Why people study statistics

By skepticlawyer

It’s no secret around the place that I’m a bit of a cricket nerd, but my childhood fondness for cricket opened the door to a whole field of knowledge: that field was statistics, and to this day it’s the only area of mathematics in which I dare claim any competency. Batting averages. Bowling averages. Run rates. Net run-rates. Averages in first and second innings. Duckworth-Lewis. Etc. It’s nice when an otherwise useless sporting interest has such a useful payoff.

In the video below, watch Swedish statistician Hans Rosling combine the oldest form of regular statistical collation known to man (census data recording life expectancy, occupation and income, data-gathering concepts going back to Roman times) with the newest: some truly bedazzling computer technology.

And be amazed by the story he tells.

Once you’ve watched it once, go back and watch out for some individual details: China bouncing around like a rubber ball during the ‘Great Leap Backwards’, the effect of the 1918 ‘flu epidemic, HIV in some African countries but not others. And, above all, the Industrial Revolution, which has made monarchs of us all.

18 Comments

  1. Posted February 1, 2011 at 12:38 pm | Permalink

    Hans Rosling is good. We like Hans Rosling.

    A prominent mathematician has suggested that high school mathematics should be oriented towards teaching people basic statistics, as far more generally life useful than, for example, calculus. Gets my vote.

    BTW SL and other lurking classicists may appreciate this interview with Deidre McCloskey, a particularly historically minded economist who argues that dignity, speech and rhetoric matter in economic life. She is writing about why the Industrial Revolution happens and what it has meant, so it is even on topic.

  2. Patrick
    Posted February 1, 2011 at 1:37 pm | Permalink

    I love that one.

    I like the bit about chimpanzees (is that in this or only the 2006 TED talk version?) and his students, too!

  3. conrad
    Posted February 1, 2011 at 3:53 pm | Permalink

    That was wonderful

  4. Posted February 1, 2011 at 11:09 pm | Permalink

    This is even better than his TED demonstration; there is a perception that only things like chaos theory produce true mathematical beauty. This shows how statistics can be just as beautiful.

  5. desipis
    Posted February 2, 2011 at 9:18 am | Permalink

    I think I’ve seen the presentation before, but it is an impressive visualisation of data. There’s always plenty of historical events to try and spot when watching it again. I’d think there’d be an element of chaos within the data being statistically visualised that could the basis of the beauty.

    …high school mathematics should be oriented towards teaching people basic statistics, as far more generally life useful than, for example, calculus.

    The problem with statistics is it’s useless if you don’t understand the nature of the underlying data. For example attempting to analyse population data without understanding the relationship between exponential rates of change and total populations (calculus). I’m not exactly sure going beyond the existing high school curriculum would be of benefit without a further broad education.

  6. Posted February 2, 2011 at 11:07 am | Permalink

    [email protected] And teaching folk to be careful about data is not a useful educational aim? We are talking about pretty basic statistics, not the more complex levels.

  7. Patrick
    Posted February 2, 2011 at 11:08 am | Permalink

    The statistics part of my high school mathematics was much harder than the calculus! Maybe this was because my teacher was a former Russian nuclear physicist whose strength was trajectories and the likes…

  8. Hugivza
    Posted February 2, 2011 at 11:33 am | Permalink

    [email protected] Calabi yau manifolds, which are used in string theory analysis, have a beauty which is comparable to Mandelbrot sets. We owe much to modern computing capabilities.

  9. desipis
    Posted February 2, 2011 at 12:20 pm | Permalink

    Lorenzo,

    I agree that the concepts you’re talking about are important, but I’m not sure the type of qualitative data analysis that would cover “being careful about data” would come under the syllabus heading “statistics”. To me “statistics” is the process of putting numbers in and getting numbers out, while dealing with whether those numbers have meaning is something separate. Similar to the distinction between calculus (differentiation, integration) and physics (velocity, acceleration, force, kinetic energy, etc).

  10. kvd
    Posted February 2, 2011 at 12:50 pm | Permalink

    Well, not to distract from desipis/Lorenzo, but these are a couple of things that occurred to me while watching this fascinating show:

    – in 1942 one of the euro balls bounced below 25 life expectancy
    – $400 to $4000 in 200 years is about 1.12% interest accumulating
    – $400 as an average income seems quite high for 1810 abouts
    – in “real” terms I’d expect $400 in 1810 to be worth a hell of a lot nowadays
    – could you conclude, with regard to the above two “facts”, the cost of living now is in fact much less than it was back then?
    – Rosling equates vertical axis years of life to “health”. I’d need to think about that.

    But overall a terrific use of graphic data display.

  11. Posted February 3, 2011 at 1:34 pm | Permalink

    [email protected] Having worked as a statistics consultant, your division scares me. The point the mathematician I cited was making (sorry, I don’t have a reference) was about usefulness in life and as citizens. That would very much involve ‘what does this mean?’ questions.

