Skip to main content

View Diary: GunFAIL VI (282 comments)

Comment Preferences

  •  I think I understand what you are getting at (4+ / 0-)

    but your reasoning is based in a logical and mathematical fallacy. If one wants to talk about something being "statistically significant", then that term has a very specific meaning.  As as matter of fact, the relevant numbers begin to get into the Law of Large Numbers territory, which is a whole 'nother concept.  

    Many arguments and debates can be constructed without using statistics incorrectly.  It is one of my pet peeves.

    The general who wins the battle makes many calculations in his temple before the battle is fought. The general who loses makes but few calculations beforehand. - Sun Tzu

    by Otteray Scribe on Fri Feb 22, 2013 at 05:55:41 PM PST

    [ Parent ]

    •  About statistics (1+ / 0-)
      Recommended by:
      coquiero

      I understand a few things about variables and large numbers.

      Such as why collision of earth with an asteroid is an absolute  certainty at some point (but statistically very unlikely this year or next or even in the next couple millenia).

      Or that every year we have a 1% chance of a 100 year storm, even if we had a 100 year storm last year.

      I hear you about misuse of math, some people are reaching for factual support and interchange "rare" and "statistically insignificant" when they try to claim something is "unimportant" - not realizing that statistical significance has nothing to do with importance in a society's legitimate policy decisions.

      The number of successful assassinations is exceedingly rare but they are incredibly important events.

      My objection to your example, is that there are also people who, as you know, intentionally choose large or small reference numbers to support a claim about importance. It's easy to report a scary sounding rise in something, e.g.  of 200% if the starting number is very small. And the 200% rise can still be both unimportant and statistically insignificant if it's way down in the noise of normal variation.

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site