Take home messages

. Think for yourself

. Respect variation / noise .. tame it if possible/appropriate

. Don't capitalize on noise/chance

. The p-value is not the holy grail -
  (tell it like it is, not like it isn't)

. Don't rely on . . .
 
  - a particular statistical method just because
    other authors (or your supervisors) have

  - statistics the way a drunk relies on the lamppost

jh 2013.11.27