حواس پرتی در یک الگوی چند منحرف بصری: رفتاری و اثرات بالقوه مربوط به رویداد
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|38737||2009||7 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : International Journal of Psychophysiology, Volume 72, Issue 3, June 2009, Pages 260–266
Abstract The present study aimed at investigating visual distraction in a serial, multi-deviant oddball paradigm with deviant stimuli occurring regularly (every third trial), having a larger overall probability (1/3), and low dimension-specific probability (1/9). Participants performed a categorization task (odd/even) on centrally presented digits. Task-irrelevant geometrical forms were presented concurrently in the left and right periphery of the target. These forms were green triangles that, in every third trial, contained a deviancy either in location, color, or shape at the left or right peripheral position. Behavioral performance and event-related potentials (ERPs) were measured during the multi-deviant blocks and during corresponding control blocks to compensate for physical differences. Results revealed prolonged reaction times for the categorization task in trials containing a deviant feature relative to the respective control condition. Furthermore, two negative shifts were observed in the ERPs for deviant compared to control stimuli, the early one at the ascending part of the N1 component, and the later one at the onset latency of the N2 component. Deviant displays violating a sequential regularity on one of the dimensions thus elicit respective posterior ERP components of change detection and a deterioration in task performance even when they occur as frequently as in every third trial of a sequence. In analogy to findings in audition, these results reveal the importance of regularity processing and its immediate consequences for adaptive behavior also in vision.
Introduction Detecting and evaluating unexpected changes in the sensory environment is a crucial prerequisite for adaptive behavior. The human auditory system, for instance, has the ability to automatically extract regularities unfolding over time in the acoustic input, and to detect violations from such regularities (e.g. Alho et al., 2003, Escera et al., 1998, Escera et al., 2000 and Schröger and Wolff, 1998) even when they are not relevant for one's current goals, such as during the performance of a specific task. This ability is important because regularity violations signal a change in the environment that may require behavioral adaptation. The task-irrelevant violations may actually lead to a temporary impairment of performance, which is usually explained by an involuntary shift of attention towards the irregular, unexpected event (cf. Escera and Corral, 2007). Similar distraction effects have also been shown for infrequent and unexpected stimuli in the visual modality (Berti and Schröger, 2001, Berti and Schröger, 2004 and Berti and Schröger, 2006). In the present study, we were interested in the processing of regularity violations with a newly developed multi-deviant distraction paradigm resulting in a low probability for a specific deviant but high overall deviant-probability. If under these conditions distraction effects were obtained, this new paradigm could be used as a time-saving alternative to previous approaches for studying the processing of visual regularity violations, especially when testing populations with a need for short and simple experiments. Moreover, if indeed performance were disrupted by distractors, despite the fact that distracting events occur relatively frequent and highly regular, it will suggest that regularities on different visual dimensions are evaluated in parallel and in at least partially independent manner. Similar attempts have recently been made for the auditory distraction paradigm (Grimm et al., 2008 and Jankowiak and Berti, 2007). In its original version introduced by Schröger and Wolff (1998), participants classify short and long sounds (occurring equiprobably) by their duration. Randomly and infrequently (with 10% probability), deviant sounds occur that differ in spectral frequency. Typically, the frequency deviants are classified with a prolonged reaction time. In the ERP, when compared to the standard sounds, deviants elicit a fronto-central negative component