A comparison of the effects of formal similarity among trigrams and among word triads

Abstract

The central question of this report concerned thewie of formal similarity in free recall of lists of trigrams andlists of three-letter word triads. Similarity was manipulated among trigrams by duplicating letters and among triads bv duplicating words. An initial study showed that lists of 16 letters were learned more rapidly thana listof 16three-Ietter words. Therefore, in the major experiment, the Ss were given all appropriate elements on test trials so that only associative learning was required. Increases in formal similarity caused decreases in rate of learning for both types of lists, and the mechanisms of the interference seemed to be the same for both types of lists. However, the learning of the trigram lists was more rapid than thelearning of the triad lists, the difference being maximal with low similarity.

Semantic Encoding and the Effects of Formal Intralist Similarity

“…This suggests that the curvilinear relationship might be found only when response learning is minimized. However, in an earlier study (Underwood & Zimmerman, 1974) that used the methods of the present experiment, no curvilinear relationship was found when the triads consisted of three-letter words and similarity was varied by repeating words. In the same study there was no evidence of curvilinearity when CCC trigrams were used and similarity was varied by varying the number of different letters used to construct the trigrams.…”

“…A technique of study worked out in a previous experiment was used (Underwood & Zimmerman, 1974). The subjects in this earlier study were presented a list of word triads to learn, e.g., ale-fadtab.…”