How to Write with Evidence in a Time of Fake News

We’re getting increasingly used to the idea that the information around us may not be reliable.

It’s long been normal to assume that many of the pictures we see are faked, whether that’s models with implausible figures and impossibly flawless skin, or a hilarious image going viral of something that never actually happened. But in the past year, we’ve started to get to grips with the idea that not only can we not believe everything that purports to be a real photo, we can’t believe everything that purports to be a real news story either. It might be fake news deliberately invented for kicks, or a story essentially based on fact that has been distorted through a partisan lens so much that it’s lost all resemblance to reality.
If you’re writing in any area that requires online research, this can be extremely challenging to navigate, even with the best of intentions. Even news outlets with hundreds of staff and a dedication to reporting only the truth have been caught out by fake news – such as the Washington Post in January this year, with a fake story about Russian hacking. So what can you, as one individual, do to keep your own writing reliable and fake news-free? Here are our top tips.

1. Consider your sources very carefully

The obvious first step to take when tackling fake news is to take extra care in selecting your sources. As the Washington Post example above shows, it’s not even always enough to ensure that you’ve chosen a source that it generally reliable. Broadsheet newspapers can be fooled just as easily as the rest of us. Instead, do a bit of research not just into the website but into the author themselves. If you’re reading a science article, was it written by someone with a science background, or is there at least some commentary included from someone with expertise in the area? If it’s written by a journalist without a background in that particular area, is it at least a topic that they usually cover? As this demonstrates, much of what you can use to judge a source is just common sense.
You should not only look at the reliability of one source, but at the quantity of people covering the topic – if it’s only being covered in one place, that’s not a good sign. Also note whether you can find different information from different sources; if there’s near-complete overlap, it could imply that all of your sources rewrote the same press release. While this would suggest that they trust the press release, you should regard this more critically than if there’s evidence of independent reporting.

2. If something sounds too good to be true, it probably is

The main reason fake news attracts clicks and spreads so rapidly is because we’re all inclined not to look too closely at information that supports our own political positions. Most of the time, we decide to believe something that favours our own side before we’ve even consciously realised that we’re doing it. It’s not a case of seeing something we like the sound of, briefly considering its accuracy, and then deciding not to investigate further; we seldom get as far as the stage of critically assessing it if it’s something we’d instinctively agree with.
This means that in order to filter out fake news from our own side, we need to get into the habit of questioning everything that we see, and being extra critical of information that comes from our own ‘side’ of the debate, whatever that may be. If you come across the perfect ‘fact’ for the argument that you’re writing, that’s the moment in which you should consider whether it’s perfect because someone made it up with your views in mind. The real world is messy and lacking in heroes and villains; any stories that seem to produce them, especially if they align with your thoughts too neatly, should be treated with particular care.

3. Draw a clear line between fake news, partisan reporting, satire and kayfabe

It’s useful to distinguish between the different sources of inaccurate news when you’re working out what you can rely on; sometimes fake news gets used as an umbrella term, covering content that isn’t fake news in the classic sense at all.
‘Proper’ fake news is the content that’s been created by someone who knows full well that what they’re writing isn’t true. They’ve written it like a work of fiction, carefully designed so that people will fall for it and repeat it, and that will translate into advertising revenue for the sites in question. It’s created cynically, with a profit motive.

That can be hard to distinguish from extremely partisan reporting; they can look identical in practice. But where they differ is in motive. A fake news site will invent a poll that never happened, showing something that a particular political group wish to believe (e.g. that there’s a strong correlation between people in Scotland who support independence and support gay marriage, for instance). An extremely partisan site might run the same poll among their readership and get the same results, and repost as if it was representative and well-researched. You can’t use these sites as evidence of the facts, but you can certainly use them as evidence of what their supporters believe (in this example, the desire to believe that people who opposed independence are also more likely to be homophobic).
Satire, like the Onion, is also pure fiction, but written by people who never intended anyone to believe it to be true. When people do believe it and repost it as truth, then it can look a lot like fake news – but it’s less evidence of our news system becoming corrupted, and more evidence of people failing to check their sources; most satirical news sites are not subtle about being satire.
Finally, a type of ‘fake news’ that gets ignored is what could be called ‘kayfabe’ (a term borrowed from wrestling). This is where a story is fake, or at least exaggerated, and everyone knows it, but people still repeat it because they (consciously) like the idea it might be true. The story of Michael Heseltine killing his mother’s alsatian falls into that category – it wasn’t entirely true, most people repeating it knew it wasn’t entirely true, but they were too entertained by it not to. Kayfabe isn’t satire – the stories tend to be distorted, not invented – but it also isn’t classic fake news because everyone involved in perpetuating the story assumes the reader is in on the joke.  

