Friday Highlights

Good morning.

  1. Another take on the Twain edits.
  2. And yet another here.
  3. Old Believers in Siberia (HT).
  4. History, Bayes and the Resurrection.
  5. Authority is not the problem, only if its implementation requires coercion. 
  6. An interview noted.
  7. Some curious remarks from an IPCC author.
  8. Mr DeLong needs a saddle. (The saddle reference refers to a former Presidential quip (Truman?), “If one man calls you an ass, pay him no mind. If two or three do, buy yourself a saddle.”)
  9. An interesting use of Obamacare, which practice with a future GOP President will bite the Democratic hand that birthed it.
  10. Minorities and their status.
  11. A principle key to the Christian life.
  12. Behind the woodshed (HT).

Leave a Reply

Your email address will not be published. Required fields are marked *


  1. Boonton says:

    Another take on the Twain edits.

    I tend to agree, the book is in public domain which means people are in fact free to change it, ‘mash it up’ and whatnot. The publisher’s intent was not to get schools to swap out the ‘true book’ with the cleaned up one but to get schools not already using the book to do so.

    Another good point I heard a few nights ago on is that edited versions of books and art is used all the time. I remember the edition of Moby Dick that we read was edited for length. A lot of modern music is edited for the radio to remove swears. So all in all I don’t think its a bad thing.

  2. Boonton says:

    Authority is not the problem, only if its implementation requires coercion.

    Hey it’s the drug sniffing dog I used a while ago in my example of Type I and II errors!

  3. Mark says:

    You never answered my question of whether there was a distinction between your nomenclature of Type I/II and false positive/negative. It seems to me that calling it Type I/II just obscures the issue where false positive/negative at least gives clues to which is which.

  4. Boonton says:

    I don’t think there is a distinction. Type I and II are used in statistical tests whereas positive/negatives are used in diagnostics tests. But I think they both mean essentially the same thing except in statistics you’re not really talking about ‘positive and negatives’ but in cases of drug sniffing dogs or bomb detectors you are. (BTW, I tutored epidemiology for the first time last year and was happy to find lots of good material there more or less in the same boat).

    I appreciate the Reason writer’s angst against drug laws but in terms of probable cause I think he’s missing the point about the dog. The dog who gets 40% plus of his ‘positives’ wrong is not necessarily no different than flipping a coin. If this dog barks at your car it probably is more likely to have drugs in it and like it or not that is probable cause. Yes its a high error rate and having the police do a full search of your car is very intrusive but is this false positive rate really worse than, say, confidential police informants, which have been used for a very long time as enough probable cause to issue search warrants?

  5. Boonton says:

    #8, look like Brad de Long misread the guys email and apologized for it on his blog so Brad isn’t so ‘not nice’ after all (see IMO I think Brad is correct, there is no future of plentiful oil unless a radical innovation makes oil unnecessary thereby driving down demand dramatically. The added oil fields coming on line in places like Africa and Brazil and the Canadian tar sands are all ok but do not represent a massive increase in the supply of oil relative to growing demand in developing countries (and in the case of the tar sands, are a very costly way to extract oil that’s only viable in a high price environment).

  6. Mark says:

    You mean like us finding large methane deposits and moving much to consume that?

    Oh, wait … that’s happened. Could something else like that occur?