2 experiments with a total of 50 right-handed college students were conducted to test a number of finite-state, self-terminating memory-scanning models for choice reaction times (RTs). The 1st class of models considered may be conceptualized as using probabilistic push-down stack memory mechanisms. 2-state and 4-state models of this class were investigated. The other class of models tested was represented by a 3-state model which assumed that the memorial stimulus-response associations could be found in 1 of 3 distinct states: high expectancy, short-term memory, or long-term memory. 4 stimuli and responses were used in Exp I and 6 in Exp II. In order to obtain tractable solutions to the predictions from the push-down stack models, each S had 1 stimulus designated as the "key" stimulus, which was presented with probability r. The remaining stimuli were all presented with probability (1-r)/3 in Exp I and probability (1-r)/5 in Exp II. In both experiments, the independent variable was r, which was varied as .10, .25, .40, .55, and .70. The data (mean RT, sequential RT, an distributional fixed-point property) uniformly supported the 2-state model (which in fact lies in the intersection of the 2 classes of models investigated). (PsycINFO Database Record (c) 2006 APA, all rights reserved).
|Number of pages||10|
|Journal||Journal of Experimental Psychology: Human Perception and Performance|
|Publication status||Published - May 1975|
- 2- &
- 4-state vs 3-state self terminating memory-scanning models, choice RT, college students