The idle ramblings of a Jack of some trades, Master of none

Nov 5, 2007


Recently, I was struck by the relationship between opinion polls and the use of statistics in the popular culture and the news. In particular, this post is encouraged by Ravikiran Rao's recent article in his blog. Polls and statistics are manipulative and open to abuse. Even if you ignore the problems of sample selection and the propensity of the interviewee not to tell the whole truth, opinion polls are often driven by a very real bias - that the pollster has a hidden agenda. The appropriate framing of a sequence of questions can elicit from the respondent the desired response to the key issue.

A delightful parody of the methods applied by unscrupulous pollsters appears in that wonderful political comedy of the 1980s, Yes, Prime Minister. Jim Hacker (strangely prescient!) wants to cancel Trident, a hugely expensive ballistic missile system, and introduce conscription into the British armed forces. The bureaucrats, led by Humphrey Appleby, are appalled. Since the Prime Minister bases his decisions mainly on the results of opinion polls, Appleby shows Bernard Woolley, the PM's assistant, how polls can be conducted to obtain the desired result:

Humphrey: You know what happens: nice young lady comes up to you. Obviously you want to create a good impression, you don't want to look a fool, do you? So she starts asking you some questions: "Mr. Woolley, are you worried about the number of young people without jobs?"

Bernard: Yes

Humphrey: "Are you worried about the rise in crime among teenagers?"

Bernard: Yes

Humphrey: "Do you think there is a lack of discipline in our Comprehensive schools?"

Bernard: Yes

Humphrey: "Do you think young people welcome some authority and leadership in their lives?"

Bernard: Yes

Humphrey: "Do you think they respond to a challenge?"

Bernard: Yes

Humphrey: "Would you be in favour of reintroducing National Service?"

Bernard: Oh...well, I suppose I might be.

Humphrey: "Yes or no?"

Bernard: Yes

Humphrey: Of course you would, Bernard. After all you told her you can't say no to that. So they don't mention the first five questions and they publish the last one.

Bernard: Is that really what they do?

Humphrey: Well, not the reputable ones, no, but there aren't many of those. So alternatively the young lady can get the opposite result.

Bernard: How?

Humphrey: "Mr. Woolley, are you worried about the danger of war?"

Bernard: Yes

Humphrey: "Are you worried about the growth of armaments?"

Bernard: Yes

Humphrey: "Do you think there is a danger in giving young people guns and teaching them how to kill?"

Bernard: Yes

Humphrey: "Do you think it is wrong to force people to take up arms against their will?"

Bernard: Yes

Humphrey: "Would you oppose the reintroduction of National Service?"

Bernard: Yes

Humphrey: There you are, you see, Bernard. The perfect balanced sample.

With a suitable display of statistics or a graph, a presenter can hoodwink the unwashed public. Consider, for instance, this piece in the Scotsman. One in twenty children do not know where their food comes from! They have no idea we get eggs from hens! They think milk arises spontaneously in supermarkets! Surely this sounds the death-knell of the standards of our education! An entire industry of hand-wringers and pundits can make a living out of presenting the facts in this negative fashion.

But what exactly is the cause for panic? If one in twenty children are ignorant, then that means that 95% of them are not. That, however, is not news.

Manwhile, here is a question. Suppose you are the HR manager of a company, and you find that 40% of employees take sick-leave on Mondays or Fridays. Immediately you suspect that there is a large-scale abuse of the benefit, and that a large proportion of employees are shirkers, using the excuse of sickness to garner long weekends. But a little thought should reveal - if you assume reasonably that people tend to fall sick on random days with no predisposition to any one day - that this is no cause for alarm. Monday and Friday are two out of five working days, so it is reasonable that people do fall ill on these days 40% (two-fifths) of the time.

What would be a cause for concern is if 60% of employees took Fridays or Mondays off. Or if a given employee took her sick leave on Mondays or Fridays more than 40% of the time.

Unfortunately, even this level of understanding seems to elude the general public.


Post a Comment