| Abstract: | Presents an analogy between the learning theorist and the test-theorist, suggesting that present models for test prediction do not provide for generalizability of test results fron 1 situation to another or for explaining failures of tests to predict. It is proposed that future models include motivational variables. As a 1st step, it is proposed that Hull's model of learning be adapted for test prediction. The new model would be as follows: P = A * M * L, where P = performance; A = some weighted sum of aptitude, knowledge, and skills; M = some weighted sum of internal and external motivations; and L = situational variables. (PsycINFO Database Record (c) 2010 APA, all rights reserved) |
