JOINT EFFECTS OF MEMORY LOAD, STIMULUS QUALITY, AND WORD FREQUENCY ON LEXICAL DECISION PERFORMANCE

Abstract

The joint effects of word frequency and stimulus quality in the lexical decision task (LDT) is a common research interest due to the robust additive effect observed which implies a staged processing model. This joint effect can however be moderated by the type of non-word, such as transposed letter (TL) non-words, which produce an over-additive effect at distributional level because the system reverts from a familiarity mechanism to analytical checking. This study examines the effect of memory load on LDT with either legal non-words (eg. FLIRP) or TL non-words (eg. BRANE). The memory load should have little impact on LDT with legal non-words and produce the established additive effect but should cause over-additive effect to be observed at the mean and distributional level for TL non-words. However, results show that additive effect at the mean and distributional level were obtained for both legal non-words and TL non-words. This suggests that processing follows a staged model in both situations. This is unexpected for TL non-words but could indicate an inability to pivot to an attentionally demanding analytic checking process due to the memory load imposed, providing insights on the role of memory load in LDT.