Sunday, February 19, 2006
Rasch: An idea whose time has come?
There are 4 elements necessary for an idea "whose time has come" to run rampant: connection, validation, selling and luck. In his recent Kenneth Myer Lecture, Tim Costello reminded us of Malcolm Gladwell's theory of a Tipping Point.
"Connection" means that people propagating the idea must be well connected to opinion formers and the influential.
"Validation" means that the idea must be supported by some people perceived to be experts.
"Selling" means that the idea must have dynamic advocates who can present the idea clearly and persuasively.
"Luck" means that some of those accidental sets of circumstances, that occur all the time, can be turned to assist the promotion of the idea.
Rasch measurement is advancing on all fronts. We have had 50 years of preparation so that, when we seize that "lucky moment", Rasch will become an "overnight sensation"!
"Connection" means that people propagating the idea must be well connected to opinion formers and the influential.
"Validation" means that the idea must be supported by some people perceived to be experts.
"Selling" means that the idea must have dynamic advocates who can present the idea clearly and persuasively.
"Luck" means that some of those accidental sets of circumstances, that occur all the time, can be turned to assist the promotion of the idea.
Rasch measurement is advancing on all fronts. We have had 50 years of preparation so that, when we seize that "lucky moment", Rasch will become an "overnight sensation"!
Monday, February 13, 2006
Olympic Figure Skating Judging
The International Skating Union has introduced yet another rating system in the interests of "fairness". It's aim is definitely to be obscure, and to hide judge behavior from analysts like me. But the only new regulation that might have defeated the sophisticated scheming that went on at Salt Lake City in 2002 is the "only 9 out of 12 judges count". The 9 are supposed to be assigned randomly. But, if I was a crooked judge, this would not deter me. I've a 75% chance that my ratings count ... so I'll go ahead and cheat. If they don't count, then I've lost nothing. If they do count, then I've gained what I wanted.
Does anyone know if the random assignment of judges takes account of the well-established East-West judge bias? Is this stratified "East, West" random sampling, or could we end up with 6 Eastern judges and 3 Western judges making the decisions?
Does anyone know if the random assignment of judges takes account of the well-established East-West judge bias? Is this stratified "East, West" random sampling, or could we end up with 6 Eastern judges and 3 Western judges making the decisions?
Online Rasch Course: 31 participants!
One week into the Practical Rasch Measurement Course sponsored by Statistics.com. 31 participants with a wide variety of expertise and experience. Some are old friends. It's a challenge for us all. Bond & Fox's "Applying the Rasch Model" is the basic text, but with plenty of supplemental readings and exercises. Also two Special Projects, the US PGA golf measures and rankings and the Winter Olympics Women's Moguls ratings. We're also experimenting with a chat room [but this didn't work out well] ... It's an adventure for everyone! If you are a participant (or want to be one), please comment ....