Dec. 8th, 2015

In the policy statement where Trump attempts to justify his "no Muslims allowed" nonsense (which I'd assumed had to be being exaggerated by the press until I saw it for myself), he cites a survey of US muslims to justify the assertion that "there is great hatred towards Americans by large segments of the Muslim population".

Yay, a citation! I love it when people try to uphold stupid crap with citations, because it leads to some excellent object lessons in why you should always go to the data, not what some numbskull is telling you about the data. I'm home sick today and my brain has segfaulted on doing real work for a while, so instead let's play Fun Facts About This Survey!

Fun fact #1: 600 people were surveyed. 600! That's .0003% of the 2.77 million muslims living in the US (as of 2010, according to Pew Research). To put that in perspective, .0003% of the total US population is approximately equivalent to the number of people in the US who speak Hungarian at home (that's according to 2013 census data, and yes, I spent way too long trying to find a sufficiently obscure statistic to make this point).

(EDIT: I talked about this with a statistician friend, and it sounds like 600, while weirdly small, can be valid sample size for a population of this size depending on how the sample is selected, though it still raises questions about why so few people were polled. Anyway, mea culpa for the assumption. You never know what you don't know, and all that. But even if the sample is representative...)

Fun fact #2: Much is made in Trump's statement about the fact that out of that .0003% of the US muslim population, 51% said muslims in the US should "have the choice of being governed by shariah". And yet, for some reason he doesn't mention the most popular answers to questions like "how would you characterize shariah" ("a guide to the personal practice of islam"), and "if shariah and the US constitution were to conflict, which should win?" (the constitution), as well as a majority opinion that "it is up to the individual muslim to define shariah". There may still be a conversation to be had there about church and state, but it's not the one he's trying to have.

Basically, if you can't define your terms because doing so would sink your argument, you don't have an argument. And that's if you're working from statistically significant data, which isn't the case here either.

I know the fact that Trump is an idiot isn't likely to be news to anyone reading this, but citing a survey so obviously flawed and so obviously at odds with what he's saying really irks me because it just seems to highlight the fact that he doesn't expect people to actually read the damn thing. So read it, and get in the habit of reading, and seeking out if necessary, and critically analyzing, primary data all the time.

You never know what it's actually going to tell you.



October 2016

232425 26272829

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Jul. 21st, 2017 12:38 am
Powered by Dreamwidth Studios