SEQUENTIAL RECALL OF A MIXED LIST
Robert T. Green, Graham F. A. Harding · British Journal of Psychology · 1962
A previously reported finding that an item with a high probability of recall will not necessarily be emitted any earlier in a recall series was submitted to closer investigation. The anomalous result was reproduced, using the original experimental design. Two further groups were introduced to control context effects and item uniqueness. More careful analysis shows that the factors governing order of emission during recall may be distinguished from those affecting the overall probability of recall. The two processes are distinct and, in some circumstances, independent. After eliminating various alternative hypotheses the outcome in the present instance was found to depend largely on previously learned associations of meaningful material. The implications of this with regard to the von Restorff effect are discussed. During the course of this experiment an interesting side effect intruded itself on the data. For reasons not at present clear, there was a significant tendency for the recall of paralogs to be disturbed in the presence of a noun, and, to a lesser extent, for the recall of nouns to be disturbed by the presence of a paralog.