social commentary

more statistical loathing

I have grown to dislike and even distrust the use of statistics, particularly in the form of percentages. (See here and here for examples of why this is so). Admittedly, some uses are rather comical. For instance, just the other day there was an advert on TV for mascara (Maxfactor, I think it was) and the voice-over made grand statements about women’s views on the product. Based on the VO’s claims, one would think that he was speaking for the whole of womankind. Hardly. At the bottom of the screen appeared the percentage of women supporting the claims made and number surveyed. The figures? 74% of 70 women!¹ No typo. Seven, zero. My first (cynical) thought was: “So, you [Maxfactor] basically got your own PR department to offer some opinions.”

As benign or even banal as this instance might be, it adheres to or relies on (and possibly even perpetuates) a rather malignant rhetorical ploy: shape opinion on the basis of persuasively strong claims supported by high percentages.² For example: “The majority of people (78%) believe _[insert hot-button issue here]_ should be permitted” or “…feel that _____ is unfair.” The underlying assumption appears to be: with language such as “majority” or “most people” and percentages exceeding 50%, we can make the issue appear to be prevailing and widespread, and if we can get people to believe the language and percentages, then we can shape public opinion in a particular direction. To remain in my cynicism, this usually means: the “majority” we’re documenting is the cultural norm, so you might want to get on board rather than fight against the “majority” view.

However, because I am that annoying person who asks, when confronted with percentages: ” ‘x’% of how many surveyed?”, and because the survey pool is hardly ever deep and wide enough or representative of the whole, I will neither be persuaded by the claims made (because they do not represent the whole they claim to) nor accept the data to be empirical evidence of public views/opinions (because it’s not). Moreover, I will not pretend that, say, 1500 people surveyed constitute a “majority” view and that I must accept their view, which is really a “minority” one (based on comparative figures), simply because they’re touted as the majority in a particular survey. In fact, I don’t have to accept anything simply because a few say so–and do so rather loudly (i.e. delusionally pretending to be the many).³ I accept things because they are worthy of acceptance, but that requires an entirely different (and more substantial) kind of conversation.

¹ I’m not sure if they’re doing this in the States (or anywhere else), but here in the UK it is now common practice in adverts to show both the percentage and the number of people surveyed.
² It is, therefore, no wonder that the survey-data is either tucked away at the end of the article or on a completely different site.
³ Current issues in American politics illustrate what happens when a few are allowed to shape the many, and do so on the assumption that the few are portrayed as more powerful than the many.

another one left behind

I have four e-mail addresses: personal, University, joint (i.e. my lovely wife and me), and Yahoo. The first is the one I prefer and use most often, the second is used out of necessity, the third is for random things, and the last is lovingly called, my “crap e-mail.” It’s the one I use when I have to sign up for things or give out a valid e-mail address. Thus, it’s the one that receives the most crap (or spam).

However, because it’s a Yahoo address this means I occasionally have to endure the Yahoo homepage, which is usually nothing more than a cesspool of banal dronings about superficialities masquerading as real news. Admittedly, every so often there is something important and newsworthy, but it’s typically late in the feed or sandwiched between useless slices of stale drivel. (Can you tell I have a slight disdain for Yahoo?)

In exceptional cases (and by that I mean their rarity, not their quality), I find something that is simply laughable–not because it is funny but because it is painfully pathetic. Case in point: this morning I was slapped in the face with this story. My angst is not with the story itself. I truly feel for the woman who was inadvertently harpooned and I do pray that she makes a full recovery. My beef, however, is with the ineptitude at keeping the details straight–let alone consistent. Or valid. Or right. Here’s what I mean.

Take a look at the way in which the story is advertised in the newsfeed:

Screen Shot 2013-05-09 at 09.01.07

Now, I ask you, dear reader: do you remember seeing anything in the original story about the women being “inches away from death”? I don’t. I do, however, remember seeing that “the harpoon came within 1cm of killing the woman.” In case you missed it:

Screen Shot 2013-05-09 at 09.01.12

Last time I checked, “1cm” is a far cry from “inches”–the plural form implying more than one inch. Come on, people; it’s basic education! You know, the stuff you learned when you were about 5. (Another child left behind).

I think what bugs me even more is that the newsfeed presents the “inches from death” as though it’s a legitimate quotation, but it’s not. No one in the article says that! Ever. So, not only are the technical details inconsistent (i.e. patently wrong) yet presented as though nothing’s really problematic, the quotation doesn’t even exist yet it is proffered as though it does. That’s just poor and painfully pathetic reporting.

I think I need more coffee….

a modicum of solace

Yesterday The Guardian provided the results of an American-based survey on various conspiracy theories. Interestingly, the headline focused only on one of the 26 questions: Obama as the Antichrist. Two things–one about the headline, and one about the survey–and then a third, which is less troubling.

With regard to the headline, it’s a bit misleading. According to the numbers in the results (the full break-down is here), only 13% believe Obama is the Antichrist. That’s hardly “one in four Americans.” The only way you can get close to a “one in four” charge (i.e. 25%) is if you lump the 13% from the “not sure” category, which is what Paul Harris does in the article. He does this on the basis that “not sure” = “possibly so” or “I could be convinced”. A bit shady on the method, but understandable. “One in four” sounds better and (slightly) more widespread than “one in seven(ish)”.

