Nutella Facebook page users prove: "Facebook Fueled Anti-Refugee Attacks in Germany"

Hah, got you. The complete NYT headline is

"Facebook Fueled Anti-Refugee Attacks in Germany, New Research Suggests"

Here is the study

Fanning the Flames of Hate: Social Media and Hate Crime

By Karsten Muller and Carlo Schwarz

Racy title indeed.

Here is the expert of striped taylor-made shirts and connoisseur of wines up to $ 50 at Slate.

He is not quite buying it and has some valid points. Unfortunately, Felix misses the forest for the trees. By reading this concocted, convoluted German study it gets obvious pretty early on that the result of the study was a given. The only question was, how do we back it up?

Here is Felix:

"The idea here is that by looking at how many people are active on the Nutella Facebook page, you can get a good indication of how active the broader population is on social media."

One might as well look at the Tour de France FB page or a pop star's page and could perhaps draw the same conclusions. Or as Tyler Cowen suggests, why not Zwetschgenkuchen?

"And areas in the top third of Nutella activity on Facebook do seem to have more attacks on refugees." 

Beware of sweet tooths!

Their methodology is weird. This is from their paper:

"Finally, we obtain data from Google trends for the overall interest in the search terms “Brexit”, “Trump” and “UEFA Euro 2016” in Germany to proxy for distracting news events. Google scales the weekly number of searches for these terms on a scale from 0 to 100, where 100 marks the week with the highest search interest in the preceding 5 years. Time series plots suggest that these measures are sound approximations for attention paid to Brexit, the Trump election, and the UEFA European Championship (one of the most widely followed sports events in Germany)."

And next it says:

"We start our analysis by documenting simple correlations between social media and attacks on refugees in Germany. The results in this section should be interpreted as purely suggestive and do not allow for causal inference. Nevertheless, we consider the findings insightful, because we are not aware of previous empirical evidence on the cross-sectional and time series relationships between social media and hate crime."

Tyler Cowen has the best take on this paper and the graph on his post says it all.

And here is XKCD.