Subcultural engagement for the 21st Century...
Barbelith is a new kind of community (find out more)...
You can login or register.

Underground > Head Shop >

Statistical literacy

No star here laces

11:38 / 13.05.03

We pay a lot of attention to numeracy and literacy in the west - it's a stated goal of the UK government to ensure all school leavers are literate and numerate.

The implicit assumption here is that literacy and numeracy are essential skills for modern life. We need to be able to read and write to negociate bureaucracy and earn a living. We need to be able to add up to run our financial affairs.

However I'm beginning to wonder about whether familiarity with some of the basic prinicples of statistics might be just as important. In the excellent The tyranny of numbers, David Boyle traces the use of statistics through history, leading to the inescapable conclusion that there will only be more measurement and more statistic in our future.

I'd argue that some understanding of statistics is now essential to properly comprehend science, economics and politics.

By basic statistics, to clarify, I mean understanding concepts such as mean, median and mode; probability; risk; the normal distribution and statistical significance.

Without this toolkit we cannot understand what is meant by statements like "red wine could be good for the heart", "hospitals have got better under a Labour government" and "women are more empathetic than men".

Yet statistics is currently considered a specialist ghetto. It's not even on the national curriculum, despite our mania for statistically measuring pretty much everything. Weather forecasters are not allowed to use probabilities on UK television because as a nation our understanding of such things is considered too weak to cope.

Anyway, this is probably of interest to precisely nobody, but am curious to what extent others think that stats is a cornerstone of modern life...

12:08 / 13.05.03

I think you are probably quite right, MCB. Statistics, and basic scientific literacy, are probably as essential as literacy of the more usual kind. I doubt you can really understand science without a minimal understanding for statistics. So, given that science and technology present lots of relevant ethical problems and solutions, a meaningful democracy requires a minimal understanding of science and therefore statistics.

Having said that, statistics are pretty slippery and liable to manipulation, deliberate or otherwise. So one possibly risks entrenching statistics as inviolable truth, by explaining the basics. I don't think this is really a danger, as I think one is more likely to appreciate the complexity of certain statements if one has a grasp of the basics.

13:25 / 13.05.03

I completely agree, misunderstanding statistics is one of my pet hates. It should be taught in school but the basics should be included in Maths and Science at high school anyway.
It's a part of the literacy/numeracy problem- if someone is innumerate there's no point trying to explain significance as a percentage etc.
Also (tangentially) accounting, bookkeeping, the mechanics of mortgages and bank loans etc. should all be taught in school. Many people cannot even grasp compound interest.

14:31 / 13.05.03

Oh how easy it is to lie with statistics, if we all understood statistics better, why, our respective gubments would have to be more honest. I think there's a reason statistics aren't taught in depth, woe to the common man that can't grasp the average.

17:13 / 13.05.03

Are you saying some people leave school without knowing how to calculate a mean? Really, i think its very uncommon for someone to leave not knowing how.

What i really think should be taught is bank stuff- interest rates, stuff about shares, isas, taxes etc. I briefly glossed over it in standard grade maths, but only just. Whereas i think i would have found 2 years of that more useful than 2 years of trigonometery.

As for someone who finds it difficult to grasp the mean- surely they're not gonna persue any kind of carreer in maths, so why start with useless trig stuff? If they dont learn bout taxes & mortgages in school, surely they'll struggle to learn about them elsewhere.

Oh and while we're at it, im all up for more cooking, and more languages. And possibly a bit less of bigoted, one sided R.E.

ONLY NICE THINGS

17:57 / 13.05.03

Mod hat: Unless people want a moderator to move that the title and abstract of this thread be changed, can we keep it to the usefulness of statistical literacy? Cheers.

Interested by Oh how easy it is to lie with statistics, if we all understood statistics better, why, our respective gubments would have to be more honest. I think there's a reason statistics aren't taught in depth, woe to the common man that can't grasp the average.

This seems to be saying that it is not in the best interests of our governments to allow the population to have statistical knowledge, since to do so would give them a better understanding of the political and economic processes (presuambly) and how they affect themselves, the potential voters.

So, doesn't this find the government in a bit of a bind? In order to maximise the wealth of the nation they need competent mathematicians to become economists, scientists, managers etc, and they need those firther down not to behave in a way that will damage the economy, so numeracy is a vital skill for the largest possible sector of the population to have. On the other hand the polity, precisely the people most in need of statistical knowledge, are also the people the government needs to keep ignorant of these things...

18:08 / 13.05.03

, statistics are pretty slippery and liable to manipulation, deliberate or otherwise.

Either D'Israeli or Twain said, "There are three kinds of lies: lies, damned lies, and statistics."

