Meta & Children's Online SafetyChat: We proved Daisy does not improve well-being
Thread View
quoted in an excerpt in the archive.
Monday 10 August 2020
06:34Dr. Kristin Shay HendrixI haven't had a chance to read it yet unfortunately.
We had an extensive research and analytics plan compiled already.
How are you staffing Daisy? Is Melissa tackling the work?
I think we should both speak up during the review about our concerns.
07:04Dr. Kristin Shay HendrixOK just read it. I'm concerned about the plan they put forth. We did design a survey to dig into teens attitudes (that was part of the study we'd presented to Schultz et al.). I do not think we should study only teens, so this proposal gives me strong pause.
07:04Roberto MedriI am exactly where you are
07:05Roberto MedriAlso zero researchers worked on this deck
07:05Dr. Kristin Shay HendrixSecond, they talk about a decision by leadership in Q3 for IG to pursue Purple Daisy independently which suggests de-coupling from FB, but then they have a question in the discussion points asking if IG would ship independently from FB...so I'm concerned about that.
07:05Roberto MedriThere are zero details on the test
07:05Roberto MedriIt's a "come on let's ship this" deck
07:09Dr. Kristin Shay HendrixTotally. The teens findings were very, very concerning, though. So I want to do this right and not screw it up.
07:10Dr. Kristin Shay HendrixIt was a pretty controversial/contentious Zuck review with Adam and Fidji not being in alignment.
07:10Roberto MedriAnd this is still true
07:10Dr. Kristin Shay HendrixAnd Schultz was as diplomatic as he possibly could be knowing that all of this was not handled well.
07:10Dr. Kristin Shay HendrixAgree Definitely still true.
07:10Roberto MedriFidji completely parked it FB-SENY-00000353
07:11Dr. Kristin Shay HendrixThe decision to test on both FB again was largely comms motivated.
07:11Roberto MedriSchultz told me: "We proved Daisy does not improve well-being"
07:11Dr. Kristin Shay HendrixCorrect all of our effect sizes were pretty small (though, notably larger than other effect sizes we see in research).
07:11Dr. Kristin Shay HendrixInteresting Moira Burke in CDS has HUGELY pushed for Daisy to ship.
07:12Dr. Kristin Shay HendrixIt actually caused a lot of thrash at the last minute.
07:12Roberto MedriI'm not surprised
07:12Dr. Kristin Shay HendrixBecause she was having convos with leadership that those working on Daisy weren't a part of.
07:12Roberto MedriWhat bothers me is everyone seems more motivated by incentives
07:12Dr. Kristin Shay HendrixI really appreciate and respect Moira's expertise. However, not aligning with myself and the FB Res Dir was really thrashy. And Schultz was not thrilled.
07:13Roberto MedriThan by evidence
07:13Dr. Kristin Shay HendrixTotally agree.
07:13Roberto MedriCDS has the incentive "publish an academic paper on this"
07:13Dr. Kristin Shay HendrixWe flagged this to Cox in our researcher Q&A with him how leads make decisions not based on evidence or people-first values.
07:13Roberto Medrino finding, no paper
07:13Dr. Kristin Shay HendrixWell, hopefully any peer reviewer who knows anything would catch all the red flags on the bungled approaches we've taken.
07:14Dr. Kristin Shay HendrixMaybe not...
07:14Dr. Kristin Shay HendrixI've had awful peer reviewers before. Lol.
07:14Roberto MedriI like to believe they would :)
07:14Dr. Kristin Shay HendrixKyle Andrews who was the quant UXR is pretty adamant about not working on this again.
FB-SENY-00000354
07:14Dr. Kristin Shay HendrixThe qual UXR who worked on it goes on parental leave in a couple of weeks.
07:14Roberto Medrithis should be made clear
07:15Roberto MedriThat the project was trashy before
07:15Dr. Kristin Shay HendrixSo I'm either going to pull Kyle off of other PO work or ask
07:15Roberto MedriI had a similar story in ds
07:15Dr. Kristin Shay HendrixYeah, I felt bad for and Aaron.
07:15Roberto MedriPausing feed recs for this would be Kafkian. And yet here we are
07:16Roberto MedriI told vishal yesterday night to pause Feed Recs for a bit to work on Daisy.
07:16Dr. Kristin Shay HendrixWe were in an awful position of Adam having publicly talked about this, all but promised we'd do this, and then trying to be as objective as possible.
07:16Roberto Medri"I don't think this test will tell us much more than the previous one, and I don't think we have research staffed"
07:17Roberto MedriHe told me: "thanks, that was not apparent from the document"
07:17Roberto Medri"EXACTLY"
07:21Dr. Kristin Shay HendrixI pinged him as well with details.
07:22Roberto MedriI think he's in a difficult position because of what Adam committed to publicly
07:22Roberto MedriI feel for him
07:22Dr. Kristin Shay HendrixSame
12:26Roberto MedriTHANK YOU
12:27Roberto Medrithank you thank you thank you
12:29Dr. Kristin Shay HendrixThe teen model call out is fair.
FB-SENY-00000355
12:30Roberto Medridefinitely
12:38Dr. Kristin Shay HendrixAnd we were barely powered enough to detect the highly-active teens effect we did.
12:39Roberto MedriI honestly would hug you if it weren't for covid
12:39Roberto Medrimaybe we've got one less thing to do
12:40Dr. Kristin Shay HendrixShared an image