4. Track facts through the internet

Genuine stories will have something of a paper trail, if you bother to follow it. We can take the example above of a misleading poll to demonstrate this. Let’s say you read about such a poll in a mainstream newspaper like the Guardian. If so, you’d expect the story to give details of the polling company – whether it was YouGov, Ipsos-MORI, ComRes or any other well-known pollsters. Once you have that, you can investigate the polling company’s website and see if you can get further details of the poll itself.

If it’s a reputable polling company, then you can usually assume the sample size will be sufficient, but it’s worth taking a look at the questions asked. What’s reported by the newspapers might not accurately reflect what the newspapers reported – for instance, a majority in favour of a well-developed plan of citizen service, including a military option but also voluntary work in organisations such as charities or old people’s homes, might be reported as a majority in favour of military service generally.
Many stories that aren’t exclusives should have a similar trail that you can follow. You might be able to chase them back to a press release, or to a scientific paper, and then you can look for yourself to see how much the reporting you’ve read resembles the original. In the same way, if you trace a fact through the internet and you never find any further detail, consider the possibility that it’s a lie that’s simply been repeated verbatim lots of times.

5. Don’t believe anything that’s skimpy with its evidence

If you have a case of fake news on your hands, the above approach should help you to work it out; when you go to look for evidence, you won’t find any. However, it’s not always the case that a lack of a trail corresponds to fake news, such as in the case of exclusives. But one indicator both of proper fake news and of extreme partisanship is a story that avoids referring to evidence.

Let’s take the poll example again. In a good, accurate article, you might expect to find exact percentages, sample sizes, the date that the poll was conducted and the method used – the latter being important because sometimes different methods throw up different results. This was the case in the EU referendum, when telephone polling consistently had a higher percentage for Remain than online polling. In an article that’s trying to disguise its lack of evidence, you’ll read things like ‘a majority’ or even ‘a clear majority’ without any indication of how large this majority is. Information such as the margin of error is also likely to be missing.
Real news wants to give you this information, because information provision is what the news is all about. Fake news of whichever kind doesn’t care about that.

6. Get used to looking for loopholes

Lots of fake news stories are what’s sometimes called ‘trivially true’. For example, it’s well established that eating healthily helps reduce your risk of developing dementia in old age. It’s also well established that fruit and vegetables are an important part of a healthy diet. So it’s entirely true to say that eating a tomato every day, if you weren’t doing previously, will reduce your risk of developing dementia, at least a little. Nonetheless, a headline that said “Eat a tomato a day to cut your risk of dementia” is decidedly misleading. It’s trivially true, but it makes tomato-eating sound like some kind of wonder drug, while in fact it’s nothing of the sort; you could replace the tomato in that statement with any fruit or veg and it would still be true.

Plenty of sites use this kind of loophole to produce a story; health stories are a particularly good example because there are plenty of things that can change your risk of getting a disease (which makes for a good headline) while not necessarily being meaningful. But there are plenty of other loopholes that writers can use to disguise the fact that their story isn’t as significant as it sounds.
For instance, there’s the technique of reporting on a normal event as if it’s in some way significant. You could easily imagine a story that says something like, “Labour agree to Tory demands to withdraw MPs from crucial vote” – lots of shock value and clickbait potential there. But in fact, vote-pairing – where one MP agrees not to vote to balance an MP from the opposing side who cannot vote due to illness or travel – is a completely normal part of the parliamentary process. It’s just one that most people aren’t aware of, so is ripe for a ‘trivially true’ fake news story.

7. Use the fact-checking tools available to you

It’s important to remember that you’re not alone in the battle against fake news. There are plenty of fact-checking tools out there to help. You can start with the classic fact-checkers such as Snopes, but it’s worth doing some DIY work as well – for instance, Google’s ‘search by image’ option is brilliant if you’d like to check whether the photo illustrating an article is of what it’s supposed to be of, or if it’s a repurposed stock image.
For the most technologically advanced, you can even download browser extensions that will flag fake news sites or identify dodgy Facebook posts – though you then have to trust that these extensions were built by a neutral provider in the first place. Inevitably, there are now fake fact-checking sites, that debunk stories that are true and elevate stories that are false. Sorting fact from fiction is hard work in 2017, and seems set only to get worse.

Image credits: i wish this were fake news; newspaper; alarm bell; comic books; alsatian; magnifying glass; statistics; tomato; truth sign.