at around 150 to 250 ms (MMN, N2b) reflecting the detection of the deviation, and a positive component at around 280–400 ms (P3a) indicating an attentional shift ( Schröger and Wolff, 1998). In contrast to the original approach, the multi-deviant version of the paradigm presents different deviants within a sequence, such as frequency, intensity, and location deviants during a duration categorization task. Jankowiak and Berti (2007) showed that reaction time prolongation and deviance-related ERP components are obtained even when the overall probability for the occurrence of a deviant is increased up to 33%. More specifically, every third tone in the sequence contained a feature change that randomly concerned one of the three deviant dimensions. Grimm et al. (2008) additionally showed that with this multi-deviant approach, behavioral and electrophysiological distraction effects are of the same amplitude as in the original single-deviant paradigm, despite the fact that deviants are three times as likely to occur, and their occurrence in the sequence is predictable. The present study is the first one to test whether this multi-deviant approach can also be applied in the visual modality. Obtaining reliable distraction effects would argue in favor of a uniform theory of involuntary orientation of attention towards unexpected events across the two modalities of vision and audition. Comparison between the visual and auditory domain is often restricted owing to the usage of different paradigms. In the visual modality, distraction has predominantly been measured in response-competition paradigms (e.g., see Kim et al., 2005, Kraft et al., 2007, Lavie, 2005 and Lavie et al., 2004) and attentional-capture paradigms (e.g. Eimer and Kiss, 2008 and Folk and Remington, 1998; for a review, see Ruz and Lupiáñez, 2002). However, distractibility can also be investigated using distractor stimuli that violate a sequential regularity as usually done in auditory distraction paradigms (with the regularity being the repetition of a standard stimulus or parameter; Escera et al., 1998 and Schröger and Wolff, 1998). Following this approach, it has already been shown that the visual system is able to automatically detect infrequent sequential changes in a series of consecutive, otherwise regular displays (e.g. Czigler et al., 2004, Czigler et al., 2006, Kimura et al., 2009, Pazo-Alvarez et al., 2003 and Tales et al., 1999). Moreover, visual change detection leads to similar distraction effects as in audition, when infrequent regularity violations are presented while participants perform a task. For instance, Berti and Schröger, 2001 and Berti and Schröger, 2004 presented a sequence of squares containing a triangle at an exposure duration of 200 ms or 600 ms, and participants were instructed to categorize the visual stimuli by their duration. In 12% of the trials, the position of the triangle inside the square was shifted. In these deviant trials, participants responded more slowly, and compared to the ERP elicited in the regular standard trials, the deviant ERPs showed a negativity over parieto-occipital electrodes at around 200 ms (visual MMN, N2b) and a positivity over frontal electrodes at around 400 ms (P3a). Analogous to the auditory modality, the first component is supposed to reflect a process of visual change detection, whereas the second is assumed to reflect a shift of attention towards the irregular change. Even though the effects were smaller than those measured in the auditory modality (cf. Berti and Schröger, 2001), the findings confirm that the visual system likewise extracts sequential regularities in stimulus sequences and interferes with concurrent mental processes (distraction) as soon as a regularity violation appears. The present test of a visual multi-deviant paradigm will reveal whether distraction effects are still obtained when violations on different visual stimulus features (location, color, and shape) are presented during the same visual sequence with low probability of each feature change per se, but high overall probability of deviant displays (33% as in the auditory multi-deviant paradigm). The presence of deviance-related effects would suggest analogous mechanisms of visual change detection and attentional orienting as previously shown for the auditory modality.