What I'm wondering is if statistics aren't taught with basic math because they can't be taught -- because they deal with a certain amount of error and ambiguity, and those hard things make our little brains to seize up.

(Note: I've never taken a statistics class, so if anyone wants to write some basic definitions of terms, I'd be all for it.)

18:16 / 13.05.03

I'm fairly sure that averages, probability and basic graph interpretation are taught as part of the Maths GCSE package in the UK these days. What they don't go into is how easy it is to produce biased and inaccurate stats. Given enough different analytical tools, and an imagination, you can make sets of data say pretty much anything you want. The actual statistical results just become more and more irrelevent as you learn about the mechanics of producing them but it gets interesting trying to figure out how on earth the 'statisticians' manage to lend credibility to their findings.

In my opinion, basic literacy is a good thing - you can choose what you read and be reasonably sure that the words taught to you at school still mean the same thing. Numeracy is also fine, 1+1 should always equal two. Knowledge of basic statistical terms and meanings is not essential, if anything it makes it easier to be convinced by misleading statistical statements.

Maybe what I'm trying to say is that good and healthy cynicism should be included in the curriculum instead. Learning how and why to question apparently respectable sources of information is a useful life skill which usually only gets covered in History lessons.

(Except for the very good point about the wine. I'm happy to run with that one until I get a better suggestion, perhaps involving chocolate.)

18:21 / 13.05.03

...one possibly risks entrenching statistics as inviolable truth, by explaining the basics. I don't think this is really a danger, as I think one is more likely to appreciate the complexity of certain statements if one has a grasp of the basics.

Oh, heavens yes. Even if one isn't steeped in statistics, just wondering what the margin of error is, or knowing that USA Today polls are self-selected and therefore 80% likely to be 20% as good as numbers I just made up, is half the battle.

Just this weekend, I was in a speech tournament in which I gave an editorial commentary (the "pretend you're ranting on the radio" event) on this very subject; specifically, we're easily manipulated by fear and the flashbulb effect because most people don't realize that a 9/11's worth of people die every month in the US from car accidents. (Not that tragedies aren't, um, tragic, but a tragic event is quite different from an imminent personal threat which necessitates immediate shock&awe.)

...they need those firther down not to behave in a way that will damage the economy, so numeracy is a vital skill...

Depending on what you mean by "further down"... is there really anything that the teeming masses do that have an effect, positive or negative, on the economy? I'd think most folks seek decent-paying jobs and really ought to be saving more, regardless of what they know or don't know about macroeconomics. So while I don't have much of a conspiratorial slant, I'd think the government gets more out of ignorance than numeracy.

18:30 / 13.05.03

In order to maximise the wealth of the nation they need competent mathematicians to become economists, scientists, managers etc, and they need those firther down not to behave in a way that will damage the economy, so numeracy is a vital skill for the largest possible sector of the population to have

I'm not sure I buy that. Specialists who need to be numerate can probably be given the relevant skills at a point where they decide to pursue the specialism. A "good" economic agent need not be particularly numerate, though it may depend on where you draw some kind of competance line. Moreover, statistics are quite a specialised brand of numeracy to do with the systematic evaluation of data, so are less tied in with personal gain than being able to add up, say.

Having said that, I don't see a big government conspiracy to keep us from finding about about how to use statistics. They just distort the statistics themselves and, to some extent, rely on a common dislike of numbers.

What I'm wondering is if statistics aren't taught with basic math because they can't be taught -- because they deal with a certain amount of error and ambiguity, and those hard things make our little brains to seize up.

Nah. If anything it is the other way round, it is the hard rigid stuff that people hate. Having said that, stats isn't all that removed from basic maths so the contrast doesn't really arise, IMO. There are, however, some pretty tricky issues when dealing with stats - what is random? how do you know things are independent? - though you'll never got told any of that, so its not that much of a problem.

19:45 / 13.05.03

True enough all, and I do agree perhaps the mathmatics of money would be the best focus. But there is a differance between grasping what an average and mean is and being able to regurgitate it onto a test and then forget, which in my limited expieriance is just what high school students do, at least here in the states. Brits may be getting a better education, and it wouldn't surprise me in the slightest, and the presentation of statistics to influence thought or to advertise is something different entirely. A commercial may say, "4 out of 5 dentists prefer this or that brand tooth paste", but the commercial doesn't say, "thats because we gave them loads of free samples."

21:34 / 13.05.03

Hermes you have it entirely. It is the thought process behind all statistical figures that is important, not the method. If you can detatch one from the other then you have a decent basis for discussion.
The issue with statistics is not how to produce them, but how to use them. I guess that the general public are not stupid and will consider the source of the figures, not the content.

Do you happen to know how many people will buy a toothpaste based on just statistical reccomendation? Now that would be more interesting.

