The effect of rate of presentation, substitution, and mode of response in paired-associate learning.

The meaning of the phrase "to form an association" serves as a focus for the expe-riments to be reported here. Inferences about the processes that enable S to learn paired-associates range from analogies to the classical conditioning paradigm to assumptions of highly cognitive mediating acts.l In the conditioning analogy, there are, presumably, incremental increases in stimulus-response strength over successive repetitions of the pairs. Our own biases tend toward the view that most performance changes in the human adult depend on cognitive processes utilizing information previously acquired. The data presented by Rock were therefore of particular interest to us.2 It was Rock's contention that an association between the elements of a pair is formed in a single trial. Repetition is not necessary for the formation of associations, but does assume importance in that only a few associations may be formed on any one trial. Hence, to learn a long list of items, the list must be repeated. Two experiments were carried out using the general procedure employed by Rock. In Experiment I, the Ss were required to spell the nonsensesyllables, instead of pronouncing them, at two different rates of presentation. In Experiment II, a direct comparison of the spell vs. pronounce modes of response was made at a single rate of presentation identical to that used by Rock. The purpose of these procedural changes was to determine the generality of the phenomenon reported by Rock. Specifically, it was hypothesized that in a more diScult task Rock's conclusions will not be supported.