Rapid serial visual presentation (RSVP) and Naturalistic reading paradigms are two of the most popular paradigms used in neuroimaging studies of reading. In RSVP experiments single words or phrases are presented on screen one at a time before quickly being replaced by the next word/phrase. In naturalistic reading tasks bodies of text are presented and participants read at their own pace and their eye movements are tracked. While both of these methods are used, literature comparing them is sparse. Our research is examining the affordances and limitations of each paradigm in order to better understand their strengths and weaknesses.
In this review of RSVP and naturalistic reading paradigms we conclude that while RSVP affords more control over stimuli and what participants are reading it is difficult to get readers to form a text model. This method is also quite different from how we typically read and thus lacks ecological validity. In naturalistic reading paradigms readers more easily construct text models making this a good way to study comprehension. However, with these paradigms it is difficult to determine what a reader is focusing on or processing at any given time (even with eye tracking). Challenges with stimulus design for each and subtractive methods also complicate any direct comparisons. We suggest that further work comparing brain activation in sentence-based RSVP experiments to that observed with naturalistic reading tasks and studies designed to verify in vs. out of scanner eye-tracking will help us to better understand and compare these paradigms.