THE EFFECT OF COMPLEXITY OF NATURAL LANGUAGE MEDIATORS AND THE ASSOCIABILITY OF PAIRS ON PAIRED-ASSOCIATE LEARNING,
Alexander J. Wearing, William E. Montague · Illinois Digital Environment for Access to Learning and Scholarship (University of Illinois at Urbana-Champaign) · 1967
Natural language mediators (NIiMs) are widely used by _Ss in pairedassociate learning.Experiments which have documented their effect on learning haves however, largely ignored qualitative differences between them.Two large groups each learned a different CVC-word list after which they reported any NLMs they had used.Judges rated the complexity of NLMs using a scale developed by Martin, Boersma and Cox (1965) with different materials.The results agree with theirs in that complex NLMs produced fewer errors in learning.However, some categories on the scale were used infrequently which may indicate that, at least with highly meaningful material, a simpler dichotomous categorization (NLM or Rote) may be preferable.Because the two lists were different in learning difficulty, a second experiment was carried out in which the probability of NLM for mation for each of the stimulus pairs was determined, as this variable, called associability (AS), has been shown to be related to the rate of paired associate (PA) learning, (Montague and Kiess, 1966).An independent group of 50 Ss were presented with the CVC-word pairs used in the first experiment, and the Ss wrote down any NLM they had for each pair in 15 seconds.The proportion of Ss giving an NLM for a pair was de fined as its AS value.It was found that for the items in both lists