Confirmation, correction, and contiguity

Summary Three experiments were run to investigate events within a presentation of an S-R pair in verbal learning. The first experiment compared the anticipation technique with a technique eliminating both confirmations and corrections. The latter technique was found to produce faster learning. A second experiment compared a condition in which an erroneous pairing was forced just before the correct pairing with conditions in which one or two correct pairings were the rule. The condition simulating a correction technique resulted in slower learning than either of the other two conditions. The third experiment compared a correct anticipation which was not confirmed with one that was confirmed. No significant difference in retention after 60 sec. was found between the two conditions. It was concluded that there is a difference in effectiveness between a confirmation and a correction. The difference could be interpreted in terms of the number of pairings involved, and it was not considered necessary to add assumptions concerning the influence of special motivational effects. An assumption of postremity was not considered to be appropriate in a general theory of verbal learning.