Correspondence concerning normalized gain scores
On approaches to assessing change. This piece is the text of an email I sent to Prof. Richard Hake on March 25, 2006, in response to his posting on the AERA-D listserv at <http://lists.asu.edu/cgi-bin/wa?A2=ind0603&L=aera-d&T=0&F=&S=&P=339> concerning the use of normalized gain scores <g> to measure learning. Prof. Hake posited that psychometricians seemed unaware of <g>. Based in part on his conversations with me, Physicist Joe Redish at the University of Maryland suggested that the case was less one of unawareness than one of uninterest, due to its limitations. In the first piece below I offer below some amplifications of this line of thought. The topics I discuss are, in turn, item response theory (IRT) as a way of dealing with floor and ceiling effects, IRT as a way of characterizing change in terms of conceptual development, further extensions, and lines of contemporary work.