Anecdotes do not show the full picture

Hang on for a minute...we're trying to find some more stories you might like.


Email This Story






(John Macdougall/Pool/DPA/Abaca Press/TNS)

On Wednesday, the President of the United States tweeted a series of videos from Britain First, a far-right political party, ripping three confusing scenes out of context and purporting them to be part of a trend of violence by people of Muslim faith. If the clickbait headline didn’t give it away — “Muslim migrant beats up Dutch boy on crutches!” — the first video’s title is brazenly inaccurate, showing a Dutchborn citizen of an untold religion beating up another Dutch boy on crutches, with all dialogue spoken in Dutch.

The other video, “Muslim Destroys a Statue of Virgin Mary!” is an accurate title but neglects to mention that destroying idols is looked down upon by most non-extremist Muslim societies, especially since Mary is revered in the Islamic faith. And then there is the disturbing scene titled, “Islamist mob pushes teenage boy off roof and beats him to death!” It’s real, but shows a scene of the strife in post-revolution Egypt that has already been litigated.

But rather than settling on calling it fake news, let’s put a pin on how these videos are presented in a deceptive manner, and consider the logic behind tweeting them in the first place. Let’s pretend we did witness a Muslim migrant beat up a Dutch boy on crutches. Is that proof that we should fear Muslims?

Of course not. It’s pretty easy to find a video of any sort of person doing the same crime, often with the identity in the headline. It’s also quite feasible to put a camera in a country riddled with conflict and amplify a convenient detail. In a sense, this is a deeper problem than fake news: We are ripe for believing in anecdotal evidence that, regardless of how real or fake it is, impairs our ability to compare the magnitude of different phenomena.

To be clear, we are not discussing anecdotes that are being used as evidence for any discrete accusation. For example, if a video is usable as evidence for a crime, the perpetrators are accountable to the fullest extent of the law. We are talking anecdotes that fuel sweeping accusations: ‘we should fear Islam,’ ‘liberal activists are violent,’ ‘Trump voters are all racists’ and other convenient illustrations of groups in society.

The typical pattern is that a video will circulate, it will be shared with one of these accusations attached and that accusation will carry a narrative that consolidates in the reader’s brain with shocking imagery. Then rinse and repeat.

Even when recognizing the flaw in trusting a single anecdote to make sweeping judgements about society, the impulse is to find more anecdotes to be credible. Sometimes there are a hefty number of examples, but a motivated search for anecdotes does not lead to an informed conclusion. A dominance of examples telling one story are not necessarily indicative of society as a whole; the sample may be skewed toward societal biases, or someone with an agenda might be actively hunting for one sort of idea over the rest.

The goal of finding examples is not to prove a point but to support one. If one seeks to quantify some kind of point—that this ‘x factor’ leads to crime, for example—then the only way to find the truth is to find breakdowns of collected data. But that’s boring, right?

Let us return to the idea that Britain should treat Muslims with more scrutiny. According to the 2017 European Union Terrorism Situation & Trend Report, Islamist attacks decreased to 13 in 2016 (from 17 in 2015), and only six were credited to the Islamic State. This is nothing to overlook, but it is considerably less than the number of “ethno-nationalist and separatist extremists” that same year, which was 99. It also notes that assaults “targeting asylum seekers and ethnic minorities in general” are simultaneously on the rise. There are more videos purporting Muslim violence picked up by right-wing social media accounts like Britain First. That, however, means little more than how many ‘likes’ these accounts have accrued.

Does this revelation mean we should ignore evidence of a crime committed by a Muslim individual, if it is found? No. Does it mean we stop trying to fight terrorism? No, but it might mean Western countries are fighting Islamist terrorism disproportionately in comparison with other forms.

It’s true that even with statistics, we don’t have a full picture; there will always be bias in collection, a margin of error, problems of statistical integrity and conflicting explanations for those numbers. In fact, a single study or statistic can be misleading. But even so, they carry far more weight than 100 arbitrarily-obtained videos and allegations. Thanks to randomness, when used correctly, statistics are able to quantify and compare phenomena and set a standard of evidence for all to work upon.

The unfortunate truth, of course, is that the president wouldn’t find much capital in evidence. That is unlikely to change. But teaching ourselves to question the significance of what we are seeing, even if doing so challenges our biases, is of the utmost importance in post-truth America.

James Mazarakis is a Collegian columnist and can be reached at [email protected]