John Bunch asked for a citation and I have found one so far: Coombs. Dawes &
Tversky (1970, 257) reports, "The first paper to attempt to provide a serious
rationale for the learning curve was that of Thurstone in 1930 [Jrnl G. Psyc,
"The Learning function". He considered the two principal variables in learning
to be practice and atainment, the former represented by trials or time and the
latter by the probability that the act will be successful." []'s mine.
Prior to that, the earliest mathematical model is said to have appeared in 1907
(Gulliksen, 1934) and 1907-1930 was an era of several empirical models
documenting the learning curve. That period was also the major period of
Lillian Moller Gilbreath's work, which included some documentation of practice
effects. I'll search some other Pioneers of Psychology.
Hope that helps, whcih probably depends on whether the psyc folks read history.
--
Prof. John L. Naman naman+@pitt.edu