I know that this may be considered threadrot because it isn't about statistical literacy. But it is about statistical influence on the general population and that's what's importance in everyday life.

15:07 / 14.05.03

In the last five, six years, the Royal College of Psychiatrists has included, as part of its membership exam, a 'Critical Reading' paper which focusses on statistics as used in medical research. The aim is to enable psychiatrists to evaluate any given research article in terms of bias, significance, etc., and thus distinguish findings of genuine value from slickly-presented crap.

I've always hated statistics, and it was a pisser of a paper to pass, but I've found my (comparitively low-level) critical reading skills transform the way I approach statements like 'Mozart makes you cleverer', 'the vast majority believe in dignity' or 'bread makes you gay'. It seems slightly retrograde that I didn't learn this stuff until my postgraduate life...

11:42 / 15.05.03

Innumeracy.com is a good resource. You can read all about how statistics can lie, the Monty Hall problem (a personal favourite) and even play the Monty Hall game. One thing about stats (and probability) is that we are pretty bad at it. So you take the following problem:

A test is developed to see whether a person has taken some illegal drug. This test gives an accurate result 95 per cent of the time and you know that 5 per cent of the population take the drug. A person gets a positive test result. What is the chance that they are a user of the drug?

And most everyone will get it wrong. of course, it isn't clear how realsitic the assumptions are, but still...

No star here laces

12:06 / 15.05.03

I've been thinking some more about this.

To what extent do you think greater use of statistics was necessitated by scientific and philosophical developments that placed greater and greater emphasis on the fallibility of human perception and logic?

I think postmodernism requires statistics, as it implies none of us are capable of reaching an objective conclusion due to innate bias. Plus we now have a scientific basis for critiquing human reasoning in a systematic manner, leading to a demand for external justification in decision-making.

Ergo statistics supplement our distorted perceptions in much the same way that books and the internet supplement our faulty memories and fact-retrieval systems.

12:16 / 15.05.03

I don't think that human fallibility is the only issue. We also have that people can't process large amounts of raw data. Even with small amounts of data, detecting trends is quite hard. Having said that, some reasonably objective measuring stick is needed. I don't think you neeed to invoke postmodernism to see that though.

No star here laces

12:46 / 15.05.03

I disagree. If you take the enlightenment schtick that rational human thought and technological progress will eventually produce a 'civilised' world, the requirement for statistics is far less, and the type of discourse is correspondingly far less reliant on statistics.

This worldview places abstract human thought at the top of the pyramid and enshrines the scientist archetype as an objective seeker after truth. The great scientific battles of the enlightenment were fought with mathematics and logic, not with statistics.

Marx, as possibly the most archetypal embodiment of this worldview used a rational sociological argument, without recourse to statistics, to create his model of history. Today it is absolutely unthinkable that anyone could possibly create a work of political and economic theory without vast amounts of statistical evidence.

Of course the relationship is not linear and directly causal, but more complex. However I am fairly convinced that statistics' rise to prominence and even necessity is linked to changes in the way we view the world and our own minds...

13:12 / 15.05.03

If you take the enlightenment schtick that rational human thought and technological progress will eventually produce a 'civilised' world, the requirement for statistics is far less, and the type of discourse is correspondingly far less reliant on statistics.

I disagree back. Lets leave aside your caricature of the "enlightenment schtick". Statistics didn't arise out of postmodernism, it arose out of trying to make rational decisions from data. Yeah, people are biased, but this is a pretty elementary point.

As far as I can tell, statistics were used fairly early. From here, there is a quote about statistics by Maxwell in 1871. Here it lists the formation of the Statistical Society of London in 1834. The title of one paper is "Statistics of the Metropolitan Commission in Lunacy. By Lieutenant Colonel Sykes.." 1840.

The widespread use of statistics in the social sciences may be more recent, but I think it is more clearly part of the rise of numeracy, than a postmodern realisation. Probably even more relevant is the advent of the computer, which makes statistical analyses possible.

No star here laces

13:33 / 15.05.03

Of course the computer has had a major effect. But statistical techniques and the ability to gather statistical volumes of data predate the actual usage of statistics in most fields, and particularly the use of statistics in argument and polemic which is a relatively modern phenomenon.

I'm not arguing that postmodernism is the sole cause of the stats boom, but only that these changes happening at the same time is not coincidental.

13:51 / 15.05.03

But statistics aren't objective, the statistic depends on how the study is put together, and this can often times be misleading, may even make a logical decision more difficult.

14:03 / 15.05.03

We are going to have to agree to disagree, 'laces. I think it is far too tenuous to go from postmodernism to an attempt at objectivity. Historically, I doubt you could find any connections at all and I'm not even sure if postmodernists would agree with you.