[Rasch] Question - Probabilistic Information Transfer

Mike Linacre mike at winsteps.com
Fri Feb 21 10:24:00 EST 2014


thank you for sharing the announcement of the book "Probabilistic 
Information Transfer" (which I could not find on Amazon.com). the 
announcement claims "a new mathematical theory, entirely developed by 
SCM". The theory appears to be the Experimental Probabilistic 
Hypersurface (EPH).

Section V of 
describes EPH. According to my reading of the algebra, this is a process 
for predicting a new observation based on a weighted item series, with 
normally distributed observations at each previous point of the time series.

The previous observations could be Rasch measures, but otherwise EPH 
does not seem to relate to Rasch methodology.

Does anyone have a different understanding of this theory?

Mike L.

More information about the Rasch mailing list