Abstract

This study was purported to reaffirm the existence of mediation effects in the three stage transfer situation. Experiment I was conducted to find out mediation effects in the eight paradigms using paired associate learning. Dropping method was used to equalize the degree of learning of each pair in stage I and stage II. The results showed that among eight paradigms, significant mediation effect was found only in the paradigm I and mediation-tendency was observed in the paradigm V, though not signiffcant. Experiment II was to retest the mediation effect in the group situation using only three paradigms; I, V and VII. Contrary to the findings of Experiment I, mediation-tendency was observed only in the paradigm VII. The results of these two experiments were rather inconsistent with thestrong mediation effects in the other experiments (e.g., Horton and Kjeldergaard, 1961; Cramer (1967). Our results might be due to the use of dropping method (Experiment I) in which the degree of prior learning might not be sufficient to enhance mediation effects in the test stage. Many other reasons such as the confusion between implicit mediation of each stage and over mediation were also presented to explain the findings of this study. It was suggested that the study of these mediation related variables might be more appropriate for the further study on the mediation than the experiments to test the existence of mediation effects might. It was also suggested that the nature and locus of mediation might be analyzed and interpreted in terms of memory process, especially the "naming process" in STM and LTM.