Clicked through to source¹, expecting it to say "...of streamers with more than a billion cumulative viewing hours" or something.
Nope.
https://snipboard.io/BCkmKf.jpg literally what it says on the tin. The only issue I can imagine is that the sample size is 550 respondents and study recruitment was biased ("hey wanna fill out a survey about online harassment?" imagine the odds you'll respond if you've been swatted vs. if you haven't had any negative experiences; and/or the link was shared among specific communities).
Online self-reporting is also prone to multiple responses from the same person, straight up prank entries, or honest misreadings of questions.
Edit: also, 13% says they "were exposed to" the topic of QAnon on the Internet in the past 12 months, 29% to anti-vax movement, etc. I guess the remaining three quarters just hasn't used the Internet outside of chatting with their family and looking up recipes? Interesting study on how not to run studies.
¹ PDF download (:/) https://www.adl.org/sites/default/files/pdfs/2023-06/Online-... page 23