CASE STUDY: Interactive and Additive Effects of Word Frequency and Predictability
The process of reading involves a dynamic interplay of bottom-up and top-down mechanisms. Bottom-up processing refers to how our brains interpret basic sensory information, like recognizing individual letters and words. Top-down processing, on the other hand, involves using prior knowledge and context to influence perception. A recent study by Schuster et al. (2025) titled “Interactive and additive effects of word frequency and predictability: A fixation-related fMRI study,” sheds light on these mechanisms, and highlights how the integration of eye tracking with fMRI can provide invaluable insights into the cognitive processes underlying word recognition.
Reading Research Methodology with Eye Tracking and fMRI
The core question addressed by Schuster et al. was whether word frequency (a bottom-up indicator) and word predictability (a top-down indicator) have an additive or interactive effect on processing speed and neural correlates during reading. An additive effect would suggest that contextual top-down processing occurs post-lexically, while an interactive effect would imply concurrent processing of both variables in early stages of word recognition.
To investigate this, the researchers employed a sophisticated methodology: fixation-related fMRI. This approach allowed participants to engage in self-paced silent reading, mimicking natural reading conditions, while their eye movements and brain responses were co-registered. The onset of each first fixation on the target words was used to model the canonical hemodynamic response function in the fMRI data. The study involved 24 native German readers who read 144 sentences designed to manipulate word frequency and predictability. Eye movements were recorded using a long-range SR Research Eyelink 1000 eye tracker system, and fMRI data was acquired with a Siemens PRISMA 3T scanner. This dual-modality approach was critical for capturing the subtle, rapid shifts in attention and cognitive processing that occur during reading.
Primacy of Bottom-Up Processing and Top-Down Information Integrated Post-Lexically
The gaze data showed significant main effects of both word frequency and predictability. Participants exhibited shorter first-fixation durations for high-frequency words compared to low-frequency words, and for high-predictability words compared to low-predictability words. Crucially, the eye-tracking data showed no interaction between word frequency and predictability. This suggests that these two factors had independent effects on reading behavior.
The fMRI results corroborated these findings, showing exclusively additive effects. Higher activation for low-frequency words was observed in the orbital part of the inferior frontal gyrus and the left fusiform gyrus, regions associated with lexical access and visual word form processing. Similarly, unpredictable words elicited higher activation in the inferior frontal gyrus, left middle frontal gyrus, left insula, and supplementary motor cortex, indicating increased cognitive load for semantic integration. The absence of an interaction effect at the whole-brain level, even at lowered thresholds, further supported the additive nature of these processes.
Fixation related fMRI is a very powerful technique. By precisely measuring fixation durations and locations, the researchers could identify when and where processing difficulties occurred. This allowed them to differentiate between early lexical processing (indexed by word frequency effects) and later post-lexical semantic integration (indexed by predictability effects). Without the precise temporal and spatial resolution offered by the EyeLink eye tracker, disentangling these effects would have been significantly more challenging.
The study by Schuster et al. provides compelling evidence for the additive effects of word frequency and predictability during silent reading, suggesting a primacy of bottom-up processing with contextual top-down information being integrated post-lexically. This research underscores the vital role of eye tracking as a methodological tool in cognitive neuroscience. Its ability to provide real-time, fine-grained measurements of reading behavior is indispensable for understanding the complex interplay of cognitive mechanisms during language comprehension.
For information regarding how eye tracking can help your research, check out our solutions and product pages or contact us. We are happy to help!