Serial pattern learning: Higher order transitions

Abstract

Defines a transformation as a way of changing a serial pattern by acting not directly upon its elements, but on its structure. 3 transformations were tested experimentally using 64 undergraduates with musical knowledge: extension (e.g., 1 1 2 2 2 3 3 3 3), elaboration of new lower order nodes (e.g., 1 2 1 1 2 2), and interspersal of 2 different substructures (e.g., 1 5 2 5 3 5). Extension considerably facilitated learning of a long sequence, and any deviation from regular extension was harmful. The process of elaboration, as tested, had no effect on learning. The process of interspersal benefited learning, but the results left open the possibility that Ss may just learn one dominant theme and not pay much attention to the other subordinate theme. It is concluded that such higher order transformations were used by Ss in learning serial patterns and that the specific unique properties of each operation was important. The abstract theory of hierarchical trees is unable to give a complete account of serial pattern learning. (PsycINFO Database Record (c) 2006 APA, all rights reserved). © 1973 American Psychological Association.