consumption + audience engagement + contextual cues -- <svg viewBox="0 0 512 512" style="height:1em;position:relative;display:inline-block;top:.1em;" xmlns="http://www.w3.org/2000/svg"> <path d="M504 256C504 119 393 8 256 8S8 119 8 256s111 248 248 248 248-111 248-248zm-448 0c0-110.5 89.5-200 200-200s200 89.5 200 200-89.5 200-200 200S56 366.5 56 256zm72 20v-40c0-6.6 5.4-12 12-12h116v-67c0-10.7 12.9-16 20.5-8.5l99 99c4.7 4.7 4.7 12.3 0 17l-99 99c-7.6 7.6-20.5 2.2-20.5-8.5v-67H140c-6.6 0-12-5.4-12-12z"></path></svg> **Online Field Experiment + Comments Data from Facebook** - Experiment: Vary platform where users watch Democrats primary debate - Social Media Data: Scrapped comments from Facebook lives. -- <svg viewBox="0 0 512 512" style="height:1em;position:relative;display:inline-block;top:.1em;" xmlns="http://www.w3.org/2000/svg"> <path d="M504 256C504 119 393 8 256 8S8 119 8 256s111 248 248 248 248-111 248-248zm-448 0c0-110.5 89.5-200 200-200s200 89.5 200 200-89.5 200-200 200S56 366.5 56 256zm72 20v-40c0-6.6 5.4-12 12-12h116v-67c0-10.7 12.9-16 20.5-8.5l99 99c4.7 4.7 4.7 12.3 0 17l-99 99c-7.6 7.6-20.5 2.2-20.5-8.5v-67H140c-6.6 0-12-5.4-12-12z"></path></svg> **Main results - Comments accentuate misperceptions:** - Lower feeling thermometer to certain candidates subject to toxic, negative comments. - Distort inferences about candidate viability. --- class:middle ## Streaming Chats -- **Video Feed + Social Chat: All in one screen.** -- Popular among the younger generations. - Twitch (Amazon) - YouTube/YouTube Gaming (Google) - Mixer (Microsoft) - Facebook Lives -- --- ## Political Debates  --- ### Bolsonaro's Facebook Lives <img src="./figs/bolso.png" width="100%" /> --- class:middle ### Research Design **We combine a large scale ''field" experiment" with data scrapped from Facebook streaming chats.** -- **Three main experimental conditions** - Control (standard NBC broadcast) - Expert chat (538 website) - Streaming chat (Facebook) -- .red[To complement our analysis, we scrapped the comments from Facebook Lives ] -- --- ### Treatment Arms .pull-left[] .pull-right[] .center[ <img src="./figs/screenshot-abc.png" width="60%" /> ] --- class: center, middle, inverse ## Understanding comments on streaming chats. --- ### High Frequency .center[ <img src="./figs/freq-bind.png" width="80%" /> ] --- ### Much more toxic: .center[ <img src="./figs/proportion_toxicity.png" width="80%" /> ].footnote[Deep Learning Models from Google Perspective API] --- ### Contains Mostly Negative Primes .center[ <img src="./figs/topics_3by_cand.png" width="80%" /> ] --- ### Mostly Negative Polarity About the Candidates .center[ <img src="./figs/polarity_candidates.png" width="80%" /> ] --- class:inverse, middle, center ## Experimental Results --- ### Subjective Well-Being .center[ Facebook chat somewhat less informative, enjoyable, and engaging <img src="./figs/frequencyplot.png" width="80%" /> ] --- ### Feeling Thermomethers about the Candidates <img src="./figs/ftplot.png" width="80%" /> --- ### Content Effect .center[ <img src="./figs/ft_on_negative_comments.png" width="80%" /> ] --- ### Poll Performance <img src="./figs/pollplot.png" width="80%" /> --- ### Context Effect .center[ <img src="./figs/polls_on_positive_comments.png" width="80%" /> ] --- ## Results **Main Findings**: - Creates worse viewing experience. - Affects perceptions: - Disproportionately negatively affect certain candidates subject to toxic, negative comments. - Distort inferences about candidate viability. --- ## Next steps: contextual cues and majority illusion. .pull-left[ <img src="./figs/tweet.png" width="100%" /> ] .pull-right[ <img src="./figs/qttweet.png" width="100%" /> ] --- class:center, middle, inverse # Thank you!