نتیجه گیری انگلیسی
3. Results 3.1. Behavioral data Overall, subjects gave a correct response in the digit categorization task in 96.8% of the trials, with an average reaction time of 444.4 ms. The percentage of correct responses was not influenced by any of the experimental factors (F < 1.4, p > 0.27). Mean RT was found to be slowed in response to deviant trials (447.84 ms) when compared to the corresponding control trials (440.95 ms) as indicated by an effect of STIMULUS TYPE in the repeated measures ANOVA (F(1,15) = 10.57, p = 0.005). Furthermore, there was a main effect of FEATURE (F(2,30) = 4.27, p = 0.032). Both deviant and control displays containing a location asymmetry were responded to faster (440.27 ms) than those containing a color asymmetry (447.88 ms, F(1,15) = 6.89, p = 0.019) or a shape asymmetry (445.04 ms, F(1,15) = 6.10, p = 0.026). However, the distraction effect itself (the RT difference between deviant and control trials) showed no difference between the three features as can be derived from the missing interaction STIMULUS TYPE x FEATURE (F(2,30) = 2.44, p = 0.11). No effect of SIDE or any other interaction was obtained (Fs < 2.5, p > 0.18). Reaction times for deviant and control stimuli and their difference are given in Table 1 separately for the three feature dimensions. Table 1. Mean reaction times and mean amplitudes of ERP components in response to deviant and control stimuli for displays containing a location, color, and shape asymmetry are shown. Location Color Shape Dev Con Dev – con Dev Con Dev – con Dev Con Dev – con Reaction times (ms) 447.3 (10.4) 433.3 (8.8) 14.0 (4.1)** 448.4 (9.5) 447.4 (10.9) 1.0 (4.6) n.s. 447.3 (9.7) 442.2 (8.6) 5.6 (3.3)* Mean amplitude in the early time window (µV) Left occipital ROI − 5.25 (1.01) − 4.46 (0.86) − 0.87 (0.38) * − 4.27 (0.82) − 3.72 (0.82) − 0.54 (0.32) n.s. − 3.83 (0.82) − 3.91 (0.71) 0.08 (0.40) n.s. Right occipital ROI − 7.36 (1.16) − 6.46 (1.09) − 0.90 (0.42) * − 6.16 (1.05) − 5.42 (1.11) − 0.74 (0.28) ** − 5.88 (1.20) − 5.45 (1.03) − 0.43 (0.35) n.s. Mean amplitude in the late time window (µV) Left occipital ROI 2.78 (1.13) 3.43 (0.98) − 0.64 (0.43) n.s. 1.05 (0.72) 1.80 (0.76) − 0.75 (0.40) * 2.10 (1.00) 2.58 (1.04) − 0.47 (0.44) n.s. Right occipital ROI 2.64 (0.95) 3.57 (0.85) − 0.93 (0.36) * 0.89 (0.55) 1.80 (0.76) − 0.87 (0.38) * 1.96 (0.77) 2.65 (0.55) − 0.70 (0.37) * Additionally, values for the difference between deviant and control response and corresponding indicators for statistically significant differences are given. To verify differences between the responses to deviants and controls, one-sample one-tailed Student's t tests were applied to test against zero level (***p < .001,**p < .01, *p < .05, n.s not significant). Table options 3.2. ERP data Grand-average ERPs in response to the two stimulus types (deviants and controls) are depicted in Fig. 2, Fig. 3 and Fig. 4, respectively for the location, color, and shape dimension. In addition, scalp topographies and scalp current density (SCD) maps of the difference wave between deviant and control ERPs are given for the early (115–145 ms) and the late latency ranges (for location: 230–260 ms; for color: 265–295 ms; for shape: 245–275 ms). Mean amplitudes of the ERPs for deviants and controls as well as for the difference waves for each feature can be found in Table 1. Full-size image (74 K) Fig. 2. Left: Grand-average ERPs elicited by location deviants (solid line) and location control stimuli (dashed line) and the corresponding difference waveform over the right frontal (upper diagraph), left occipital (middle diagraph) and right occipital (lower diagraph) electrode cluster. ERPs differ in an early (115 to 145 ms) and a late time window (230 to 260 ms). Vertical dashed lines mark the latencies of the relevant differences. Right: Scalp potential (upper row) and scalp current density distributions (middle row) of the difference between deviant and control displays shown for the early and late time window. Additionally, mean amplitudes of the difference waveforms depending on the stimulus presentation side and the electrode cluster are given in bar charts (lowest row). Figure options Full-size image (72 K) Fig. 3. Left: Grand-average ERPs elicited by color deviants (solid line) and color control stimuli (dashed line) and the corresponding difference waveform over the right frontal (upper diagraph), left occipital (middle diagraph) and right occipital (lower diagraph) electrode cluster. ERPs differ in an early (115 to 145 ms) and a late time window (265 to 295 ms). Vertical dashed lines mark the latencies of the relevant differences. Right: Scalp potential (upper row) and scalp current density distributions (middle row) of the difference between deviant and control displays shown for the early and late time window. Additionally, mean amplitudes of the difference waveforms depending on the stimulus presentation side and the electrode cluster are given in bar charts (lowest row). Figure options Full-size image (71 K) Fig. 4. Left: Grand-average ERPs elicited by shape deviants (solid line) and shape control stimuli (dashed line) and the corresponding difference waveform over the right frontal (upper diagraph), left occipital (middle diagraph) and right occipital (lower diagraph) electrode cluster. ERPs differ in an early (115 to 145 ms) and a late time window (245 to 275 ms). Vertical dashed lines mark the latencies of the relevant differences. Right: Scalp potential (upper row) and scalp current density distributions (middle row) of the difference between deviant and control displays shown for the early and late time window. Additionally, mean amplitudes of the difference waveforms depending on the stimulus presentation side and the electrode cluster are given in bar charts (lowest row). Figure options In the early time range, coincident with the early half of the N1 response, ERPs elicited by deviant stimuli showed a stronger negativity (− 5.5 μV) in comparison to the control stimuli (− 4.9 μV). The effect was small but reliable, as reflected by a main effect of STIMULUS TYPE (F(1,15) = 6.61, p = 0.021) in the repeated measures ANOVA. This effect of STIMULUS TYPE interacted with the factor CLUSTER (F(1,15) = 5.33, p = 0.036). That is, differences between deviant and control displays were larger at the right electrode cluster (− 0.7 μV, t(15) = − 3.21, p = 0.003), however also present at the left cluster (− 0.4 μV, t(15) = − 1.80, p = 0.046). All other significant effects and interactions were not related to the status of deviancy, but to the physical differences in stimuli or to differences between the electrode clusters depending on the side of presentation and on the specific feature presented. Thus, we found a main effect of FEATURE (F(1,15) = 13.39, p = 0.001) reflecting more negative mean amplitudes in response to displays containing a location asymmetry (− 5.9 V) compared to displays containing a color asymmetry (− 4.9 V, p = 0.001) or a shape asymmetry (− 4.8 , p = 0.02), a main effect of CLUSTER (F(1,15) = 5.82, p = 0.029) indicating that, in general, larger mean amplitudes were elicited over the right occipital cluster of electrodes (− 6.1 V compared to − 4.2 V over the left cluster), and an interaction between CLUSTER and SIDE (F(1,15) =15.78, p = 0.001). These effects were qualified however by the 3-way interaction between FEATURE, CLUSTER, and SIDE (F(1,15) = 11.4430, p = 0.001) which reflects the fact that only for displays with a location asymmetry we found an interaction between SIDE and CLUSTER (F(1,15) =15.81, p = 0.001; larger over occipital electrode sites contra-lateral to the side of presentation of the critical feature), but not for color asymmetries (F(1,15) = 2.07, p = 0.171) or shape asymmetries (F(1,15) = 0.09, p = 0.763) (compare bar charts in Fig. 2, Fig. 3 and Fig. 4, respectively). The late time window overlapped the first part of a negative-going (but not actually negative) deflection in the ERPs. Since this deflection follows a clear P2 response, we refer to it as an N2. Deviant stimulus displays elicited more negative going potentials at this window (1.9 μV) than control stimulus displays (2.6 μV) as indicated by a main effect of STIMULUS TYPE (F(1,15) = 9.16, p = 0.008). All further effects were again only related to differences between the physical features: main effect of FEATURE (F(2,30) = 7.80, p = 0.002), and differences between the electrode clusters dependent on the critical feature and the side of presentation (interaction of FEATURE, CLUSTER and SIDE (F(2,30) = 3.62, p = 0.039). 1 To summarize the results involving the STIMULUS TYPE factor (i.e., related to change detection), a deviance-related response was found in both time windows, but was more pronounced on the right electrode cluster for the early time window.