Thread by @Amelia_RH

Over a week ago the Wall Street Journal asked me to comment on the report that occasioned this story. Sadly none of my comments made the cut. But that's ok, I'm happy to repeat what I told WSJ (and more!) on twitter in this thread you're gonna want to read
wsj.com/articles/congr…

The report by Public Good Projects(funded by Nicholson foundation) claims "bots insert pro-vaping message in anti-smoking Twitter conversations or promote misinformation regarding [ecigs]"

Here's the TL;DR: PGP's report is TERRIBLE & their main conclusion is arguably fraudulent.

Basically, when WSJ asked me to comment, it was pretty clear they were looking for an opinion from "the other side." I don't criticize stuff I haven't read. To their credit, they let me read the report first, on the condition I didn't share or discuss it until after the story ran

So the story came out yesterday. So did PGP's press release (prnewswire.com/news-releases/…). Note the time stamp on the press release is a full 5 hours later than the time stamp on the WSJ article. It appears PGP offered them an exclusive.

The press release links to a copy of the report on PGP's website. It looks identical to the version I read/commented on last week. Except for one important difference. The version I saw had 6 additional pages. These 6 pages were the focus of my comments to WSJ. They're gone now.

The pages were an entire section of the report called "Bot Networks" Here's PGP's preamble to that section (cut from website version). Excuse the highlights, an artifact of my diligent effort to understand the study's significance
so as to comment intelligently to a newspaper lol

So this was followed the other 5 (now removed) pages. Each page was a self-contained case study on one of five exemplary "bots" (according to PGP) and their "botnets," that PGP selected to show "different types of content promoted by bots and the size and scope of their reach."

Can you guess where I'm going with this?

So PGP redacted the names on all these exemplary bots that got so much damn real-estate in the report they gave WSJ. Apparently they did this to protect privacy.. of bots... because this makes sense. But they did a shitty job & I was able to identify all of them in under 30 mins.

Aside: I have no way of knowing what PGP's actual motivation in redacting account names from the central "bot" in their major "bot nets" was, but I doubt it was to protect anyone's privacy, because the same was littered with screen shots with identifying info on other accounts.

So, I identified the 5 accounts & guess what? I'm 99.9% sure NONE of them are bots. My only uncertainty is with a commercial CBD account. It doesn't seem like a bot and I'd bet $$ it isn't, but I've never heard of until these sloppy researchers gave me a reason to look them up...

As for the other 4, I'm damn sure they're not bots because I know them. I've engaged with all of them for years. But you don't need to know these people to know they're obviously not bots. It's clear from their profile info, account activity, post content, follower ratios, etc.

I don't know what to do about the privacy of the accounts I ID'd BTW. Maybe DM them screenshots & they can share if they're comfortable? I don't think it's my place to "out" anyone.

But I do know one account won't mind if I use their "bot net" to exemplify how nuts this is...

We could make this a game. First person to correctly tag account 1 wins...

If you guessed @vaping360, you win. Meet Jim McDonald, V360 news editor/guy who quite publicly tweets from the @vaping360 account.

Jim isn't a bot. He let me sleep on his couch last week before I testified in Detroit, so I'm pretty sure he's a real dude..

https://twitter.com/whycherrywhy/status/1184286479796842496?s=20

@Vaping360 The thing that bugs me about this is I told WSJ all of this. Like - I told them immediately I know these accounts. I KNOW they aren't bots. I told them one of them was actually a news publication's account managed by an advocate who is a reputable journalist in the space.

And I told them that this was like... really serious, and I didn't know how it was possible to get this so wrong.

I also told them that PGP was completely opaque about what they did making it impossible to tell if they made an error or were just lying.

I gave them the IDs of all of the accounts so they could check my work. I encouraged them to contact these accounts themselves, especially Jim, who's a "public figure"

I gave them names of other social media researchers who aren't as flagrantly "pro vaping" as I probably seem.

I'm not bitter that I wasn't quoted (And to be fair the section that I found so alarmingly dishonest, as opposed to just incompetent, opaque or stupid, is now gone from the report).

I tried to make it clear to the journalist that what he showed me was worse than a flawed study.

Let's call this what it is: Dehumanizing propaganda from a special interest group, trying to de-platform real people who are using twitter to protest bans on harm reduction.

The propagandists succeed in triggering a congressional investigation on this tomorrow. That's fucked.

For posterity here's lobby group CTFK's tweet about the WSJ story. CTFK has pushed this absurd "bot" narrative from the get-go. I don't think they love having all these very real & very vocal critics debunking their BS

The 116 replies are all people telling them they're not bots

This myth of bot-dominated online vape advocacy is cut from whole cloth. But credible actors are spreading it & people who don't know better believe them. This is causing massive harm. Bad social science is just as dangerous as bad "hard" science.

(❤️the reply to this tweet btw)