Date Published: March 15, 2019
Publisher: Public Library of Science
Author(s): Canhuang Luo, Wei Chen, Ye Zhang, Carl Michael Gaspar, Jérôme Prado.
The recognition potential (RP) is an early visually evoked response (~250 ms) whose magnitude is sensitive to object recognizability and related factors. The RP is often measured when objects are embedded in a rapid stream of masking stimuli (the RSS paradigm), especially in reading research. The idea is that RSS provides greater stimulus-dependent variations in RP, compared to the corresponding variations without RSS. However, this idea has never been subject to systematic evaluation. We directly test whether RSS can enhance 2 types of RP stimulus selectivity, by measuring the RP in conditions that only differ in the presence or absence of a masking stream and in the type of stimulus shown. We measure the effect of image inversion on RP for Chinese characters (experiment 1); the effect of orthographical correctness on RP for Chinese characters (experiment 3); and as a control study, the effect of image inversion on the N170 response to faces (experiment 2). To ensure a fair comparison, the earliest negative deflections (RP/N170) measured with and without RSS should at least have similar channel ranges, and topographical distributions of both amplitude and selectivity. Our first set of results clearly supports this. Our main results clearly show an increase in stimulus-selectivity for RSS over non-RSS in both of the RP (Chinese character) experiments, but no such enhancement in the N170 (face) experiment. This provides incentive for investigations into the underlying mechanisms of selectivity enhancement. Our findings may also help to explain contradictory findings between RP/N170 studies that differ only in the use of noise masks, which is sometimes treated as trivial detail in papers that do not reference the RP/RSS literature.
The rapid-stream-stimulation procedure (RSS) is a paradigm for stimulus presentation that is thought to enhance early electrophysiological responses to recognizable stimuli. In this study, we test whether RSS increases the recognition potential (mostly measured in response to linguistic material), and whether it also affects selectivity measured with other stimuli (the N170 response to faces in this study).
This study presents the first systematic and fair comparison of both RP and N170 selectivity between RSS and non-RSS paradigms. We confirm prior assumptions about RSS by demonstrating that the RSS paradigm can enhance the stimulus selectivity of RP component amplitude in 2 different experiments that use Chinese characters as stimuli. Our enhanced ‘orthography effect’ in particular (experiment 3), may help to resolve contradictory results by showing noise-masking may be necessary to even obtain a stimulus selectivity. These positive results with RP responses to Chinese characters contrast with our face study, which shows that RSS does not enhance the N170 selectivity to face orientation. Unlike the familiar Chinese characters and radicals that we use in experiments 1 and 3, our Chinese faces were unfamiliar to our participants. Therefore, our results with unfamiliar faces is consistent with the notion that RSS enhances selectivity to aspects of the stimuli that are distinctly recognizable, and richer with semantic association. Finally, we include a complementary body of results, including some novel analyses, which addresses the important question of whether N1 components measured with and without RSS are comparable in the first place. Our results show that, at least for the stimuli we use (faces and Chinese characters), the answer is yes.