This conversation is closed.

What sources do you get good statistics from?

Hans Rosling Tweeted a site which I really liked:
And then I came to think about his extraordinary website
And in Sweden we have the Statistics Sweden

So, what sources do you use for good statistics? And could you provide a link?

  • thumb
    Jul 26 2011: - Wikipedia has good stats
    - Wolfram alpha can produce some good data (
    - World health organisation bundles data from across the globe. This is often quite accurate (though some measures have broad confidence intervals)
    - A lot of country government websites release a lot of data
    - Google analytics is good
    - scientific research sometimes publish raw data

    There are a lot of projects around open data ( for example), I suggest to dig in there, you are likely to find a lot of sites.
    like this:

    As for statistics (numbers derived from data-sets): depends on how they are constructed... so if in doubt about the validity, try and find out how it was derived. That can give a good estimate of the validity and reliability.

    As for getting good statistics: I construct them myself ;-)
    • thumb
      Jul 29 2011: Thank you for sharing some great sources Christophe!

      These are sites that all should have a look at! ;)
  • thumb
    Jul 29 2011: Everyone is welcome to SHARE THEIR SOURCES! (Just a friendly reminder)
  • thumb
    Jul 28 2011: from my gut!
  • thumb
    Jul 28 2011: 85% of statistics cited in casual conversation are made up. ;)

    Statistics are usually just a means to an end. expert pollsters know how to frame the questions to get the answers they want to skew the results.
    • thumb
      Jul 28 2011: True, it is quite easy to sneak in leading questions, and answer tendencies in anything you wish to have certain answers from...

      Sadly, that happens.

      I abhor such practices, as it damages the good researchers and scientists out there... and it confuses the lay-men who can't distinguish between good and bad research, polls and surveys.
  • thumb
    Jul 28 2011: Statistics are meaningless.
    • thumb
      Jul 28 2011: Oh no you DIDN'T Scott!

      This is one statement that we will debate, elaborate!!
      • thumb
        Jul 28 2011: I'm talking from my perspective which is education and I can't see the point of collating averages and so on.

        You can identify trends to your hearts content but it is only ever about the stats, it never seems to be applicable or relevant to the people from which the stats were derived.

        Also, unless you survey everyone under the exact conditions and exactly the same frame of mind with an absolute control on the truth being told then they are just a talking point over the teacups at best.
      • thumb
        Jul 29 2011: You know me and my sweeping statements. Always a good way to kick an animated discussion off.

        On further thought, my gripe is really the misappropriation of statistics, by which I mean, in the case of education systems, student learning data is gathered for the purposes of the Ministry of Education and rarely for the purpose of enhancing actual classroom experience.

        This consumes a lot of valuable teacher/student time for the purposes of bureaucracy.
    • thumb
      Jul 28 2011: Scott: I assume you meant to say

      " I cant find meaning in statistics"?
      " I don't understand statistics sufficiently to derive the meaning from it"?

      Ok, this was ad hominem, I only wanted to indicate I disagree with your statement
      • thumb
        Jul 29 2011: As it is your profession, I can understand your frustration Christophe.
        To me saying that statistics are meaningless is almost equal to saying that there is no knowledge...
      • thumb
        Jul 29 2011: Christophe, a bit of both, to be honest.

        I was totally confounded by the subject Statistics at highschool so I dropped it and took History instead.