    [email protected] Your points in order
    (1) I’m guessing Poland and the Holocaust
    (2) That is real, not nominal, interest.
    (3) Living on an average of $400 per year ($1600 for family of four) seems pretty poor to me. Average income growth was very, very flat for a very, very long time. An Arnhem land aborigine probably had a better standard of living than a Georgian England agricultural labourer: ignoring relative vulnerability to disease.
    (4) Not quite sure what you mean: that is $400 in today’s money.
    (5) Yes, in terms of the time it takes to earn x worth of something. Ted Ridley has a nice example with light. As Ridley says, prosperity is the time it takes you to earn x. To earn an hour of light now takes about 0.5 seconds at average wages. In 1800, it took 6 hours at average wages to earn a candle to burn for an hour.
    (6) As a distillation, life expectancy is an excellent proxy for health. It is hard to be healthy if you are dead.

  12. kvd
    Posted February 3, 2011 at 2:17 pm | Permalink

    [email protected] thanks for accepting my musings as “real” not nominal. I find the graphical representation fascinating – wish I’d had the vision to incorporate such stuff back in the days… My two remaining complaints are that the presenter talked too fast – thus my mistake in his “health and wealth” comment produced my thought that it was health vs earnings; and secondly – to drag out a specific sub region is impressive, but it would need to situate itself within the grid at proper position to be really effective.

    As to your comments – and forgive length

    1) I guessed Poland as well, for same reasons, but that was nothing to do with health per se.
    2) Yes
    3) I now understand the $400 is “wealth” not income. And it still seems very high to me.
    4) Yes you are right. My misunderstanding at time of posting.
    5) Thanks for that link – really interesting. Must admit I surprised myself by coming to that tentative conclusion. Another reason to like this blog!
    6) Life expectancy to the extent it is graphed taking into account wars is not representative of “health”. Minor quibble, but I’d have preferred the graph “ignore” the World Wars for this purpose. And I can’t believe I just wrote that.

  13. Posted February 3, 2011 at 9:11 pm | Permalink

    [email protected] Correction of myself: I meant Matt Ridley on TED, not Ted Ridley.

    Personally, I am cool with the notion that the Holocaust and the World Wars were unhealthy.

  14. Mel
    Posted February 4, 2011 at 4:08 pm | Permalink

    I think Rosling is being seriously overoptimistic in concluding that we can all make it into the top right hand corner. I doubt the planet has the resources to allow the 9.5 billion people projected for 2050 to all consume like folk in rich countries.

  15. Posted February 4, 2011 at 7:17 pm | Permalink

    [email protected] On current technology, possibly not. However, there is no reason to presume that current technology is the limit.

  16. Mel
    Posted February 4, 2011 at 9:08 pm | Permalink

    I don’t Lorenzo but as Prof Quiggin and others have pointed out, technology in almost all areas other than IT and communications has only improved by small increments over the past few decades.

    See here: http://johnquiggin.com/index.php/archives/2003/10/18/slowing-down/

    And the whole shebang relies on ever diminishing fossil fuels.

    The end is nigh I tells ya’.

  17. Posted February 4, 2011 at 10:16 pm | Permalink

    Yes… like to try drawing some lines through the scatter, but it /does/ look like the angle was once 45 degrees, but flattening off more recently.

    Saw another interesting stats and health – over much shorter timescales – mortality based on change of temperature from one day to the next (more than 3 degree change up or down makes a significant difference to mortality rates). Some of the Brisbane v LosAngeles differences puzzle me.

    But Melbourne wasn’t in the study – given that the weather changes every 5 minutes, mean daily temperatures aren’t meaningful I guess.

  18. Movius
    Posted February 5, 2011 at 12:30 am | Permalink

    @18 “technology in almost all areas other than IT and communications has only improved by small increments over the past few decades.”

    With respect, that is complete bullshit. Of course, some disciplines are advancing quicker (and more obviously) than others, but in general, technological and scientific advancement , while mostly incremental, is massive.

    To illustrate my point with a personal anecdote that proves nothing: I recently rewatched Cosmos on a DVD set. It is still an amazing document about the wonders of science, nature, blah blah, etc. and it stands up well even now. (the stats video reminds me a lot of Sagan explaining a 4th spatial dimension)

    At the end of many episodes was a 1990 update by Carl Sagan discussing advances in the fields discussed in the episode. One of which was the imaging of Betelgeuse in optical light.

    In 1980 there was not a single optical light image of the disc of a star other than the sun. Yet in 2011, thousands of exoplanets orbitting other stars have been discovered.

    Other fields have advanced similarly. 10 years added to the life of AIDS sufferers (medicine,) half a tonne removed from the typical car weight (manufacturing,) Universal flu vaccine conceivably within reach (medicine again,) The pyramid has been completely inverted (Association Football tactics,) and so on.

    Even the old standard “The lightbulb hasn’t changed in a million years due to freemasons and the bildeburg group!” Doesn’t hold true anymore.

    As such I hold no grave fears for humanity’s future that involve an inability for technology to advance.

    Though I can imagine (probably largely localised,) issues that involve a lack of understanding of science. Be it vaccine preventable disease, famine due to GM paranoia or even climate-change denial and similar catastrophes.

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*