Author Topic: SGU 5x5 #105  (Read 1298 times)

0 Members and 1 Guest are viewing this topic.

Offline Mike

  • Frequent Poster
  • ******
  • Posts: 2367
    • SGUFans.net
SGU 5x5 #105
« on: March 13, 2012, 07:23:35 PM »
Representativeness Heuristic
"We're just so damn exciting." - Dr. Steven Novella, MD

Offline werecow

  • Cryptobovinologist
  • Stopped Going Outside
  • *******
  • Posts: 4215
  • mooh
Re: SGU 5x5 #105
« Reply #1 on: March 14, 2012, 04:40:58 AM »
This one was great guys, cheers! 
Mooohn!

Offline Trinoc

  • Frequent Poster
  • ******
  • Posts: 3975
  • Dumb, in a pocket, and proud of it.
Re: SGU 5x5 #105
« Reply #2 on: March 14, 2012, 09:21:37 AM »
I think they are lumping rather a lot of things which probably require their own explanations, under a single heading called "representativeness heuristic". I have heard of all of the fallacies they listed, but I don't recall ever having heard of the RH before, and I'm not sure this blanket heading helps at all other than perhaps to provide us skeptics with a bit of a lazy get-out: simply tell someone with whom we disagree that it is an example of RH rather than examining the details further.

BTW, any statisticians out there: Shouldn't "regression to the mean" actually be called "regression to the median"? That is, doesn't a parameter tend to regress towards the point where 50% of samples would be above and 50% below, rather than regressing to the arithmetic average (the mean) of the value of those samples.

If I'm moderately well paid (if only!) and my income is regressing to the mean then it would go up because all of the millionaires and billionaires out there are skewing the average upwards. What is actually likely to happen of course is that my income will go down to something more like the point where 50% of comparable people are making more than me and 50% less. That is, the median.

Or perhaps it might regress to the mode, i.e. the income range of most of the people. One thing it doesn't necessarily regress to is the mean.
I'm a skeptic. Not a "skepdude". Not a "man skeptic". Just a skeptic.