A Source is a Source of Course of Course

Any volunteers for today’s exemplary of crap journalism? Anyone? Mary Elizabeth Williams, is that your hand I see up? Alright then!

Rule one in journalism – at least if you’re even remotely interested in getting at the truth – is to always check your facts. Hey, I’m not just taking opportunistic potshots here. That actually is the first rule listed on the Society of Professional Journalists’ ethics page. The first section comes with a header that reads "Seek Truth and Report It," and it contains a list of things that "journalists should:" do. The first of those is:

Take responsibility for the accuracy of their work. Verify information before releasing it. Use original sources whenever possible.

Mary Elizabeth Williams seems to have gotten and read the memo, but she didn’t take it very seriously, at least, not if her recent Salon piece It officially sucks to be female on the Internet: 95 percent of online abuse is aimed at women is anything to go by.

You’ve already seen, from the headline alone, where I’m going with this, right? Does it strike you as even remotely plausible that "95% of online abuse is aimed at women?" Even remotely? Of course not. No one who’s been to an internet comments section recently can fail to have noticed that a lot of abuse gets directed at men as well. Does more of it get piled on women? I don’t know. I don’t find that – considerably milder – claim implausible (though I do wonder how you’d verify it? What "units of abuse" are we talking here, and what method of guessing a target’s sex are we talking, given that it’s relatively easy to hide one’s sex as a commenter online?). But 95% of internet abuse is aimed at women? Well, it might be true, but it’s so thoroughly at odds with my daily, lived experience on the internet – where abuse directed at men is also rampant – that I’d really like to see how we got at that number.

So, you’d think an article making a shocking claim like that would be heavy on the verification, right?

Yeah, well, this one not so much.

Here’s the bit where Williams gives us the "source" of that shocking number:

And the next time someone tries to tell you it’s not as bad as you say or it happens to everybody or go somewhere else if you can’t handle it, please repeat this line: “The UN estimates 95% of all aggressive and denigrating behavior in online spaces is aimed at women.” That’s the United Nations saying that — about 95 percent of the crap that goes down. (The BBC has since removed but not issued a correction for the original statistic.)


Salon is truly beyond shame. Let’s dive into the details of why this is unacceptable.

If you’re going to write an entire column based on the premise that 95% of abuse online is directed at women, you should at least be fairly certain it’s actually true. And see, "be fairly certain it’s actually true" doesn’t mean "find a webpage – even one – that claims it is." The BBC, like all news organizations everywhere, makes mistakes. It’s not as though anyone is unfamiliar with examples of otherwise-reputable news sources that went to press too soon and got caught with their pants down, if you’ll excuse the metaphor. So, no, actually, you’re not done with your homework just because you read something on the BBC. A responsible reporter would’ve, you know, gotten the BBC’s source and checked it out.

But even if you didn’t do the minimum that should be expected of anyone writing a column based around a single number in a major online publication and verify the source, you should still be bigger than trying to skate by on the technicality that "no formal retraction means it’s still technically true!" Honestly. I mean, honestly ask yourself why you think the BBC took that number down. Is it because the prose just didn’t flow at that point? Is it because they ran out of pixel space? No – they took it down because they couldn’t verify it. That they should have issued a formal correction but didn’t doesn’t give you a license to compound their error. It just makes you complicit in it for forwarding it on.

Since Williams was reduced to linking a Wayback Machine version of the BBC article, the live one having expunged her favorite number almost immediately after going online, we’re able to see the original quote and check the source for ourselves and it’s …


Because there wasn’t one.


Mary Elizabeth Williams is such a crap journalist that she based an entire column around a retracted number that was never sourced in the original to begin with. Wow. And Salon lets it stay there. Wow. Just stop to marvel.

But it’s true. If you follow the link she gives to the BBC original, the claim really is just that "the UN" said it. But we’re not given any link to a UN report, much less the name of a UN report or official, making that claim. Now there IS a UN source cited for some other claims given in infographics – something called the Association for Progressive Communication /UN Broadband Commission for Digital Development, 2015 – and there IS an executive summary report available online for that particular conference. But that summary amusingly fails to back up the claims made in the infographics it’s cited in – in fact it directly contradicts some of them. For example, the infographic on the BBC original claims that women are 27 times more likely than men to be victims of online abuse, but the number 27 never appears in the executive summary, much less in the context of gender disparities in the likelihood of suffering abuse online. As it happens, the report does make a claim about gender disparities in the likelihood of suffering abuse online, but the number it gives is that 57% of online abuse victims are women. Which is, to anyone who does their math outside of Hogwarts, nothing remotely approaching "27 times." If we use actual math, as opposed to the out-of-someones-ass variety, 57/43 = 1.33 – i.e. women are 1.33 times more likely than men to be victims of online abuse. 1.33. Not 27. Or, if I can do a little more elementary school math here without offending anyone, roughly 1/20th as often as the BBC was claiming.

It takes some serious gumption to cite a report right there in an infographic that directly contradicts it.

Mary Elizabeth Williams is the kind of person who will continue to recycle a number that even THAT low standard of evidence won’t stand by. Just think about that.

So, ladies and gentlemen of the left, if you’re ever wondering why many of us fail to take you seriously when you claim that there is an epidemic of online abuse against women, it’s not just because we spend a lot of time online every day and fail to observe any gender disparity in the kind of abuse that exists there (there seems to be plenty of rather vicious abuse directed at pretty much anyone). It’s also because we have several examples of people like the BBC and Mary Elizabeth Williams making up shocking statistics about it whole cloth. If you want to convince anyone of this stuff, you’ll have to start checking your numbers and citing credible sources. I mean for real, though. You really actually do. Citing unsourced retracted numbers just ends up having the opposite effect of convincing anyone you have the first clue you know what you’re talking about, you know?

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>