Miscalibrated Internet Receptor Stalks

Pew, these statistics stink.

With respect to the neo-tweet article on drone approval on the main page Huge gender gap in opinions about drone strikes I would like to point out a few reasons why this should be taken with the usual Khewra Mines annual salt production.


The questions were asked in "39 countries among 37,653 respondents". That works out at about 6 in every million of the population. The margins of error they report are:

±3.3-3.9: 12 countries |||| ±4.0-4.9: 26 countries |||| ±5.3 El Salvador, ±5.2 Chile, ±7.7 Turkey.

That means that for Turkey the range of possible actual figures is 15.4 percentage points, not percentage of the results, actual points of percentage. For the country with the least margin of error (well done chaps and chapesses, by Jove) the spread would be 6.6 percentage points. The vast majority of countries have a spread of between 8 and 9.6 percentage points.

So for those two the figures in the table above should read as:

Britain: Disapprove @ 47.7 - 54.3, approve @ 35.7 - 42.2

Turkey: Disapprove @ 74.3 - 89.7, approve @ minus 0.7 - 14.4

If you look into it there were four questions asked about drones in the Pew Global Attitudes Project for Spring 2013. Unfortunately it looks like three were only asked in Pakistan according to what can be seen on their website. The only one that was asked in other countries appears to be: "Do you approve or disapprove of the United States conducting missile strikes from pilotless aircraft called drones to target extremists in countries such as Pakistan, Yemen and Somalia?" This is somewhat different from the "% approval of drone strikes" given in the news reports or reblogs or retweets.


Here is the full report from Pew.

I hope you will forgive me droning on, but I feel very strongly about the media's responsibility in reporting press releases. I wish more would do journalism as I remember it.

Share This Story