Assistant Professor @UTAustin @UTiSchool. @Stanford and @OhioState alum.
- Tweets672
- Following1.4K
- Followers999
- Likes4K
ALT Title page of journal article "The Consumption of Pink Slime Journalism: Who, What, When, Where, and Why?" by Ross Dahlke, Ryan C. Moore, Priyanjana Bengani, and Jeffrey T. Hancock, published in Digital Journalism. Abstract describes a study combining web browsing behavior of 1,238 American adults (21M website visits) during the 2020 election with a database of 1,627 pink slime outlets and 132,237 articles. Finds 3.7% of American adults (~9.5M people) visited pink slime sites; consumption differs from general misinformation and local news; topics produced and promoted differ from topics consumed; Google and Facebook were prominent referrers; news desert residents were not more likely to consume pink slime. Keywords: pink slime, digital trace data, audience analysis, local news, misinformation, journalism, 2020 election, imagined audiences.
ALT Figure 1: nine-panel grid showing proportion of participants exposed to Pink Slime (left column), Misinformation (middle), and Local News (right) websites, broken out by Age (top row), Media Diet Slant on a 1–10 liberal-to-conservative scale (middle row), and Race (bottom row). Each panel plots point estimates with confidence intervals. Pink slime exposure is low across all subgroups (mostly under 10%) with no strong demographic pattern. Misinformation exposure rises sharply with age, peaking around 50% for the 65+ group, is higher among more conservative media diets, and is lower among Black respondents (~17%). Local news exposure is fairly even across age and race, slightly higher for more liberal media diets.
ALT Figure 4: four U.S. maps in a 2x2 grid showing where individuals live (dots) and where the local news or pink slime sites they visited claim to be based (arrow tips). Top row shows local news visits; bottom row shows pink slime visits. Left column shows non-Trump supporters (blue); right column shows Trump supporters (red). Local news maps for both groups are densely covered with arrows across the country, indicating widespread visits. The two pink slime maps are nearly empty by comparison, with only a handful of arrows, and those arrows frequently span long distances across state lines—suggesting people typically visit pink slime sites that are not local to where they live. Alaska and Hawaii are not included.
ALT Table 1: Topic Model Results for 18 topics, with columns for whether the topic is programmatically generated, percent of articles produced, percent consumed by readers, percent of Facebook ads, and percent of tweets. Notable mismatches between production and consumption: Gas Prices is 21.5% of articles produced but 0% consumed; Iron and Metals is 18.4% produced, 0% consumed; Professional Licensing is 12.5% produced, 0% consumed. Conversely, Politics is just 1.4% of articles but 55.9% of consumption. Illinois Public Salary Data is 4.1% produced and 12.7% consumed. COVID and Election Conspiracies is 2.0% produced, 10.8% consumed, and 61.2% of Twitter promotion. Local News is 50.5% of Facebook ads and 33.7% of tweets but only 4.5% of articles and 5.9% consumed.
ALT Three texts from the same person. The first reads "Do you have advice on what I should put I the thread promoting commit on twitter?" The second reads "I could use some help." After those texts is the read receipt. A third, unread, message follows "Guys????"
