How variable are the classic ERP effects during sentence processing? A systematic resampling analysis of the N400 and...

Area of Science:

• Cognitive Neuroscience
• Psycholinguistics
• Electrophysiology

Background:

• Event-related potentials (ERPs) like N400 and P600 are crucial for studying language processing.
• Previous research has focused on the existence of N400 and P600 effects, but their measurement properties remain underexplored.

Purpose of the Study:

• To investigate the measurement properties of N400 and P600 ERP components.
• To assess the variability in temporal and spatial parameters of these ERP effects.
• To evaluate statistical power and effect magnitude variability in language processing research.

Main Methods:

• Utilized bootstrap resampling on a dataset of 187 participants and 40 stimulus sentences per condition.
• Analyzed variability in effect magnitudes, onset times, and scalp distributions across resamples.
• Focused on N400 effects for semantic anomalies and P600 effects for syntactic anomalies.

Main Results:

• Grand averages revealed robust N400 and P600 effects.
• Significant variability was observed across resamples in effect magnitudes, onset times, and scalp distributions, particularly for P600.
• Statistical power and effect variability may be higher than commonly assumed.

Conclusions:

• The study highlights substantial variability in N400 and P600 measurement properties, impacting interpretation.
• Findings provide a foundation for designing future ERP studies in language processing.
• Addresses challenges in ERP data analysis, including potential p-hacking concerns.