Learning of Random Shapes and Two Unique Sets of Concrete Nouns

In paired-associates learning, ten random shapes were used as stimuli and two sets of concrete nouns as responses. One set of nouns was high in descriptive appropriateness, da, the other set low. Two different groups learned the two combinations. Later, two additional groups learned with the pairings scrambled. Still later, a replication was made. The replication yielded results which, in most ways, were basically like those obtained originally; but there was an important exception that was related to the learning of the scrambled pairs for nouns of low da value