With regard to the survey, it too is a bit misleading. If we went on the title alone, we get the impression that 25% of (all) Americans believe Obama is the Antichrist. Given the population of the US (315,610,625, as of 8.30 this morning), that would mean something around 78,902,656.25. (Who is the .25 of a person!?). But that can’t be right. Outside of Garnier, who would survey that many people? Harris does explain that the survey involved only “a sample of American voters”. Okie dokie. According to this site, last year there were 146,311,000 registered voters in the US (a number that seems too clean for my taste, but no matter). So if we use that number, then, according to Harris’ “one in four” charge, that would mean 36,577,750 people surveyed believe Obama to be the Antichrist. (Thankfully, no fractions of people this time). But that can’t be right either.

Conveniently (or smartly), Harris leaves out the exact size of the sample, and he suspiciously leaves out any links to survey itself. Again, it sounds so much better and more–dare I say–condemning to say “one in four Americans” and let people assume that the number is huge. But what about facts? If you were to take the 12 seconds to do your own search and locate the survey in question (or just use the link I supplied–you’re welcome), you would see that the sample-size is . . . get ready for this . . . 1247. I didn’t leave out any numbers. 1,2,4,7. That’s all. Seriously, Harris: Asda’s got more stuff on sale.

Accordingly, if we use Harris’ bold figure of 25%, that would mean only 311.75 people believe Obama is the Antichrist. That’s not “one in four”. That’s like one in a million (I think; my maths are a little rusty this morning). But no one is going to care if Harris says that figure, so it’s no wonder that his misleadingly says, “one in four”. Things get uglier if we use the solid number of 13%, which would bring the total to a whopping 162.11 people. That’s just under one in every two million (again, I think). Seriously, Harris: you’re going to paint 1/4 of Americans with a brush admittedly used by only 162.11 people? Tsk! Tsk!

The third thing, and this is the modicum of solace: only 162.11 people admit to believing Obama to be the Antichrist. I don’t mean to sound crass, but it’s comforting to know that only 162.11 people have a faulty understanding of the Antichrist. If we follow what the Johannine Epistles say, the “antichrist” (ἀντιχριστος) is anyone who “denies the Father and the Son” (1Jn 2.22) and/or denies “Jesus Christ as coming in the flesh” (2Jn 1.7). While the Johannine Epistle seem to suggest a solitary figure–i.e. the Antichrist–appearing in the last days (cf. 1Jn 2.18; 4.3), nothing definite is said about him (or her–sorry, I have to be PC these days). However, since the language of the Epistles on this matter is apocalyptic, we can safely assume that this solitary figure is understood in apocalyptic terms–i.e. he (or she) ain’t human. And if he (or she) ain’t human, then Obama can’t be the Antichrist. Same goes for the new Pope–contra this guy–or any other person.

quote of the day (or why Santayana is still right)

Thus, eagerness on the part of the cities to receive marks of imperial favour, the desire of wealthy natives to attain the only form of public honour open to them, and the greed of the common people for sports and games, all combined to buttress the worship which the authorities had adopted as an instrument of government. But this was a way of expressing gratitude and admiration which the followers of Jesus could not take. The claim that was made on behalf of the emperor was irreconcilable with the sole right of Christ to the worship of men. Gradually it would come to the knowledge of the citizens that there was a sect in their midst that refused to join in the emperor-cultus. Astonishment would give place to anger. Every consideration that increased the enthusiasm of the citizens for the worship would make the attitude of the Christians more obnoxious in their eyes. The refusal would be construed into disloyalty; and both priests [of the emperor-cultus] and people would take every means in their power to overcome an obstinacy which would not only appear unreasonable and ungracious, but which might have the effect of making the city’s loyalty suspect in high quarters. The whole resources of the community would be employed to compel that conformity to the established usage which was not rendered voluntarily.

— J.T. Dean, The Book of Revelation (1915), 14

head-counts are representative, percentages are shady

For who knows how long, I’ve been suspect of nearly everything that comes with a percentage, especially surveys or polls. Why? Because surveys or polls use percentages in much the same way that Gorgias used rhetoric. They make grandiose claims about a multitude of concerns, with the hope that the towering figures will awe the masses. 67% of doctors agree on the cause of ‘X’, 95% of Europeans say ‘Y’ about the Euro-crisis, 78% of women prefer shampoo ‘Z’, and most recently: ‘In the US, 46% hold creationist view of human origins.’

My issue is not with the particular topic of discussion (i.e. human origins); my problem is with what the shady percentage conveys. Specifically, this figure suggests a reality that is neither consistent with the implication of its own claim nor with how things are or might be in reality. Quoi? We’ll start with the first half. As of today, at 11.09am (UK time) the US population stands at 313,658,703.* Thus, when the study says ‘46% hold creationist view of human origins’ it is implying that 144,283,003.38 Americans agree with that view.

However, and this is the second half of my problem, that 144,283,003.38 is not reality–i.e. it does not reflect the true number of those who in fact hold the creationist view. The survey indicates that only ‘a random sample of 1,012 adults, aged 18 and older, living in all 50 U.S. states and the District of Columbia’ (emphasis added). If my maths are right: 46% of that 1,012 sample is 465.52, and that figure is a pathetic 0.000148% of the US population. Call me cynical, but I fear that if the study did involve every single US citizen the number (or percentage) of those holding a creationist view of human origins would not reach the proposed 46%.

I can understand why surveys like this one opt for percentages instead or solid numbers. It sounds better to say 46%, which implies a significant portion of the country, than to say 465.52 people, who could fit into a small neighborhood in a medium-sized town. It made better sense for Gorgias to rhetorically woo the crowds into thinking he possessed great skill in a multitude of professions and avoid circumstances in which he would have to prove he had none.

* This figure is flexible for the simple fact it represents registered persons/births. Just for fun: by the time I finished this post (i.e. 11.51am), the population grew to: 313,658,902. Mazal Tov!