A comparison of process and memory-based theories of automaticity

Process and memory-based theories of automaticity are contrasted by manipulating the information contained in primary memory. Process-based theories predict that performance will be modulated by the information contained in primary memory, while memory-based theories predict that performance which is automatic will not be dependent of the contents of primary memory. The information contained in primary memory was manipulated by inserting a variable duration interference task between the memory set and the probe stimulus in a Sternberg memory search task. The interference task prevented rehearsal, necessitating the retrieval of the memory set from secondary memory in variably mapped (VM) conditions. Performance in consistently mapped (CM) conditions provided strong support for memory-based theories of automaticity. As performance became more automatic (as indicated by reductions in the effect of memory load), the effect of the interference task was reduced. As predicted by memory-based theories of automaticity, there was a temporal coupling in the reduction of these two effects with CM practice. Performance was modeled, using Monte Carlo simulations, as a race between a computational procedure (i.e., an algorithm) and a single-step direct memory access of past solutions. Manipulations which slowed the algorithm and/or decreased the minimum of the memory distribution resulted in more automatic (i.e., memory-based) performance. Automaticity is viewed as a continuum, reflecting the relative involvement of cognitive operations upon information contained in primary memory. The continuum is bounded by performance which is completely dependent on the information contained in primary memory (i.e., controlled processing) and by performance which is independent of the contents of primary memory.U of I OnlyETDs are only available to UIUC Users without author permissio