A Corpus of Joint EEG and Self-Paced Reading of Natural Dutch Texts

Abstract

We present the Tilburg corpus of Natural Dutch Texts (TiNT): A corpus of joint electroencephalography (EEG) and self-paced reading (SPR) of natural, medium-length, Dutch texts. The corpus contains recordings from 71 native Dutch speakers reading eight naturally occurring texts of around 600 words each. The texts are of varying genres and were chosen based on overall fluency and comprehensibility. To assess the quality of the corpus, we examined participant responses to comprehension questions, self-reported familiarity with the texts, and whether well-established effects replicated for both reading times and event-related potentials (ERPs) (N400 and P600). The corpus contributes to a small collection of corpora with simultaneous recording of reading times and EEG. While this is often achieved using eye-tracking, the use of SPR offers methodological advantages, particularly in aligning neural signals with word-level processing. In addition, the use of natural texts with longer dependencies makes the corpus a unique resource for psycholinguistic research. The corpus enables research into the relationship between neural and behavioral responses in naturalistic reading contexts.

Resources

Details