Description
One of the most critical components of reading is visual word recognition. The masked priming technique has been widely used to study visual word recognition because it isolates the earliest processes involved in word recognition. When combined with event related potentials (ERPs), it is a powerful way to see the millisecond by millisecond time course of the underlying processes. However, this technique lacks spatial resolution. fMRI, which has excellent spatial resolution, has yielded few results when combined with masked priming. This may be due to the fact that visual word recognition effects in masked priming are quite small. A compounding issue is that fMRI studies to date have used an event-related design with group-level analysis which considerably reduces power and can wash out effects. The current study aims to capture both the when and where of masked priming by utilizing both ERP and fMRI with a block design to increase power and individual ROI analysis to reduce the washout from group analysis. Two masked priming experiments were conducted. Participants were presented with word pairs which were either repeated or unrelated. Experiment 1 utilized ERPs to test the viability of a blocked design by presenting half the stimuli in a block design and half in an event-related design. Three time windows were examined: the N/P 150 (125-175), N250 (175-280ms) and N400 (300-500ms). There were no significant differences in effects between the block design and the event-related design. Experiment 2 utilized fMRI and the block design from Experiment 1 and also included a localizer to individually identify 6 regions involved in orthographic, phonological, and semantic processing: the visual word form area, precentral gyrus, inferior frontal gyrus (anterior and posterior), temporoparietal junction and middle temporal gyrus. All regions showed significantly greater percent signal change in the unrelated condition compared to the related condition. This provides evidence that a masked priming paradigm can indeed capture these higher-level processes in fMRI. Together, these results expand on previous literature in both ERPs and fMRI and provide an exciting finding that could lead to a more extensive literature on masked priming paradigms in fMRI.
