Abstract

One of the difficulties of the framework of levels-of-processing approach to memory is associated with the absence of an independent index of processing depth. As an attempt to identify levels of processing and to find a possible index of levels of processing, the release from proactive inhibition (RPI) paradigm was examined within a framework of the levels-of-processing view of memory. Two experiments were undertaken to examine (1) whether a shift in levels of processing (structural-semantic) would produce a RPI, (2) whether a shift in dimensions within a given level of semantic processing (action-potency-evaluation judgement) would produce a RPI, and (3) whether the nature and amount of RPI would be functionally associated with levels of processing. In Experiment 1, two levels (structural and semantic) of processing were included to examine whether a shift in. level of processing would produce a RPI. On Trial 1-3, five words for each trial were exposed one at a time under either structural processing instruction (letter identification) or semantic processing instruction (category judgement). On Trial 4, a shift in level of processing was instructed. On each trial the sbujects were required to recall the stimulus words unexpectedly. Semantic processing produced a significantly higher recall performance than structural processing. The recall performance decreased significantly over trials 1-3. For the data on Trial d, a significant main effect of Shift variable was found, suggesting that a reliable RPI was obtained following a shift in level of processing. le Experiment 2，three processing dimensions within semantic level of processing (active-passive, strong-weak, and pleasant-unpleasant judgement) were included to examine whether a shift in dimension of processing would also produce a RPI. The task and procedure were the same as those for Experiment 1, except for the processing instructions. There was a significant decrease in recall performance over trials 1-3. For the data on Trial 4, there was a significant increase in performance following a shift in dimension of processing. The above findings were discussed with in the levels-of-processing view of memory. It was concluded that the RPI paradigm could be used in order to identify different levels of processing and dimensions within levels of processing. However, further work will be necessary to state whether the direction or degree of RPI following shift in levels of processing is also functionally related with any hierachical ordering of processing levels.