Monday, October 28, 2013

Mop the SLOP

Most of what’s wrong with journalism can be traced to under-resourced newsrooms. But there are also bad practices, like pretending online polls are real. Such “polls” debase journalism and cheapen public debate.

News sites use online polls to drag eyes to their pages, and more page views means more ad revenue. Although that’s kind of crass, if that’s all they did with those fake “polls”, it wouldn’t be such a stain on journalism. The problem is that the sites/newspapers then report the poll results as if they’re actually valid.

The main problem with these “polls” ought to be obvious: Self-selection bias. This means that the people polled have selected themselves, and so, are almost never representative of the population generally. If these “polls” end up mirroring real, scientific polls, it’s purely accidental. All of which is why social scientists refer to these fake “polls” as, appropriately enough, SLOP: self-selecting opinion poll. [More on the subject of self-selection bias can be found through a Google search of scholarly articles]

The fact that these fake “polls” are wide open to manipulation means that people attempt to affect the outcome, to skew the results in the direction they want. The term for this is “freeping”, a word that’s a back formation from “Freeper”, the name used to describe the denizens of an extreme far right US website—one of the most popular rightwing sites on the Internet—called “Free [sic] Republic”. They worked out long ago that they could go to websites as a bloc and skew the results of a “poll” the way they wanted the results to go.

Eventually, the centre and left began to do this, too. For example, LGBT activists and supporters freep online “polls” about marriage equality. Popular gay blogger Joe.My.God. even has a subject tag for his posts urging readers to freep polls.

Most online “polls” have a rightwing bias most of the time. That’s because the rightwing is generally more passionate about issues, especially hot button issues like marriage equality, welfare reform, immigration, etc., that news sites deliberately poll on because they know it’ll bring eyes to their pages. One might also suggest that this is because the righwing is angrier than the left, but I couldn’t possibly comment. What’s obvious is that the rightwing tends to be more motivated.

So, what we have are polls that are biased, usually in a rightward direction, but that also can be easily manipulated toward either end of the political spectrum. These “polls” aren’t even remotely representative of the population generally. All of which means that these “polls” are utterly useless for anything other than pure entertainment. SLOP, indeed.

What makes these things so odious is that journalists cite these “polls” in real news stories as if they’re real, scientifically valid opinion polls (TVNZ’s news programmes frequently do this with their text “polls”, which are all SLOP as well). Generally, journalists don’t even inform people that the poll was a SLOP poll, leaving news consumers to think that the “polls” are real and valid when they’re never even remotely so.

When these “polls” skew toward one end of the spectrum or the other, like-minded political activists then tout the results of the fake “polls” as “proof” that the general public agrees with them. For example, in New Zealand’s marriage equality debate, our rightwing adversary constantly cited such fake “polls” as some sort of evidence that his side was winning when it clearly wasn’t the case. Of course, he also had a lot of trouble interpreting real, valid poll results, but the problem here was his touting fake “polls” as if they were real (and only the ones that allegedly supported his position, of course).

So the problem isn’t just the use of fake “polls” to drive pageview numbers up, it’s that the fake “polls” then become a topic in the news, and they become an aspect of debates on public policy, occasionally even shifting real public opinion (as measured by movement in real, scientifically valid opinion polls). This, in turn, sometimes means that journalists are creating the news, not merely reporting it, and that’s never okay.

The simplest solution would be to end all online “polls” conducted by mainstream news organisations. But, since they want eyes drawn to their websites, they won’t do that. At the very least—the barest minimum—news organisations MUST report that the polls are not scientifically valid. This point is non-negotiable, and ought to be a given for any journalist who cares about the ethics of journalism. To report these fake “polls” as if they’re real is a violation of the public trust and brings the journalist and his/her news organisation into disrepute.

The bottom line for me is simple: If a news organisation reports one of these fake online “polls” as if it was real, I then have no alternative but to assume that the rest of the story—and even perhaps everything the news organisation reports—is suspect and completely unreliable, too. This is a simple cause and effect situation: If they won’t be honest about these fake “polls”, how can anyone trust what they say about anything?

I can’t. Jounralists have a choice: Fix this or be assumed to be liars. Yes, it really is that simple.

No comments: