Reading skill modulates the effect of parafoveal distractors on foveal lexical decision in deaf students

Reading skill modulates the effect of parafoveal distractors on foveal lexical decision in deaf... a1111111111 In low-level perceptual tasks and reading tasks, deaf individuals show a redistribution of spatial visual attention toward the parafoveal and peripheral visual fields. In the present study, the experiment adopted the modified flanker paradigm and utilized a lexical decision task to investi- gate how these unique visual skills may influence foveal lexical access in deaf individuals. It OPENACCESS was predicted that irrelevant linguistic stimuli presented in parafoveal vision, during a lexical Citation: Tao J, Qin Z, Meng Z, Zhang L, Liu L, Yan decision task, would produce a larger interference effect for deaf college student readers if the G, et al. (2019) Reading skill modulates the effect of parafoveal distractors on foveal lexical decision stimuli acted as distractors during the task. The results showed there was a larger interference in deaf students. PLoS ONE 14(9): e0221891. effect in deaf college student readers compared to the interference effect observed in partici- https://doi.org/10.1371/journal.pone.0221891 pants with typical levels of hearing. Furthermore, deaf college student readers with low-skilled Editor: Susan Nittrouer, University of Florida, reading levels showed a larger interference effect than those with high-skilled reading levels. UNITED STATES The current study demonstrates that the redistribution of spatial visual attention toward the par- Received: March 25, 2019 afoveal visual regions in deaf students impacts foveal lexical processing, and this effect is mod- Accepted: August 16, 2019 ulated by reading skill. The findings are discussed in relation to the potential effect that enhanced parafoveal attention may have on everyday reading for deaf individuals. Published: September 12, 2019 Copyright:© 2019 Tao et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in Introduction any medium, provided the original author and Many studies employing low-level visual cognitive tasks suggest that, for deaf individuals, source are credited. visual attention is reorganized to compensate for the lack or absence of auditory input pro- Data Availability Statement: All relevant data are vided by a complex environment [1], and, this reorganization has been thought to result in a within the manuscript and its Supporting redistribution of spatial visual attention toward the parafoveal and peripheral visual fields in Information files. deaf individuals [2, 3]. Funding: This research was supported by National In support of this hypothesis, deaf individuals have been shown to respond more rapidly to Social Science Fund of China (16BYY074) grant flashed or moving stimuli in the periphery [4, 5], and they have been reported to have awarded to Guoli Yan. The funder had no role in study design, data collection and analysis, decision enhanced visual selective attention in the peripheral visual field [6]. There is also evidence to publish, or preparation of the manuscript. from fMRI studies that the middle temporal-medial superior temporal area of deaf individuals shows a greater response to peripheral motion stimuli compared to individuals with typical Competing interests: The authors have declared that no competing interests exist. levels of hearing [7, 8]. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 1 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students However, in some circumstances increased parafoveal or peripheral visual attention might hamper foveal visual field performance. For example, Proksch and Bavelier have shown that deaf individuals exhibit greater distractibility from parafoveal distractors in target detection tasks [9], and reduced discrimination performance was reported when a central discrimination task and a peripheral localization task had to be carried out simultaneously [10]. These studies, using low- level cognitive tasks, have shown that deaf individuals are more distracted by the presentation of task-irrelevant information in peripheral vision when the task requirements demand a central focus. An important question leading from these studies is, to what extent increased parafoveal processing in deaf individuals might impact an everyday task, such as reading. It is well documented that the illiteracy rate of deaf individuals has been very high for decades. The average reading proficiency of young deaf adults graduating from high school has been reported to be far behind that of individuals with typical levels of hearing [11]. It is also well known that there is a strong link between visual attention and reading effi- ciency in individuals with typical levels of hearing [12]. For example, it has been shown that selective spatial attention is significantly associated with reading efficiency [13], and that visual-spatial attention can effectively predict reading achievement [14] in hearing individuals. Foveal information (the central 2˚ of the visual field) is essential for word processing and for reading [15]. However, eye movement studies in reading have also indicated that linguistic information can be obtained from the parafoveal region (from the foveal region up to 5˚ of visual angle on either side of fixation). If readers’ allocate attentional resources to process par- afoveal words, then reading efficiency can be increased, but, paradoxically, if too much atten- tion is allocated to process parafoveal information, then this may actually hinder identification of the foveal word under inspection [16]. Therefore, readers have to carefully allocate their attention between the foveal region and parafoveal region, in order to obtain parafoveal infor- mation that is useful (facilitatory), but not harmful (inhibitory) for reading [17]. Given the relationship between foveal and parafoveal processing in reading, it has been speculated that the redistribution of spatial visual attention toward the parafoveal visual fields, observed for deaf individuals, may actually slow down foveal processing, and hence cause read- ing difficulties for deaf individuals [18]. In support of that view, some studies have shown that deaf readers, including Chinese deaf readers, can access parafoveal information more effi- ciently during sentence reading [19, 20]. However, although these studies show deaf individu- als to have enhanced access to parafoveal information more efficiently during sentence reading, they cannot tell us how visual processing of parafoveal information may influence foveal lexical access during reading in deaf readers. In this paper, we investigate whether the redistribution of spatial visual attention toward the parafoveal visual fields in deaf individuals impacts foveal lexical access for a higher-level cognitive task (lexical decision), with an ultimate aim of relating how the findings might affect the everyday task of reading. The current study adopted the modified flanker paradigm [21] and utilized a lexical deci- sion task to investigate the effects of task-irrelevant linguistic stimuli, presented in parafoveal locations, on central lexical processing in deaf individuals. If the redistribution of spatial visual attention toward the parafoveal visual fields in deaf readers affects the lexical processing of stimuli under foveal inspection, it would be predicted that deaf participants should take longer to make a lexical decision when a target character is presented centrally along with a distractor character simultaneously presented in parafoveal vision. We, therefore, aimed to test whether interference effects were greater for deaf college student readers compared to participants with typical levels of hearing. A further aim of this experiment was to explore whether any observed interference effects for the deaf readers were modulated by reading level ability. Although there is no empirical PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 2 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students evidence to support that claim, it could be the case that high-skilled deaf readers are able to inhibit irrelevant distractors more effectively than low-skilled deaf readers. If this is the case we would predict a greater interference effect for low-skilled deaf readers compared to high- skilled deaf readers. One of the challenges in the field is to uncover factors that distinguish skilled deaf readers from unskilled deaf readers [22]. Control over the redistribution of spatial visual attention may be a factor that distinguishes between these two groups of deaf readers. Method Ethics statement The procedures of the study have been approved by the Ethical Committee of Tianjin Normal University. All deaf college student participants and all college students with typical levels of hearing provided informed written consent prior to inclusion in the study. Parents provided informed written consent for middle school students with typical levels of hearing. Participants Forty severely to profoundly deaf college students (M = 20.69 years, SD = 1.25) were age recruited from the Technical College for the Deaf, Tianjin University of Technology, China. The deaf participants’ hearing losses were� 75dB (M = 100.45 dB, SD = 12.61) in the better ear. Participants were either born deaf or became deaf before three years of age. None of the participants had received a cochlear implant. Thirty-five deaf participants reported sign lan- guage as their primary and preferred language of communication. Five deaf participants reported spoken language as their primary and preferred language of communication. Demo- graphic information of participants was collected by a questionnaire survey. The information was provided by the participants themselves. The power analysis of the current study for an average effect size of d = 0.45 is 0.97, and that power value is greater than the minimum recom- mended level of 0.80 [23, 24]. A reading control group consisted of forty Chinese middle school students with typical lev- els of hearing (M = 13.98 years, SD = 0.79). We matched their reading comprehension level age on the Chinese Proficiency Test for Deaf College Students [25, 26]. This test was developed for deaf college students and consists of three parts: sentence comprehension, short text compre- hension, and ancient text comprehension. The test has a total of 50 multiple choice questions, with 2 points given for each question. The completion time is 30 minutes. There was no signif- icant difference in reading comprehension scores between the deaf college students (M = 67.03, SD = 5.63) and the reading control group (M = 66.48, SD = 7.92), t (78) = 0.36, p = .72. There was no significant difference in reading fluency [20] scores between the deaf college students (M = 342.34 characters/min, SD = 111.98) and the reading control group (M = 330.78 characters/min, SD = 101.20), t (78) = 0.49, p = .63. An age control group consisted of forty Chinese college students with typical levels of hear- ing (M = 20.49 years, SD = 1.24). There was no significant difference in chronological age age between the deaf college students and the age control group, t (78) = – 0.71, p = .48. There was a significant difference in reading comprehension scores between the deaf college students (M = 67.03, SD = 5.63) and the age control group (M = 86.15, SD = 5.33), t (78) = 15.61, p < .001, d = 3.49 [25, 26]. There was a significant difference in reading fluency [20] scores between the deaf college students (M = 342.34 characters/min, SD = 111.98) and the age control group (M = 458.23 characters/min, SD = 89.58), t (78) = 5.11, p < .001, d = 1.14. In addition, we matched participants from the three groups on nonverbal IQ using the Raven Test [27, 28]. There was no significant difference in standard score among the three groups, F (2,117) = 0.20, p = .82. All participants had normal or corrected-to-normal vision. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 3 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Materials There were 360 trials (see S1 Materials), of which 180 trials presented pseudo-Chinese charac- ters and were used as filler materials [29, 30]; Eighty percent of trials were target-distractor pairs, and the other 20 percent of trials were without distractors. For the target-distractor pairs, twenty undergraduate students with typical levels of hearing, who did not participate in the experiment, evaluated the relatedness of target characters and distractor characters on a 5 point scale (1 = very dissimilar, 5 = very similar). The targets and distractors shared no orthographic (M = 1.39, SD = 0.38), phonological (M = 1.19, SD = 0.26), or semantic (M = 1.37, SD = 0.38) similarities. The target characters mean frequency was 145.99 per million (SD = 430.01) and the distractor characters mean frequency was 120.72 per million (SD = 548.35), and the target characters and distractor characters mean number of strokes was 8.72 (SD = 3.23) and 8.74 (SD = 3.00), respectively [31]. Given that the parafoveal region is defined as up to 5˚ visual angle on either side of fixation, the parafoveal presentation locations for the study were operationalized as 2˚, 3˚, or 4˚ visual angle from the center of the display on either side. Half of the distractors were presented to the left parafoveal visual field and half to the right. The negative sign and positive sign represent the left and the right of the central fixation point, respectively. Design The study used a 3 (Group: deaf college student group, reading control group, age control group) ×7 (Interference condition: without interference, –4˚, –3˚, –2˚, + 2˚, +3˚, +4˚) mixed experimental design. The group was a between-subjects factor and interference condition was a within-subjects factor. Procedure Participants were seated 45cm from the 14-inch monitor of a laptop (with a 60-Hz refresh rate, 1024×768 resolution). The pixel value of each character was 34.67 × 34.67. At this distance, one character subtended about 1˚ visual angle [32]. A software package (Eprime2.0) was used to run the program and record reaction times (RTs). Each trial began with a cross at the center of the screen for 500ms. The target character was then presented at the center of the screen (with/without a distractor character being presented on either side of fixation at different locations, depending on the interference condition). Par- ticipants were asked to focus on the central fixation and were asked to make a lexical decision for the foveal target as quickly and accurately as possible when this was presented, regardless of whether a parafoveal distractor was simultaneously presented with the target. Participants used the right or left index finger to press the F or J on the keyboard, indicating character (word) or pseudo-character (non-word), respectively. Participants had a maximum of 3s to respond [32]. Each target presentation trial was followed by a random interval 1000~1300ms before the next trial sequence began; Fig 1 shows a schematic of the trial sequence. There was a break after every 72 trials and the experiment lasted approximately 30 minutes. Data analysis The analysis of RTs excluded pseudo-Chinese character filler trials, incorrectly answered trials (4.35%), and trials with an RT which were beyond 2.5 standard deviations (2.82%) from each participant’s mean RT (see S1 Dataset). Analysis of error rates excluded pseudo-Chinese char- acter filler trials and trials which were beyond 2.5 standard deviations (2.74%) from each par- ticipant’s mean RT [29, 30]. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 4 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Fig 1. Example of distractor trial sequence. Fig 1 represents a trial of +3˚ interference condition. The target character (举, lift) was presented at the center of the screen and the distractor character (址, address) was presented at 3˚ visual angle of right side from the center of the display. https://doi.org/10.1371/journal.pone.0221891.g001 Data were analyzed using the lme4 package [33], within the R environment for statistical computing [34]. Generalized linear mixed models were used to analyze error rate data (binary variables) [35]. We report b values, standard errors (SE) and t values (z values for error rates data), with |t|� 1.96 (|z|� 1.96) deemed significant at the 0.05 alpha level. In the model, partici- pants and items were specified as crossed random effects [36]. Group and interference condition were specified as fixed effects. To assess differences between pairs of groups, we set up two suc- cessive difference contrasts: comparing the age control group to the deaf college student group, and the deaf college student group to the reading control group. Analysis of interference effects was based on planned comparisons for each of the different interference conditions and the without-interference (single target) condition, which was used as a reference condition. In this model, the interactions represent the interference effect for a given pair of groups. If the maxi- mal random effects structure model failed to converge, the random structure was trimmed, and then we used a likelihood ratio test to verify the best fit of the converged models to the data [37]. Results Error rates When different interference conditions were compared to the without-interference condition separately, the random effects structure model failed to converge. Therefore, we compared dif- ferent interference conditions as a whole to the without-interference condition and the model including participant and item random intercepts converged. The deaf college student group average error rate was 3.63%; the age control group average error rate was 3.13%; the reading control group average error rate was 5.86%; See Table A in S1 Appendix. As shown in Table 1, the deaf college student group did not differ in their error rate with the age control group (b = 0.20, SE = 0.19, z = 1.05); the deaf college student group and the reading control group was significantly different in error rate (b = 0.53, SE = 0.18, z = 2.96), and the interference effect main effect was not significant (b = 0.09, SE = 0.17, z = 0.51). The deaf college student group and the age control group did not significantly interact with the interference effect (b = – 0.11, SE = 0.23, z = – 0.48), and the deaf college student group and the reading control group did not significantly interact with the interference effect (b = 0.14, PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 5 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students SE = 0.20, z = 0.69). Thus the only significant result for the error rate analysis, in relation to the deaf college student group, showed that the reading control group made more errors com- pared to the deaf college students, and this finding can most likely be explained by age differ- ences between those two groups. Reaction times A model that included participant random slope alongside participant and item random inter- cepts was deemed the best model for analysis, based on the likelihood-ratio test result [36]. As shown in Table 2, the group main effects did not differ: The deaf college student group did not differ in their RTs (see Table B in S1 Appendix) from the age control group (b = 26.03, SE = 28.78, t = 0.90); and the deaf college student group did not differ in their RTs from the reading control group (b = 1.48, SE = 28.78, t = 0.05). The interference effect main effects were significant in all interference conditions, all |ts| >1.96, showing that distractors presented at all eccentricities resulted in longer RTs compared to when no distractors were presented (the without distractor condition). As shown in Table 2, the deaf college student group and age control group significantly inter- acted with the interference effect in all interference conditions, all |ts| >1.96, indicating that the interference effect was larger for the deaf college student group than for age control group; see Fig 2. The deaf college student group and reading control group significantly interacted with the interference effect in all interference conditions, all |ts| >1.96, indicating that the interference effect was also larger for the deaf college student group than for the reading control group [20]. In summary, the results showed that the presence of a parafoveal lexical distractor resulted in poorer performance (longer RTs) for a lexical decision task in the deaf participants com- pared to the participants with typical levels of hearing. A further aim of this study was to investigate whether reading skill modulated observed interference effects in deaf individuals. We tested this hypothesis by comparing interference effects between different reading levels of deaf individuals. If deaf individuals with low-skilled reading levels allocate more attentional resources to process parafoveal information than deaf individuals with high-skilled reading levels, we predicted that low-skilled deaf readers in the current study should have taken longer to make a lexical decision compared to high-skilled deaf readers in the current study. Analyses of interference effects for deaf readers with high-skilled reading levels versus deaf readers with low-skilled reading levels We separated the deaf college students into individuals with high-skilled reading levels and individuals with low-skilled reading levels, using a median split technique [38] based on their Table 1. Groups × interference conditions results for error rates (%). b SE z Intercept -3.61 0.11 -34.11 Deaf college student group vs. Age control group 0.20 0.19 1.05 Reading control group vs. Deaf college student group 0.53 0.18 2.96 Interference vs. Without interference 0.09 0.17 0.51 Deaf college student group vs. Age control group × Interference vs. Without interference -0.11 0.23 -0.48 Reading control group vs. Deaf college student group × Interference vs. Without interference 0.14 0.20 0.69 Note. Deaf college student group: forty severely to profoundly deaf college students. Age control group: forty Chinese college students with typical levels of hearing. Reading control group: forty Chinese middle school students with typical levels of hearing. |z|� 1.96 deemed significant at the 0.05 alpha level. Significant z-values are shown in bold. https://doi.org/10.1371/journal.pone.0221891.t001 PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 6 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Table 2. Groups × interference conditions results for RTs (ms). b SE t Intercept 654.93 12.07 54.27 Deaf college student group vs. Age control group 26.03 28.78 0.90 Reading control group vs. Deaf college student group 1.48 28.78 0.05 –2˚vs.Without interference -92.76 10.45 -8.88 –3˚vs.Without interference -79.64 11.01 -7.24 –4˚vs.Without interference -79.87 10.71 -7.46 +2˚vs.Without interference -79.28 10.13 -7.83 +3˚vs.Without interference -73.46 10.55 -6.96 +4˚vs.Without interference -74.70 10.51 -7.11 Deaf college student group vs. Age control group × –2˚vs.Without interference -59.62 19.38 -3.08 Reading control group vs. Deaf college student group × –2˚vs.Without interference 58.12 19.46 2.99 Deaf college student group vs. Age control group × –3˚vs.Without interference -79.32 21.16 -3.75 Reading control group vs. Deaf college student group × –3˚vs.Without interference 75.05 21.23 3.53 Deaf college student group vs. Age control group × –4˚vs.Without interference -88.90 20.23 -4.39 Reading control group vs. Deaf college student group× –4˚vs.Without interference 86.14 20.30 4.24 Deaf college student group vs. Age control group × +2˚vs.Without interference -63.58 18.34 -3.47 Reading control group vs. Deaf college student group× +2˚vs.Without interference 55.50 18.40 3.02 Deaf college student group vs. Age control group × +3˚vs.Without interference -83.47 19.73 -4.23 Reading control group vs. Deaf college student group × +3˚vs.Without interference 88.39 19.80 4.46 Deaf college student group vs. Age control group × +4˚vs.Without interference -80.94 19.59 -4.13 Reading control group vs. Deaf college student group × +4˚vs.Without interference 78.34 19.64 3.99 Note. Deaf college student group: forty severely to profoundly deaf college students. Age control group: forty Chinese college students with typical levels of hearing. Reading control group: forty Chinese middle school students with typical levels of hearing. Interference conditions: The negative sign and positive sign represent the left and the right of the central fixation point, respectively. |t|� 1.96 deemed significant at the 0.05 alpha level. Significant t-values are shown in bold. https://doi.org/10.1371/journal.pone.0221891.t002 reading comprehension [25, 26] test scores (Given that this test was developed for deaf college students, it can be argued to reflect and distinguish reading ability of deaf college students). The two groups comprised a low/medium skilled group (n = 19) with reading comprehension test scores < 68.00, and a high-skilled group (n = 21) with all scores� 68.00. There was a sig- nificant difference in reading comprehension scores between the low-skilled deaf group (M = 62.05, SD = 3.37) and the high-skilled deaf group (M = 71.52, SD = 2.60), t (38) = 10.00, p < .001, d = 3.17. In the low-skilled deaf group, seventeen deaf participants reported sign language as their primary and preferred language of communication, and two deaf participants reported spoken language as their primary and preferred language of communication. In the high- skilled deaf group, eighteen deaf participants reported sign language as their primary and pre- ferred language of communication, and three deaf participants reported spoken language as their primary and preferred language of communication. In relation to the hearing loss of the deaf readers, there was no significant difference in the better ear between the low-skilled deaf group (M = 100.32 dB, SD = 12.13) and the high-skilled deaf group (M = 100.57 dB, SD = 13.32), t (38) = 0.06, p = .95. There was no significant differ- ence in chronological age between the low-skilled deaf group (M = 20.71, SD = 1.42) and the high-skilled deaf group (M = 20.68, SD = 1.12), t (38) = – 0.09, p = .93. Finally, there was no significant difference in IQ standard score between the low-skilled deaf group (M = 101.32, SD = 14.99) and the high-skilled deaf group (M = 109.83, SD = 13.51), t (38) = 1.89, p = .07. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 7 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Fig 2. Comparison of the interference effect for three participant groups in different interference conditions. Deaf college student group: forty severely to profoundly deaf college students. Age control group: forty Chinese college students with typical levels of hearing. Reading control group: forty Chinese middle school students with typical levels of hearing. Interference conditions: The negative sign and positive sign represent the left and the right of the central fixation point, respectively. Interference effect (ms) = RTs with interference—RTs without interference (single target). Error bars represent the standard errors of the means. https://doi.org/10.1371/journal.pone.0221891.g002 Error rates When different interference conditions were compared to the without-interference condition separately, the random effects structure model failed to converge. Therefore, we compared dif- ferent interference conditions as a whole to the without-interference condition. The low-skilled deaf group average error rate was 3.74%; the high-skilled deaf group aver- age error rate was 3.53%; See Table C in S1 Appendix. As shown in Table 3, the low-skilled deaf group did not differ in their error rate from the high-skilled deaf group (b = – 0.16, SE = 0.28, z = – 0.56). The interference effect main effect was not significant (b = – 0.03, SE = 0.27, z = – 0.13), and the low-skilled deaf group and the high-skilled deaf group did not significantly interact with the interference effect (b = 0.51, SE = 0.33, z = 1.54). Reaction times A model that included participant random slope alongside participant and item random inter- cepts was deemed the best model for analysis, based on the likelihood-ratio test result [36]. Table 3. Groups × interference conditions results for error rates (%). b SE z Intercept -3.94 0.19 -21.00 Low-skilled deaf group vs. High-skilled deaf group -0.16 0.28 -0.56 Interference vs. Without interference -0.03 0.27 -0.13 Low-skilled deaf group vs. High-skilled deaf group × Interference vs. Without interference 0.51 0.33 1.54 Note. Low-skilled deaf group: nineteen severely to profoundly deaf college students with reading comprehension test scores < 68.00. High-skilled deaf group: twenty-one severely to profoundly deaf college students with scores� 68.00.|z|� 1.96 deemed significant at the 0.05 alpha level. Significant z-value is shown in bold. https://doi.org/10.1371/journal.pone.0221891.t003 PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 8 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Table 4. Groups × interference conditions results for RTs (ms). b SE t Intercept 666.07 25.48 26.15 Low-skilled deaf group vs. High-skilled deaf group 102.47 50.45 2.03 –2˚vs.Without interference –135.48 21.17 –6.40 –3˚vs.Without interference –133.41 23.57 –5.66 –4˚vs.Without interference –140.12 22.24 –6.30 +2˚vs.Without interference –121.20 20.27 –5.98 +3˚vs.Without interference –132.59 20.74 –6.39 +4˚vs.Without interference –130.25 20.60 –6.32 Low-skilled deaf group vs. High-skilled deaf group × –2˚vs.Without interference –82.65 38.47 –2.15 Low-skilled deaf group vs. High-skilled deaf group × –3˚vs.Without interference –76.38 43.70 –1.75 Low-skilled deaf group vs. High-skilled deaf group × –4˚vs.Without interference –61.14 40.81 –1.50 Low-skilled deaf group vs. High-skilled deaf group × +2˚vs.Without interference –85.66 36.49 –2.35 Low-skilled deaf group vs. High-skilled deaf group × +3˚vs.Without interference –63.67 37.51 –1.70 Low-skilled deaf group vs. High-skilled deaf group × +4˚vs.Without interference –70.38 37.20 –1.89 Note. Low-skilled deaf group: nineteen severely to profoundly deaf college students with reading comprehension test scores < 68.00.High-skilled deaf group: twenty-one severely to profoundly deaf college students with scores� 68.00. Interference conditions: The negative sign and positive sign represent the left and the right of the central fixation point, respectively.|t|� 1.96 deemed significant at the 0.05 alpha level. Significant t-values are shown in bold. https://doi.org/10.1371/journal.pone.0221891.t004 As shown in Table 4, the low-skilled deaf group had longer RTs (see Table D in S1 Appen- dix) than the high-skilled deaf group (b = 102.47, SE = 50.54, t = 2.03), and the interference effect main effects were significant in all interference conditions, all |ts| >1.96. As shown in Table 4, the two groups of deaf readers significantly interacted with the inter- ference effects for 2˚ conditions on either side of fixation; see Fig 3, indicating that the Fig 3. Comparison of the interference effect for two deaf reader groups in different interference conditions. Low- skilled deaf group: nineteen severely to profoundly deaf college students with reading comprehension test scores < 68.00. High-skilled deaf group: twenty-one severely to profoundly deaf college students with scores� 68.00. Interference conditions: The negative sign and positive sign represent the left and the right of the central fixation point, respectively. Interference effect (ms) = RTs with interference—RTs without interference (single target). Error bars represent the standard errors of the means. https://doi.org/10.1371/journal.pone.0221891.g003 PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 9 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students interference effect was larger for the low-skilled deaf group (for –2˚ vs. without interference, 171.63 ms; for +2˚ vs. without interference, 163.98 ms) than for the high-skilled deaf group (for –2˚vs. without interference, 95.05 ms; for +2˚vs. without interference, 78.07 ms). The two groups of deaf readers marginally significantly interacted with the interference effects for 3˚ interference conditions on either side of fixation and for the +4˚ interference condition, indi- cating that the interference effect was larger for the low-skilled deaf group (for –3˚ vs. without interference, 165.60 ms; for +3˚ vs. without interference, 161.15 ms; for +4˚ vs. without inter- ference,163.10 ms) than for the high-skilled deaf group (for –3˚ vs. without interference, 95.59 ms; for +3˚ vs. without interference, 98.07 ms; for +4˚ vs. without interference, 98.36 ms) in the interference conditions. For the –4˚ interference condition, the low-skilled deaf group (for –4˚vs. without interference,168.68 ms) show an obvious data trend of a larger interference effect than for the high-skilled deaf group (for –4˚ vs. without interference, 109.50 ms) but this numerical difference was not statistically significant [20]. In summary, separating the deaf students into high and low-skilled deaf readers resulted in differences in the magnitude of distractor effects on the time taken to make a lexical decision, with low-skilled readers taking longer than high-skilled readers. Since there were equivalent numbers of deaf students who rely on sign language as their preferred way to communicate, it is unlikely that this preference impacted on the modulating effects of reading skill on distractor interference. One interesting question that arises from this finding relates to whether the increased inter- ference effect for low-skilled readers is unique to deaf readers, or whether the same effect is observed for hearing readers? To address this question, we carried out the following analyses for the reading matched hearing participants. Analyses of interference effects for reading-matched hearing participants with high-skilled reading levels versus low-skilled reading levels We separated the reading-matched hearing middle school participants into individuals with high-skilled reading levels and individuals with low-skilled reading levels, using a median split technique [38] based on their reading comprehension [25, 26] test scores. The two groups comprised a low/medium skilled group (n = 21) with reading comprehension test scores� 64.00, and a high-skilled group (n = 19) with scores > 64.00. There was a significant difference in reading comprehension test scores between the low-skilled group (M = 60.38, SD = 3.44) and the high-skilled group (M = 73.21, SD = 5.61), t (38) = 8.81, p< .001, d = 2.79. There was no significant difference of chronological age between the low-skilled group (M = 13.83, SD = 0.79) and the high-skilled group (M = 14.15, SD = 0.77), t (38) = 1.32, p = .20; and no significant difference in IQ standard score between the low-skilled group (M = 101.95, SD = 9.75) and the high-skilled group (M = 108.00, SD = 11.55), t (38) = 1.80, p = .08. Error rates We compared different interference conditions as a whole to the without-interference condi- tion. The average error rate of the low-skilled group was 6.67%; the average error rate of the high-skilled group was 4.94%; See Table E in S1 Appendix. As shown in Table 5, the low-skilled group did not differ in their error rate from the high- skilled group (b = 0.39, SE = 0.22, z = 1.81). There was no significant main effect of interference (b = 0.14, SE = 0.19, z = 0.71), and the low-skilled group and the high-skilled group did not sig- nificantly interact with the interference effect (b = 0.00, SE = 0.26, z = -0.02). PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 10 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Table 5. Groups × interference conditions results for error rates (%). b SE z Intercept -3.17 0.13 -23.65 Low-skilled group vs. High-skilled group 0.39 0.22 1.81 Interference vs. Without interference 0.14 0.19 0.71 Low-skilled group vs. High-skilled group × Interference vs. Without interference 0.00 0.26 -0.02 Note. Low-skilled group: twenty-one hearing middle school participants with reading comprehension test scores� 64.00. High-skilled group: nineteen hearing middle school participants with scores > 64.00.|z|� 1.96 deemed significant at the 0.05 alpha level. Significant z-value is shown in bold. https://doi.org/10.1371/journal.pone.0221891.t005 Reaction times The maximal random effects structure, including participant and item random intercepts, converged. As shown in Table 6, the low-skilled group did not differ in their RTs (see Table F in S1 Appendix) from the high-skilled group (b = 33.30, SE = 28.20, t = 1.18), and the interfer- ence effect main effects were significant in all interference conditions, all |ts| >1.96. As shown in Table 6, the low-skilled group and the high-skilled group did not significantly interact with the interference effect in five condtions (|ts| < 1.96). The two groups of middle school participants significantly interacted with the interference effects for +4˚ condition; see Fig 4, indicating that the interference effect was larger for the low-skilled group (for +4˚ vs. without interference, 64.20 ms) than for the high-skilled group (for +4˚ vs. without interfer- ence, 34.74 ms). For the +2˚ and +3˚ interference conditions, the low-skilled group (for +2˚ vs. without interference, 76.30 ms; +3˚ vs. without interference, 54.66 ms) show a data trend of a larger interference effect than for the high-skilled group (for +2˚ vs. without interference, 50.08 ms; +3˚ vs. without interference, 32.18 ms), but this numerical difference was not statisti- cally significant [20]. Table 6. Groups × interference conditions results for RTs (ms). b SE t Intercept 663.30 14.34 46.26 Low-skilled group vs. High-skilled group 33.30 28.20 1.18 –2˚vs.Without interference -74.93 9.22 -8.13 –3˚vs.Without interference -57.62 9.21 -6.25 –4˚vs.Without interference -53.22 9.21 -5.78 +2˚vs.Without interference -63.91 9.21 -6.94 +3˚vs.Without interference -42.98 9.20 -4.67 +4˚vs.Without interference -48.87 9.18 -5.33 Low-skilled group vs. High-skilled group × –2˚vs.Without interference -7.65 13.14 -0.58 Low-skilled group vs. High-skilled group × –3˚vs.Without interference -4.50 13.15 -0.34 Low-skilled group vs. High-skilled group × –4˚vs.Without interference 5.29 13.11 0.40 Low-skilled group vs. High-skilled group × +2˚vs.Without interference -21.71 13.14 -1.65 Low-skilled group vs. High-skilled group × +3˚vs.Without interference -22.20 13.09 -1.70 Low-skilled group vs. High-skilled group × +4˚vs.Without interference -28.10 13.02 -2.16 Note. Low-skilled group: twenty-one hearing middle school participants with reading comprehension test scores� 64.00. High-skilled group: nineteen hearing middle school participants with scores > 64.00. Interference conditions: The negative sign and positive sign represent the left and the right of the central fixation point, respectively.|t|� 1.96 deemed significant at the 0.05 alpha level. Significant t-values are shown in bold. https://doi.org/10.1371/journal.pone.0221891.t006 PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 11 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Fig 4. Comparison of the interference effect for two hearing middle school reader groups in different interference conditions. Low-skilled group: twenty-one hearing middle school participants with reading comprehension test scores� 64.00. High-skilled group: nineteen hearing middle school participants with the scores > 64.00. Interference conditions: The negative sign and positive sign represent the left and the right of the central fixation point, respectively. Interference effect (ms) = RTs with interference—RTs without interference (single target). Error bars represent the standard errors of the means. https://doi.org/10.1371/journal.pone.0221891.g004 The pattern for the comparison between interference effects observed in the high-skilled and low-skilled deaf students appears to be different to that observed for the same comparison in the hearing students. In the deaf students interference effects are greater for the low-skilled readers when distractors are presented at different parafoveal eccentricities both to the right and to the left of fixation. However, for the hearing students, greater interference effects for the low-skilled read- ers in that group are obsreved exclusiveley for distractors presented to the right of fixation. Discussion The aim of this study was to investigate whether the well documented redistribution of spatial visual attention toward the parafoveal visual fields in deaf individuals would influence the time taken to process lexical stimuli under foveal inspection. Firstly, we predicted that deaf partici- pants should take longer to make a lexical decision when a target character was presented cen- trally along with a distractor character simultaneously presented in parafoveal vision. Secondly, we proposed that low-skilled deaf readers would show a greater interference effect compared to high-skilled deaf readers. The results support our experimental hypotheses. We found a larger parafoveal interference effect for deaf college students (overall) than for participants with typical levels of hearing. We interpreted this finding to result from deaf individuals allocating more attentional resources to process parafoveal information. Hence, the redistribution of spatial visual attention in the deaf readers affects lexical processing of stimuli under foveal inspection. In reading, the goal is to process the individual words sequentially and to integrate the meaning of these words in order to build up a full representation and comprehensive under- standing of the text [39]. If readers’ allocate attentional resources to process parafoveal words, then reading efficiency can be increased, but, if too much attention is allocated to process par- afoveal information, then this may actually hinder identification of the foveal word under inspection [16]. Therefore, readers have to carefully allocate their attention between the foveal region and parafoveal region, in order to obtain parafoveal information that is useful (facilita- tory), but not harmful (inhibitory) for the reading process [17]. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 12 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Some studies have found that deaf individuals can access parafoveal information more effi- ciently during sentence reading [19, 20], however, little is yet known about how visual process- ing of parafoveal information may influence foveal lexical access during reading for deaf readers. The findings from the current study support the hypothesis of the redistribution of spatial visual attention toward the parafoveal visual fields in deaf individuals. An obvious ques- tion raised by the findings would be to investigate whether the observed slow-down of foveal processing for lexical decision tasks could impact in everyday reading difficulties in deaf indi- viduals, and whether any difficulties would be modulated by reading level in the deaf individuals. However, it should also be noted that in normal reading parafoveal information is often related to foveal information, and hence has the potential to speed foveal processing. In that case it is possible that deaf readers might show a larger priming effect than hearing readers for parafoveal facilitatory information. Since the current study only tested an interfering stimulus (an unrelated character), the question as to whether greater parafoveal attention could enhance reading processes in the deaf readers remains to be tested. Furthermore, the interference effects observed for the deaf students were modulated by reading level. Low-skilled deaf readers showed a greater interference effect compared to high- skilled deaf readers. This finding suggests that there may be differences in the way that low- skilled reading deaf individuals and high-skilled reading deaf individuals allocate attentional resources to parafoveal information. In order to test whether that finding is unique to deaf readers, we also separated the read- ing-matched hearing middle school participants into individuals with high-skilled reading lev- els and individuals with low-skilled reading levels. The low-skilled middle school participants showed a greater interference effect (although not all effects were significant) compared to high-skilled middle school readers in interference conditions where distractors were presented to the right of fixation. This finding is observed exclusively for distractors presented to the right of fixation in low-skilled hearing individuals, whereas the effect is observed for distrac- tors presented to the left and to the right of fixation in the low-skilled deaf readers. We inter- pret this finding to reflect a general effect of reading skill on modulating attention to the parafovea, but also that this modulation will be different for hearing and deaf individuals. A study of lexical decision flanker tasks with hearing readers has revealed a rightward bias that is argued to be related to reading direction [30], and in sentence reading tasks, the size of the per- ceptual span is asymmetric [15, 17] to the right for hearing individuals. Therefore, it is likely that an effect of reading skill would be detected for distractors presented to the right of fixation for hearing readers, but for deaf individuals an effect of reading skill would be detected for dis- tractors presented to the right and the left of fixation. Compared with hearing readers, deaf individuals allocate more attentional resources to process parafoveal information presented to the left and to the right of fixation. This is in line with the findings from the current lexical decision study which clearly demonstrates an effect of reading skill on the ability to ignore par- afoveal distractors, both to the left and to the right of fixation, for deaf readers. A related finding from a case study of a hearing reader who was diagnosed with develop- mental dyslexia showed that the main reason for his reading problem was a selective atten- tional deficit, whereby letters from words available in parafoveal vision interfered with his processing of the currently fixated word [12]. This finding is similar to the results from the cur- rent study for deaf readers. An important question raised by the findings relates to how the ability to control what is attended to and processed in the parafovea might affect reading skill development in deaf individuals. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 13 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Conclusion In sum, the current study confirms that deaf readers show a larger interference effect from irrelevant parafoveal distractors, during a lexical decision task. This is likely to result from the greater distribution of spatial visual attention toward the parafoveal visual field. Furthermore, low-skilled deaf readers show a larger interference effect which suggests that the allocation of attentional resources toward the parafovea, may reflect less volitional control in low-skilled deaf readers. The findings lead us to speculate that in deaf individuals, the ability to control what is processed or inhibited in the parafovea has the potential to impact in everyday reading in the deaf population. Supporting information S1 Appendix. (DOCX) S1 Dataset. Compiled raw data used for analysis the current study. (XLSX) S1 Materials. Materials used in the experiment. (ZIP) Author Contributions Conceptualization: Jiayu Tao, Zhao Qin, Li Zhang, Lu Liu, Guoli Yan, Valerie Benson. Formal analysis: Jiayu Tao. Funding acquisition: Guoli Yan. Methodology: Jiayu Tao, Zhao Qin, Zhu Meng, Li Zhang, Guoli Yan, Valerie Benson. Supervision: Guoli Yan, Valerie Benson. Writing – original draft: Jiayu Tao, Zhao Qin, Lu Liu, Guoli Yan, Valerie Benson. Writing – review & editing: Jiayu Tao, Zhu Meng, Li Zhang, Guoli Yan, Valerie Benson. References 1. Bavelier D, Neville HJ. Cross-modal plasticity:where and how? Nature Reviews Neuroscience 2002; 3 (6):443–52. Epub 2002/06/04. https://doi.org/10.1038/nrn848 PMID: 12042879. 2. Dye MW, Bavelier D. Attentional enhancements and deficits in deaf populations: An integrative review. Restorative Neurology and Neuroscience. 2010; 28(2):181–92. Epub 2010/04/21. https://doi.org/10. 3233/RNN-2010-0501 PMID: 20404407. 3. Bavelier D, Dye MW, Hauser PC. Do deaf individuals see better? Trends in Cognitive Sciences. 2006; 10(11):512–8. Epub 2006/10/04. https://doi.org/10.1016/j.tics.2006.09.006 PMID: 17015029. 4. Codina CJ, Pascalis O, Baseler HA, Levine AT, Buckley D. Peripheral visual reaction time is faster in deaf adults and British sign language interpreters than in hearing adults. Frontiers in Psychology. 2017; 8:50. Epub 2017/02/22. https://doi.org/10.3389/fpsyg.2017.00050 PMID: 28220085. 5. Hauthal N, Sandmann P, Debener S, Thorne JD. Visual movement perception in deaf and hearing indi- viduals. Advances in Cognitive Psychology. 2013; 9(2):53–61. Epub 2013/07/05. https://doi.org/10. 2478/v10053-008-0131-z PMID: 23826037. 6. Dye MW, Hauser PC, Bavelier D. Is visual selective attention in deaf individuals enhanced or deficient? The case of the useful field of view. PLoS One. 2009; 4(5):e5640. Epub 2009/05/23. https://doi.org/10. 1371/journal.pone.0005640 PMID: 19462009. 7. Bavelier D, Tomann A, Hutton C, Mitchell T, Corina D, Liu G, et al. Visual attention to the periphery is enhanced in congenitally deaf individuals. Journal of Neuroscience. 2000; 20(17):RC93. https://doi.org/ 10.1523/jneurosci.20-17-j0001.2000 PMID: 10952732. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 14 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students 8. Bavelier D, Brozinsky C, Tomann A, Mitchell T, Neville H, Liu G. Impact of early deafness and early exposure to sign language on the cerebral organization for motion processing. Journal of Neuroscience. 2001; 21(22):8931–42. Epub 2001/11/08. https://doi.org/10.1523/jneurosci.21-22-08931.2001 PMID: 9. Proksch J, Bavelier D. Changes in the spatial distribution of visual attention after early deafness. Journal of Cognitive Neuroscience. 2002; 14(5):687–701. Epub 2002/08/09. https://doi.org/10.1162/ 08989290260138591 PMID: 12167254. 10. Dye MW. Foveal processing under concurrent peripheral load in profoundly deaf adults. Journal of Deaf Studies and Deaf Education. 2016; 21(2):122–8. Epub 2015/12/15. https://doi.org/10.1093/deafed/ env054 PMID: 26657078. 11. Kelly LP, Barac-Cikoja D. The comprehension of skilled deaf readers: The roles of word recognition and other potentially critical aspects of competence. In: Cain K, Oakhill J, editors. Children’s comprehension problems in oral and written language: A cognitive perspective. New York: Guilford Press; 2007. p. 244–79. 12. Rayner K, Murphy LA, Henderson JM, Pollatsek A. Selective attentional dyslexia. Cognitive Neuropsy- chology. 1989; 6(4):357–78. https://doi.org/10.1080/02643298908253288 13. Casco C, Tressoldi PE, Dellantonio A. Visual selective attention and reading efficiency are related in children. Cortex. 1998; 34(4):531–46. Epub 1998/11/04. https://doi.org/10.1016/S0010-9452(08) 70512-4 PMID: 9800088. 14. Franceschini S, Gori S, Ruffino M, Pedrolli K, Facoetti A. A causal link between visual spatial attention and reading acquisition. Current Biology. 2012; 22(9):814–9. Epub 2012/04/10. https://doi.org/10.1016/ j.cub.2012.03.013 PMID: 22483940. 15. Rayner K. Eye movements and attention in reading, scene perception, and visual search. Quarterly Journal of Experimental Psychology. 2009; 62(8):1457–506. Epub 2009/05/19. https://doi.org/10.1080/ 17470210902816461 PMID: 19449261. 16. Liu L, Yan G. Effect of parafoveal visual attention enhancement in deaf reading:Evidence from disap- pearing text. Acta Psychologica Sinica. 2018; 50(7):715–26. https://doi.org/10.3724/SP.J.1041.2018. 17. Schotter ER, Angele B, Rayner K. Parafoveal processing in reading. Attention, Perception, & Psycho- physics. 2012; 74(1):5–35. Epub 2011/11/02. https://doi.org/10.3758/s13414-011-0219-2 PMID: 18. Dye MW, Hauser PC, Bavelier D. Visual skills and cross-modal plasticity in deaf readers:Possible impli- cations for acquiring meaning from print. Annals of the New York Academy of Sciences. 2008; 1145:71–82. Epub 2008/12/17. https://doi.org/10.1196/annals.1416.013 PMID: 19076390. 19. Be ´ langer NN, Slattery TJ, Mayberry RI, Rayner K. Skilled deaf readers have an enhanced perceptual span in reading. Psychological Science. 2012; 23(7):816–23. Epub 2012/06/12. https://doi.org/10.1177/ 0956797611435130 PMID: 22683830. 20. Yan M, Pan J, Be ´ langer NN, Shu H. Chinese deaf readers have early access to parafoveal semantics. Journal of Experimental Psychology:Learning, Memory, and Cognition. 2015; 41(1):254–61. Epub 2014/07/08. https://doi.org/10.1037/xlm0000035 PMID: 24999711. 21. Eriksen BA, Eriksen CW. Effects of noise letters upon the identification of a target letter in a nonsearch task. Perception & Psychophysics. 1974; 16(1):143–9. https://doi.org/10.3758/BF03203267 22. Mayberry RI, del Giudice AA, Lieberman AM. Reading achievement in relation to phonological coding and awareness in deaf readers: A meta-analysis. Journal of Deaf Studies and Deaf Education. 2011; 16 (2):164–88. Epub 2010/11/13. https://doi.org/10.1093/deafed/enq049 PMID: 21071623. 23. Westfall J. PANGEA: Power ANalysis for GEneral Anova designs. 2015. Available from: http:// jakewestfall.org/publications/pangea.pdf. 24. Zang C, Fu Y, Bai X, Yan G, Liversedge SP. Investigating word length effects in Chinese reading. Jour- nal of Experimental Psychology: Human Perception and Performance. 2018; 44(12):1831–41. https:// doi.org/10.1037/xhp0000589 PMID: 30475051. 25. Zhang F, Li D. Deaf students’ syntactic awareness of the “shi. . .de” construction in sentence reading. Acta Psychologica Sinica. 2017; 49(11):1383–91. https://doi.org/10.3724/SP.J.1041.2017.01383 26. Zhang H, Lv H, Wu L. The development of Chinese curriculum for deaf college students. Beijing: Huaxia Press; 2009. 27. Li D, Hu K, Chen G, Jin Y, Li M. Combined Raven’s Test (CRT) Shanghai area test report. Journal of Psychological Science. 1988; 11(4):27–31. https://doi.org/10.16719/j.cnki.1671-6981.1988.04.007 28. Zhang H, Wang X. Standardization research on Raven’s standard progressive matrices in China. Acta Psychologica Sinica. 1989; 21(2):113–21. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 15 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students 29. Dare N, Shillcock R. Serial and parallel processing in reading: Investigating the effects of parafoveal orthographic information on nonisolated word recognition. Quarterly Journal of Experimental Psychol- ogy. 2013; 66(3):487–504. Epub 2012/09/07. https://doi.org/10.1080/17470218.2012.703212 PMID: 30. Snell J, Grainger J. Parallel word processing in the flanker paradigm has a rightward bias. Attention Per- ception & Psychophysics. 2018; 80(6):1512–9. Epub 2018/06/03. https://doi.org/10.3758/s13414-018- 1547-2 PMID: 29858975. 31. Cai Q, Brysbaert M. SUBTLEX-CH: Chinese word and character frequencies based on film subtitles. PLoS One. 2010; 5(6):e10729. Epub 2010/06/10. https://doi.org/10.1371/journal.pone.0010729 PMID: 32. Liu L. The effect of parafoveal visual attention enhancement in deaf reading [dissertation]. Tianjin Nor- mal University; 2017. 33. Bates D, Maechler M, Bolker B, Walker S. lme4: Linear Mixed-Effects Models using ’Eigen’ and S4. 2017. Available from: https://CRAN.R-project.org/package = lme4 34. R Development Core Team. R: A language and environment for statistical computing, R Foundation for Statistical Computing, Vienna, Austria. 2017. Available from: http://www.R-project.org/. 35. Be ´ langer NN, Rayner K. Frequency and predictability effects in eye fixations for skilled and less-skilled deaf readers. Visual Cognition. 2013; 21(4):477–97. Epub 2013/08/27. https://doi.org/10.1080/ 13506285.2013.804016 PMID: 23976872. 36. Baayen RH, Davidson DJ, Bates DM. Mixed-effects modeling with crossed random effects for subjects and items. Journal of Memory and Language. 2008; 59(4):390–412. https://doi.org/10.1016/j.jml.2007. 12.005 37. Be ´ langer NN, Lee M, Schotter ER. Young skilled deaf readers have an enhanced perceptual span in reading. Quarterly Journal of Experimental Psychology. 2018; 71(1):291–301. Epub 2017/04/28. https://doi.org/10.1080/17470218.2017.1324498 PMID: 28447500. 38. Inhoff AW, Starr M, Shindler KL. Is the processing of words during eye fixations in reading strictly serial? Perception & Psychophysics. 2000; 62(7):1474–84. https://doi.org/10.3758/BF03212147 PMID: 39. Rayner K, Pollatsek A, Ashby J, Clifton C Jr. Psychology of reading. New York: Psychology Press; PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 16 / 16 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png PLoS ONE Public Library of Science (PLoS) Journal

Reading skill modulates the effect of parafoveal distractors on foveal lexical decision in deaf students

PLoS ONE, Volume 14 (9) – Sep 12, 2019

Loading next page...
 
/lp/public-library-of-science-plos-journal/reading-skill-modulates-the-effect-of-parafoveal-distractors-on-foveal-I7I9gDlovQ
Publisher
Public Library of Science (PLoS) Journal
Copyright
Copyright: © 2019 Tao et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: All relevant data are within the manuscript and its Supporting Information files. Funding: This research was supported by National Social Science Fund of China (16BYY074) grant awarded to Guoli Yan. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist.
eISSN
1932-6203
D.O.I.
10.1371/journal.pone.0221891
Publisher site
See Article on Publisher Site

Abstract

a1111111111 In low-level perceptual tasks and reading tasks, deaf individuals show a redistribution of spatial visual attention toward the parafoveal and peripheral visual fields. In the present study, the experiment adopted the modified flanker paradigm and utilized a lexical decision task to investi- gate how these unique visual skills may influence foveal lexical access in deaf individuals. It OPENACCESS was predicted that irrelevant linguistic stimuli presented in parafoveal vision, during a lexical Citation: Tao J, Qin Z, Meng Z, Zhang L, Liu L, Yan decision task, would produce a larger interference effect for deaf college student readers if the G, et al. (2019) Reading skill modulates the effect of parafoveal distractors on foveal lexical decision stimuli acted as distractors during the task. The results showed there was a larger interference in deaf students. PLoS ONE 14(9): e0221891. effect in deaf college student readers compared to the interference effect observed in partici- https://doi.org/10.1371/journal.pone.0221891 pants with typical levels of hearing. Furthermore, deaf college student readers with low-skilled Editor: Susan Nittrouer, University of Florida, reading levels showed a larger interference effect than those with high-skilled reading levels. UNITED STATES The current study demonstrates that the redistribution of spatial visual attention toward the par- Received: March 25, 2019 afoveal visual regions in deaf students impacts foveal lexical processing, and this effect is mod- Accepted: August 16, 2019 ulated by reading skill. The findings are discussed in relation to the potential effect that enhanced parafoveal attention may have on everyday reading for deaf individuals. Published: September 12, 2019 Copyright:© 2019 Tao et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in Introduction any medium, provided the original author and Many studies employing low-level visual cognitive tasks suggest that, for deaf individuals, source are credited. visual attention is reorganized to compensate for the lack or absence of auditory input pro- Data Availability Statement: All relevant data are vided by a complex environment [1], and, this reorganization has been thought to result in a within the manuscript and its Supporting redistribution of spatial visual attention toward the parafoveal and peripheral visual fields in Information files. deaf individuals [2, 3]. Funding: This research was supported by National In support of this hypothesis, deaf individuals have been shown to respond more rapidly to Social Science Fund of China (16BYY074) grant flashed or moving stimuli in the periphery [4, 5], and they have been reported to have awarded to Guoli Yan. The funder had no role in study design, data collection and analysis, decision enhanced visual selective attention in the peripheral visual field [6]. There is also evidence to publish, or preparation of the manuscript. from fMRI studies that the middle temporal-medial superior temporal area of deaf individuals shows a greater response to peripheral motion stimuli compared to individuals with typical Competing interests: The authors have declared that no competing interests exist. levels of hearing [7, 8]. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 1 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students However, in some circumstances increased parafoveal or peripheral visual attention might hamper foveal visual field performance. For example, Proksch and Bavelier have shown that deaf individuals exhibit greater distractibility from parafoveal distractors in target detection tasks [9], and reduced discrimination performance was reported when a central discrimination task and a peripheral localization task had to be carried out simultaneously [10]. These studies, using low- level cognitive tasks, have shown that deaf individuals are more distracted by the presentation of task-irrelevant information in peripheral vision when the task requirements demand a central focus. An important question leading from these studies is, to what extent increased parafoveal processing in deaf individuals might impact an everyday task, such as reading. It is well documented that the illiteracy rate of deaf individuals has been very high for decades. The average reading proficiency of young deaf adults graduating from high school has been reported to be far behind that of individuals with typical levels of hearing [11]. It is also well known that there is a strong link between visual attention and reading effi- ciency in individuals with typical levels of hearing [12]. For example, it has been shown that selective spatial attention is significantly associated with reading efficiency [13], and that visual-spatial attention can effectively predict reading achievement [14] in hearing individuals. Foveal information (the central 2˚ of the visual field) is essential for word processing and for reading [15]. However, eye movement studies in reading have also indicated that linguistic information can be obtained from the parafoveal region (from the foveal region up to 5˚ of visual angle on either side of fixation). If readers’ allocate attentional resources to process par- afoveal words, then reading efficiency can be increased, but, paradoxically, if too much atten- tion is allocated to process parafoveal information, then this may actually hinder identification of the foveal word under inspection [16]. Therefore, readers have to carefully allocate their attention between the foveal region and parafoveal region, in order to obtain parafoveal infor- mation that is useful (facilitatory), but not harmful (inhibitory) for reading [17]. Given the relationship between foveal and parafoveal processing in reading, it has been speculated that the redistribution of spatial visual attention toward the parafoveal visual fields, observed for deaf individuals, may actually slow down foveal processing, and hence cause read- ing difficulties for deaf individuals [18]. In support of that view, some studies have shown that deaf readers, including Chinese deaf readers, can access parafoveal information more effi- ciently during sentence reading [19, 20]. However, although these studies show deaf individu- als to have enhanced access to parafoveal information more efficiently during sentence reading, they cannot tell us how visual processing of parafoveal information may influence foveal lexical access during reading in deaf readers. In this paper, we investigate whether the redistribution of spatial visual attention toward the parafoveal visual fields in deaf individuals impacts foveal lexical access for a higher-level cognitive task (lexical decision), with an ultimate aim of relating how the findings might affect the everyday task of reading. The current study adopted the modified flanker paradigm [21] and utilized a lexical deci- sion task to investigate the effects of task-irrelevant linguistic stimuli, presented in parafoveal locations, on central lexical processing in deaf individuals. If the redistribution of spatial visual attention toward the parafoveal visual fields in deaf readers affects the lexical processing of stimuli under foveal inspection, it would be predicted that deaf participants should take longer to make a lexical decision when a target character is presented centrally along with a distractor character simultaneously presented in parafoveal vision. We, therefore, aimed to test whether interference effects were greater for deaf college student readers compared to participants with typical levels of hearing. A further aim of this experiment was to explore whether any observed interference effects for the deaf readers were modulated by reading level ability. Although there is no empirical PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 2 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students evidence to support that claim, it could be the case that high-skilled deaf readers are able to inhibit irrelevant distractors more effectively than low-skilled deaf readers. If this is the case we would predict a greater interference effect for low-skilled deaf readers compared to high- skilled deaf readers. One of the challenges in the field is to uncover factors that distinguish skilled deaf readers from unskilled deaf readers [22]. Control over the redistribution of spatial visual attention may be a factor that distinguishes between these two groups of deaf readers. Method Ethics statement The procedures of the study have been approved by the Ethical Committee of Tianjin Normal University. All deaf college student participants and all college students with typical levels of hearing provided informed written consent prior to inclusion in the study. Parents provided informed written consent for middle school students with typical levels of hearing. Participants Forty severely to profoundly deaf college students (M = 20.69 years, SD = 1.25) were age recruited from the Technical College for the Deaf, Tianjin University of Technology, China. The deaf participants’ hearing losses were� 75dB (M = 100.45 dB, SD = 12.61) in the better ear. Participants were either born deaf or became deaf before three years of age. None of the participants had received a cochlear implant. Thirty-five deaf participants reported sign lan- guage as their primary and preferred language of communication. Five deaf participants reported spoken language as their primary and preferred language of communication. Demo- graphic information of participants was collected by a questionnaire survey. The information was provided by the participants themselves. The power analysis of the current study for an average effect size of d = 0.45 is 0.97, and that power value is greater than the minimum recom- mended level of 0.80 [23, 24]. A reading control group consisted of forty Chinese middle school students with typical lev- els of hearing (M = 13.98 years, SD = 0.79). We matched their reading comprehension level age on the Chinese Proficiency Test for Deaf College Students [25, 26]. This test was developed for deaf college students and consists of three parts: sentence comprehension, short text compre- hension, and ancient text comprehension. The test has a total of 50 multiple choice questions, with 2 points given for each question. The completion time is 30 minutes. There was no signif- icant difference in reading comprehension scores between the deaf college students (M = 67.03, SD = 5.63) and the reading control group (M = 66.48, SD = 7.92), t (78) = 0.36, p = .72. There was no significant difference in reading fluency [20] scores between the deaf college students (M = 342.34 characters/min, SD = 111.98) and the reading control group (M = 330.78 characters/min, SD = 101.20), t (78) = 0.49, p = .63. An age control group consisted of forty Chinese college students with typical levels of hear- ing (M = 20.49 years, SD = 1.24). There was no significant difference in chronological age age between the deaf college students and the age control group, t (78) = – 0.71, p = .48. There was a significant difference in reading comprehension scores between the deaf college students (M = 67.03, SD = 5.63) and the age control group (M = 86.15, SD = 5.33), t (78) = 15.61, p < .001, d = 3.49 [25, 26]. There was a significant difference in reading fluency [20] scores between the deaf college students (M = 342.34 characters/min, SD = 111.98) and the age control group (M = 458.23 characters/min, SD = 89.58), t (78) = 5.11, p < .001, d = 1.14. In addition, we matched participants from the three groups on nonverbal IQ using the Raven Test [27, 28]. There was no significant difference in standard score among the three groups, F (2,117) = 0.20, p = .82. All participants had normal or corrected-to-normal vision. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 3 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Materials There were 360 trials (see S1 Materials), of which 180 trials presented pseudo-Chinese charac- ters and were used as filler materials [29, 30]; Eighty percent of trials were target-distractor pairs, and the other 20 percent of trials were without distractors. For the target-distractor pairs, twenty undergraduate students with typical levels of hearing, who did not participate in the experiment, evaluated the relatedness of target characters and distractor characters on a 5 point scale (1 = very dissimilar, 5 = very similar). The targets and distractors shared no orthographic (M = 1.39, SD = 0.38), phonological (M = 1.19, SD = 0.26), or semantic (M = 1.37, SD = 0.38) similarities. The target characters mean frequency was 145.99 per million (SD = 430.01) and the distractor characters mean frequency was 120.72 per million (SD = 548.35), and the target characters and distractor characters mean number of strokes was 8.72 (SD = 3.23) and 8.74 (SD = 3.00), respectively [31]. Given that the parafoveal region is defined as up to 5˚ visual angle on either side of fixation, the parafoveal presentation locations for the study were operationalized as 2˚, 3˚, or 4˚ visual angle from the center of the display on either side. Half of the distractors were presented to the left parafoveal visual field and half to the right. The negative sign and positive sign represent the left and the right of the central fixation point, respectively. Design The study used a 3 (Group: deaf college student group, reading control group, age control group) ×7 (Interference condition: without interference, –4˚, –3˚, –2˚, + 2˚, +3˚, +4˚) mixed experimental design. The group was a between-subjects factor and interference condition was a within-subjects factor. Procedure Participants were seated 45cm from the 14-inch monitor of a laptop (with a 60-Hz refresh rate, 1024×768 resolution). The pixel value of each character was 34.67 × 34.67. At this distance, one character subtended about 1˚ visual angle [32]. A software package (Eprime2.0) was used to run the program and record reaction times (RTs). Each trial began with a cross at the center of the screen for 500ms. The target character was then presented at the center of the screen (with/without a distractor character being presented on either side of fixation at different locations, depending on the interference condition). Par- ticipants were asked to focus on the central fixation and were asked to make a lexical decision for the foveal target as quickly and accurately as possible when this was presented, regardless of whether a parafoveal distractor was simultaneously presented with the target. Participants used the right or left index finger to press the F or J on the keyboard, indicating character (word) or pseudo-character (non-word), respectively. Participants had a maximum of 3s to respond [32]. Each target presentation trial was followed by a random interval 1000~1300ms before the next trial sequence began; Fig 1 shows a schematic of the trial sequence. There was a break after every 72 trials and the experiment lasted approximately 30 minutes. Data analysis The analysis of RTs excluded pseudo-Chinese character filler trials, incorrectly answered trials (4.35%), and trials with an RT which were beyond 2.5 standard deviations (2.82%) from each participant’s mean RT (see S1 Dataset). Analysis of error rates excluded pseudo-Chinese char- acter filler trials and trials which were beyond 2.5 standard deviations (2.74%) from each par- ticipant’s mean RT [29, 30]. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 4 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Fig 1. Example of distractor trial sequence. Fig 1 represents a trial of +3˚ interference condition. The target character (举, lift) was presented at the center of the screen and the distractor character (址, address) was presented at 3˚ visual angle of right side from the center of the display. https://doi.org/10.1371/journal.pone.0221891.g001 Data were analyzed using the lme4 package [33], within the R environment for statistical computing [34]. Generalized linear mixed models were used to analyze error rate data (binary variables) [35]. We report b values, standard errors (SE) and t values (z values for error rates data), with |t|� 1.96 (|z|� 1.96) deemed significant at the 0.05 alpha level. In the model, partici- pants and items were specified as crossed random effects [36]. Group and interference condition were specified as fixed effects. To assess differences between pairs of groups, we set up two suc- cessive difference contrasts: comparing the age control group to the deaf college student group, and the deaf college student group to the reading control group. Analysis of interference effects was based on planned comparisons for each of the different interference conditions and the without-interference (single target) condition, which was used as a reference condition. In this model, the interactions represent the interference effect for a given pair of groups. If the maxi- mal random effects structure model failed to converge, the random structure was trimmed, and then we used a likelihood ratio test to verify the best fit of the converged models to the data [37]. Results Error rates When different interference conditions were compared to the without-interference condition separately, the random effects structure model failed to converge. Therefore, we compared dif- ferent interference conditions as a whole to the without-interference condition and the model including participant and item random intercepts converged. The deaf college student group average error rate was 3.63%; the age control group average error rate was 3.13%; the reading control group average error rate was 5.86%; See Table A in S1 Appendix. As shown in Table 1, the deaf college student group did not differ in their error rate with the age control group (b = 0.20, SE = 0.19, z = 1.05); the deaf college student group and the reading control group was significantly different in error rate (b = 0.53, SE = 0.18, z = 2.96), and the interference effect main effect was not significant (b = 0.09, SE = 0.17, z = 0.51). The deaf college student group and the age control group did not significantly interact with the interference effect (b = – 0.11, SE = 0.23, z = – 0.48), and the deaf college student group and the reading control group did not significantly interact with the interference effect (b = 0.14, PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 5 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students SE = 0.20, z = 0.69). Thus the only significant result for the error rate analysis, in relation to the deaf college student group, showed that the reading control group made more errors com- pared to the deaf college students, and this finding can most likely be explained by age differ- ences between those two groups. Reaction times A model that included participant random slope alongside participant and item random inter- cepts was deemed the best model for analysis, based on the likelihood-ratio test result [36]. As shown in Table 2, the group main effects did not differ: The deaf college student group did not differ in their RTs (see Table B in S1 Appendix) from the age control group (b = 26.03, SE = 28.78, t = 0.90); and the deaf college student group did not differ in their RTs from the reading control group (b = 1.48, SE = 28.78, t = 0.05). The interference effect main effects were significant in all interference conditions, all |ts| >1.96, showing that distractors presented at all eccentricities resulted in longer RTs compared to when no distractors were presented (the without distractor condition). As shown in Table 2, the deaf college student group and age control group significantly inter- acted with the interference effect in all interference conditions, all |ts| >1.96, indicating that the interference effect was larger for the deaf college student group than for age control group; see Fig 2. The deaf college student group and reading control group significantly interacted with the interference effect in all interference conditions, all |ts| >1.96, indicating that the interference effect was also larger for the deaf college student group than for the reading control group [20]. In summary, the results showed that the presence of a parafoveal lexical distractor resulted in poorer performance (longer RTs) for a lexical decision task in the deaf participants com- pared to the participants with typical levels of hearing. A further aim of this study was to investigate whether reading skill modulated observed interference effects in deaf individuals. We tested this hypothesis by comparing interference effects between different reading levels of deaf individuals. If deaf individuals with low-skilled reading levels allocate more attentional resources to process parafoveal information than deaf individuals with high-skilled reading levels, we predicted that low-skilled deaf readers in the current study should have taken longer to make a lexical decision compared to high-skilled deaf readers in the current study. Analyses of interference effects for deaf readers with high-skilled reading levels versus deaf readers with low-skilled reading levels We separated the deaf college students into individuals with high-skilled reading levels and individuals with low-skilled reading levels, using a median split technique [38] based on their Table 1. Groups × interference conditions results for error rates (%). b SE z Intercept -3.61 0.11 -34.11 Deaf college student group vs. Age control group 0.20 0.19 1.05 Reading control group vs. Deaf college student group 0.53 0.18 2.96 Interference vs. Without interference 0.09 0.17 0.51 Deaf college student group vs. Age control group × Interference vs. Without interference -0.11 0.23 -0.48 Reading control group vs. Deaf college student group × Interference vs. Without interference 0.14 0.20 0.69 Note. Deaf college student group: forty severely to profoundly deaf college students. Age control group: forty Chinese college students with typical levels of hearing. Reading control group: forty Chinese middle school students with typical levels of hearing. |z|� 1.96 deemed significant at the 0.05 alpha level. Significant z-values are shown in bold. https://doi.org/10.1371/journal.pone.0221891.t001 PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 6 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Table 2. Groups × interference conditions results for RTs (ms). b SE t Intercept 654.93 12.07 54.27 Deaf college student group vs. Age control group 26.03 28.78 0.90 Reading control group vs. Deaf college student group 1.48 28.78 0.05 –2˚vs.Without interference -92.76 10.45 -8.88 –3˚vs.Without interference -79.64 11.01 -7.24 –4˚vs.Without interference -79.87 10.71 -7.46 +2˚vs.Without interference -79.28 10.13 -7.83 +3˚vs.Without interference -73.46 10.55 -6.96 +4˚vs.Without interference -74.70 10.51 -7.11 Deaf college student group vs. Age control group × –2˚vs.Without interference -59.62 19.38 -3.08 Reading control group vs. Deaf college student group × –2˚vs.Without interference 58.12 19.46 2.99 Deaf college student group vs. Age control group × –3˚vs.Without interference -79.32 21.16 -3.75 Reading control group vs. Deaf college student group × –3˚vs.Without interference 75.05 21.23 3.53 Deaf college student group vs. Age control group × –4˚vs.Without interference -88.90 20.23 -4.39 Reading control group vs. Deaf college student group× –4˚vs.Without interference 86.14 20.30 4.24 Deaf college student group vs. Age control group × +2˚vs.Without interference -63.58 18.34 -3.47 Reading control group vs. Deaf college student group× +2˚vs.Without interference 55.50 18.40 3.02 Deaf college student group vs. Age control group × +3˚vs.Without interference -83.47 19.73 -4.23 Reading control group vs. Deaf college student group × +3˚vs.Without interference 88.39 19.80 4.46 Deaf college student group vs. Age control group × +4˚vs.Without interference -80.94 19.59 -4.13 Reading control group vs. Deaf college student group × +4˚vs.Without interference 78.34 19.64 3.99 Note. Deaf college student group: forty severely to profoundly deaf college students. Age control group: forty Chinese college students with typical levels of hearing. Reading control group: forty Chinese middle school students with typical levels of hearing. Interference conditions: The negative sign and positive sign represent the left and the right of the central fixation point, respectively. |t|� 1.96 deemed significant at the 0.05 alpha level. Significant t-values are shown in bold. https://doi.org/10.1371/journal.pone.0221891.t002 reading comprehension [25, 26] test scores (Given that this test was developed for deaf college students, it can be argued to reflect and distinguish reading ability of deaf college students). The two groups comprised a low/medium skilled group (n = 19) with reading comprehension test scores < 68.00, and a high-skilled group (n = 21) with all scores� 68.00. There was a sig- nificant difference in reading comprehension scores between the low-skilled deaf group (M = 62.05, SD = 3.37) and the high-skilled deaf group (M = 71.52, SD = 2.60), t (38) = 10.00, p < .001, d = 3.17. In the low-skilled deaf group, seventeen deaf participants reported sign language as their primary and preferred language of communication, and two deaf participants reported spoken language as their primary and preferred language of communication. In the high- skilled deaf group, eighteen deaf participants reported sign language as their primary and pre- ferred language of communication, and three deaf participants reported spoken language as their primary and preferred language of communication. In relation to the hearing loss of the deaf readers, there was no significant difference in the better ear between the low-skilled deaf group (M = 100.32 dB, SD = 12.13) and the high-skilled deaf group (M = 100.57 dB, SD = 13.32), t (38) = 0.06, p = .95. There was no significant differ- ence in chronological age between the low-skilled deaf group (M = 20.71, SD = 1.42) and the high-skilled deaf group (M = 20.68, SD = 1.12), t (38) = – 0.09, p = .93. Finally, there was no significant difference in IQ standard score between the low-skilled deaf group (M = 101.32, SD = 14.99) and the high-skilled deaf group (M = 109.83, SD = 13.51), t (38) = 1.89, p = .07. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 7 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Fig 2. Comparison of the interference effect for three participant groups in different interference conditions. Deaf college student group: forty severely to profoundly deaf college students. Age control group: forty Chinese college students with typical levels of hearing. Reading control group: forty Chinese middle school students with typical levels of hearing. Interference conditions: The negative sign and positive sign represent the left and the right of the central fixation point, respectively. Interference effect (ms) = RTs with interference—RTs without interference (single target). Error bars represent the standard errors of the means. https://doi.org/10.1371/journal.pone.0221891.g002 Error rates When different interference conditions were compared to the without-interference condition separately, the random effects structure model failed to converge. Therefore, we compared dif- ferent interference conditions as a whole to the without-interference condition. The low-skilled deaf group average error rate was 3.74%; the high-skilled deaf group aver- age error rate was 3.53%; See Table C in S1 Appendix. As shown in Table 3, the low-skilled deaf group did not differ in their error rate from the high-skilled deaf group (b = – 0.16, SE = 0.28, z = – 0.56). The interference effect main effect was not significant (b = – 0.03, SE = 0.27, z = – 0.13), and the low-skilled deaf group and the high-skilled deaf group did not significantly interact with the interference effect (b = 0.51, SE = 0.33, z = 1.54). Reaction times A model that included participant random slope alongside participant and item random inter- cepts was deemed the best model for analysis, based on the likelihood-ratio test result [36]. Table 3. Groups × interference conditions results for error rates (%). b SE z Intercept -3.94 0.19 -21.00 Low-skilled deaf group vs. High-skilled deaf group -0.16 0.28 -0.56 Interference vs. Without interference -0.03 0.27 -0.13 Low-skilled deaf group vs. High-skilled deaf group × Interference vs. Without interference 0.51 0.33 1.54 Note. Low-skilled deaf group: nineteen severely to profoundly deaf college students with reading comprehension test scores < 68.00. High-skilled deaf group: twenty-one severely to profoundly deaf college students with scores� 68.00.|z|� 1.96 deemed significant at the 0.05 alpha level. Significant z-value is shown in bold. https://doi.org/10.1371/journal.pone.0221891.t003 PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 8 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Table 4. Groups × interference conditions results for RTs (ms). b SE t Intercept 666.07 25.48 26.15 Low-skilled deaf group vs. High-skilled deaf group 102.47 50.45 2.03 –2˚vs.Without interference –135.48 21.17 –6.40 –3˚vs.Without interference –133.41 23.57 –5.66 –4˚vs.Without interference –140.12 22.24 –6.30 +2˚vs.Without interference –121.20 20.27 –5.98 +3˚vs.Without interference –132.59 20.74 –6.39 +4˚vs.Without interference –130.25 20.60 –6.32 Low-skilled deaf group vs. High-skilled deaf group × –2˚vs.Without interference –82.65 38.47 –2.15 Low-skilled deaf group vs. High-skilled deaf group × –3˚vs.Without interference –76.38 43.70 –1.75 Low-skilled deaf group vs. High-skilled deaf group × –4˚vs.Without interference –61.14 40.81 –1.50 Low-skilled deaf group vs. High-skilled deaf group × +2˚vs.Without interference –85.66 36.49 –2.35 Low-skilled deaf group vs. High-skilled deaf group × +3˚vs.Without interference –63.67 37.51 –1.70 Low-skilled deaf group vs. High-skilled deaf group × +4˚vs.Without interference –70.38 37.20 –1.89 Note. Low-skilled deaf group: nineteen severely to profoundly deaf college students with reading comprehension test scores < 68.00.High-skilled deaf group: twenty-one severely to profoundly deaf college students with scores� 68.00. Interference conditions: The negative sign and positive sign represent the left and the right of the central fixation point, respectively.|t|� 1.96 deemed significant at the 0.05 alpha level. Significant t-values are shown in bold. https://doi.org/10.1371/journal.pone.0221891.t004 As shown in Table 4, the low-skilled deaf group had longer RTs (see Table D in S1 Appen- dix) than the high-skilled deaf group (b = 102.47, SE = 50.54, t = 2.03), and the interference effect main effects were significant in all interference conditions, all |ts| >1.96. As shown in Table 4, the two groups of deaf readers significantly interacted with the inter- ference effects for 2˚ conditions on either side of fixation; see Fig 3, indicating that the Fig 3. Comparison of the interference effect for two deaf reader groups in different interference conditions. Low- skilled deaf group: nineteen severely to profoundly deaf college students with reading comprehension test scores < 68.00. High-skilled deaf group: twenty-one severely to profoundly deaf college students with scores� 68.00. Interference conditions: The negative sign and positive sign represent the left and the right of the central fixation point, respectively. Interference effect (ms) = RTs with interference—RTs without interference (single target). Error bars represent the standard errors of the means. https://doi.org/10.1371/journal.pone.0221891.g003 PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 9 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students interference effect was larger for the low-skilled deaf group (for –2˚ vs. without interference, 171.63 ms; for +2˚ vs. without interference, 163.98 ms) than for the high-skilled deaf group (for –2˚vs. without interference, 95.05 ms; for +2˚vs. without interference, 78.07 ms). The two groups of deaf readers marginally significantly interacted with the interference effects for 3˚ interference conditions on either side of fixation and for the +4˚ interference condition, indi- cating that the interference effect was larger for the low-skilled deaf group (for –3˚ vs. without interference, 165.60 ms; for +3˚ vs. without interference, 161.15 ms; for +4˚ vs. without inter- ference,163.10 ms) than for the high-skilled deaf group (for –3˚ vs. without interference, 95.59 ms; for +3˚ vs. without interference, 98.07 ms; for +4˚ vs. without interference, 98.36 ms) in the interference conditions. For the –4˚ interference condition, the low-skilled deaf group (for –4˚vs. without interference,168.68 ms) show an obvious data trend of a larger interference effect than for the high-skilled deaf group (for –4˚ vs. without interference, 109.50 ms) but this numerical difference was not statistically significant [20]. In summary, separating the deaf students into high and low-skilled deaf readers resulted in differences in the magnitude of distractor effects on the time taken to make a lexical decision, with low-skilled readers taking longer than high-skilled readers. Since there were equivalent numbers of deaf students who rely on sign language as their preferred way to communicate, it is unlikely that this preference impacted on the modulating effects of reading skill on distractor interference. One interesting question that arises from this finding relates to whether the increased inter- ference effect for low-skilled readers is unique to deaf readers, or whether the same effect is observed for hearing readers? To address this question, we carried out the following analyses for the reading matched hearing participants. Analyses of interference effects for reading-matched hearing participants with high-skilled reading levels versus low-skilled reading levels We separated the reading-matched hearing middle school participants into individuals with high-skilled reading levels and individuals with low-skilled reading levels, using a median split technique [38] based on their reading comprehension [25, 26] test scores. The two groups comprised a low/medium skilled group (n = 21) with reading comprehension test scores� 64.00, and a high-skilled group (n = 19) with scores > 64.00. There was a significant difference in reading comprehension test scores between the low-skilled group (M = 60.38, SD = 3.44) and the high-skilled group (M = 73.21, SD = 5.61), t (38) = 8.81, p< .001, d = 2.79. There was no significant difference of chronological age between the low-skilled group (M = 13.83, SD = 0.79) and the high-skilled group (M = 14.15, SD = 0.77), t (38) = 1.32, p = .20; and no significant difference in IQ standard score between the low-skilled group (M = 101.95, SD = 9.75) and the high-skilled group (M = 108.00, SD = 11.55), t (38) = 1.80, p = .08. Error rates We compared different interference conditions as a whole to the without-interference condi- tion. The average error rate of the low-skilled group was 6.67%; the average error rate of the high-skilled group was 4.94%; See Table E in S1 Appendix. As shown in Table 5, the low-skilled group did not differ in their error rate from the high- skilled group (b = 0.39, SE = 0.22, z = 1.81). There was no significant main effect of interference (b = 0.14, SE = 0.19, z = 0.71), and the low-skilled group and the high-skilled group did not sig- nificantly interact with the interference effect (b = 0.00, SE = 0.26, z = -0.02). PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 10 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Table 5. Groups × interference conditions results for error rates (%). b SE z Intercept -3.17 0.13 -23.65 Low-skilled group vs. High-skilled group 0.39 0.22 1.81 Interference vs. Without interference 0.14 0.19 0.71 Low-skilled group vs. High-skilled group × Interference vs. Without interference 0.00 0.26 -0.02 Note. Low-skilled group: twenty-one hearing middle school participants with reading comprehension test scores� 64.00. High-skilled group: nineteen hearing middle school participants with scores > 64.00.|z|� 1.96 deemed significant at the 0.05 alpha level. Significant z-value is shown in bold. https://doi.org/10.1371/journal.pone.0221891.t005 Reaction times The maximal random effects structure, including participant and item random intercepts, converged. As shown in Table 6, the low-skilled group did not differ in their RTs (see Table F in S1 Appendix) from the high-skilled group (b = 33.30, SE = 28.20, t = 1.18), and the interfer- ence effect main effects were significant in all interference conditions, all |ts| >1.96. As shown in Table 6, the low-skilled group and the high-skilled group did not significantly interact with the interference effect in five condtions (|ts| < 1.96). The two groups of middle school participants significantly interacted with the interference effects for +4˚ condition; see Fig 4, indicating that the interference effect was larger for the low-skilled group (for +4˚ vs. without interference, 64.20 ms) than for the high-skilled group (for +4˚ vs. without interfer- ence, 34.74 ms). For the +2˚ and +3˚ interference conditions, the low-skilled group (for +2˚ vs. without interference, 76.30 ms; +3˚ vs. without interference, 54.66 ms) show a data trend of a larger interference effect than for the high-skilled group (for +2˚ vs. without interference, 50.08 ms; +3˚ vs. without interference, 32.18 ms), but this numerical difference was not statisti- cally significant [20]. Table 6. Groups × interference conditions results for RTs (ms). b SE t Intercept 663.30 14.34 46.26 Low-skilled group vs. High-skilled group 33.30 28.20 1.18 –2˚vs.Without interference -74.93 9.22 -8.13 –3˚vs.Without interference -57.62 9.21 -6.25 –4˚vs.Without interference -53.22 9.21 -5.78 +2˚vs.Without interference -63.91 9.21 -6.94 +3˚vs.Without interference -42.98 9.20 -4.67 +4˚vs.Without interference -48.87 9.18 -5.33 Low-skilled group vs. High-skilled group × –2˚vs.Without interference -7.65 13.14 -0.58 Low-skilled group vs. High-skilled group × –3˚vs.Without interference -4.50 13.15 -0.34 Low-skilled group vs. High-skilled group × –4˚vs.Without interference 5.29 13.11 0.40 Low-skilled group vs. High-skilled group × +2˚vs.Without interference -21.71 13.14 -1.65 Low-skilled group vs. High-skilled group × +3˚vs.Without interference -22.20 13.09 -1.70 Low-skilled group vs. High-skilled group × +4˚vs.Without interference -28.10 13.02 -2.16 Note. Low-skilled group: twenty-one hearing middle school participants with reading comprehension test scores� 64.00. High-skilled group: nineteen hearing middle school participants with scores > 64.00. Interference conditions: The negative sign and positive sign represent the left and the right of the central fixation point, respectively.|t|� 1.96 deemed significant at the 0.05 alpha level. Significant t-values are shown in bold. https://doi.org/10.1371/journal.pone.0221891.t006 PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 11 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Fig 4. Comparison of the interference effect for two hearing middle school reader groups in different interference conditions. Low-skilled group: twenty-one hearing middle school participants with reading comprehension test scores� 64.00. High-skilled group: nineteen hearing middle school participants with the scores > 64.00. Interference conditions: The negative sign and positive sign represent the left and the right of the central fixation point, respectively. Interference effect (ms) = RTs with interference—RTs without interference (single target). Error bars represent the standard errors of the means. https://doi.org/10.1371/journal.pone.0221891.g004 The pattern for the comparison between interference effects observed in the high-skilled and low-skilled deaf students appears to be different to that observed for the same comparison in the hearing students. In the deaf students interference effects are greater for the low-skilled readers when distractors are presented at different parafoveal eccentricities both to the right and to the left of fixation. However, for the hearing students, greater interference effects for the low-skilled read- ers in that group are obsreved exclusiveley for distractors presented to the right of fixation. Discussion The aim of this study was to investigate whether the well documented redistribution of spatial visual attention toward the parafoveal visual fields in deaf individuals would influence the time taken to process lexical stimuli under foveal inspection. Firstly, we predicted that deaf partici- pants should take longer to make a lexical decision when a target character was presented cen- trally along with a distractor character simultaneously presented in parafoveal vision. Secondly, we proposed that low-skilled deaf readers would show a greater interference effect compared to high-skilled deaf readers. The results support our experimental hypotheses. We found a larger parafoveal interference effect for deaf college students (overall) than for participants with typical levels of hearing. We interpreted this finding to result from deaf individuals allocating more attentional resources to process parafoveal information. Hence, the redistribution of spatial visual attention in the deaf readers affects lexical processing of stimuli under foveal inspection. In reading, the goal is to process the individual words sequentially and to integrate the meaning of these words in order to build up a full representation and comprehensive under- standing of the text [39]. If readers’ allocate attentional resources to process parafoveal words, then reading efficiency can be increased, but, if too much attention is allocated to process par- afoveal information, then this may actually hinder identification of the foveal word under inspection [16]. Therefore, readers have to carefully allocate their attention between the foveal region and parafoveal region, in order to obtain parafoveal information that is useful (facilita- tory), but not harmful (inhibitory) for the reading process [17]. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 12 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Some studies have found that deaf individuals can access parafoveal information more effi- ciently during sentence reading [19, 20], however, little is yet known about how visual process- ing of parafoveal information may influence foveal lexical access during reading for deaf readers. The findings from the current study support the hypothesis of the redistribution of spatial visual attention toward the parafoveal visual fields in deaf individuals. An obvious ques- tion raised by the findings would be to investigate whether the observed slow-down of foveal processing for lexical decision tasks could impact in everyday reading difficulties in deaf indi- viduals, and whether any difficulties would be modulated by reading level in the deaf individuals. However, it should also be noted that in normal reading parafoveal information is often related to foveal information, and hence has the potential to speed foveal processing. In that case it is possible that deaf readers might show a larger priming effect than hearing readers for parafoveal facilitatory information. Since the current study only tested an interfering stimulus (an unrelated character), the question as to whether greater parafoveal attention could enhance reading processes in the deaf readers remains to be tested. Furthermore, the interference effects observed for the deaf students were modulated by reading level. Low-skilled deaf readers showed a greater interference effect compared to high- skilled deaf readers. This finding suggests that there may be differences in the way that low- skilled reading deaf individuals and high-skilled reading deaf individuals allocate attentional resources to parafoveal information. In order to test whether that finding is unique to deaf readers, we also separated the read- ing-matched hearing middle school participants into individuals with high-skilled reading lev- els and individuals with low-skilled reading levels. The low-skilled middle school participants showed a greater interference effect (although not all effects were significant) compared to high-skilled middle school readers in interference conditions where distractors were presented to the right of fixation. This finding is observed exclusively for distractors presented to the right of fixation in low-skilled hearing individuals, whereas the effect is observed for distrac- tors presented to the left and to the right of fixation in the low-skilled deaf readers. We inter- pret this finding to reflect a general effect of reading skill on modulating attention to the parafovea, but also that this modulation will be different for hearing and deaf individuals. A study of lexical decision flanker tasks with hearing readers has revealed a rightward bias that is argued to be related to reading direction [30], and in sentence reading tasks, the size of the per- ceptual span is asymmetric [15, 17] to the right for hearing individuals. Therefore, it is likely that an effect of reading skill would be detected for distractors presented to the right of fixation for hearing readers, but for deaf individuals an effect of reading skill would be detected for dis- tractors presented to the right and the left of fixation. Compared with hearing readers, deaf individuals allocate more attentional resources to process parafoveal information presented to the left and to the right of fixation. This is in line with the findings from the current lexical decision study which clearly demonstrates an effect of reading skill on the ability to ignore par- afoveal distractors, both to the left and to the right of fixation, for deaf readers. A related finding from a case study of a hearing reader who was diagnosed with develop- mental dyslexia showed that the main reason for his reading problem was a selective atten- tional deficit, whereby letters from words available in parafoveal vision interfered with his processing of the currently fixated word [12]. This finding is similar to the results from the cur- rent study for deaf readers. An important question raised by the findings relates to how the ability to control what is attended to and processed in the parafovea might affect reading skill development in deaf individuals. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 13 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students Conclusion In sum, the current study confirms that deaf readers show a larger interference effect from irrelevant parafoveal distractors, during a lexical decision task. This is likely to result from the greater distribution of spatial visual attention toward the parafoveal visual field. Furthermore, low-skilled deaf readers show a larger interference effect which suggests that the allocation of attentional resources toward the parafovea, may reflect less volitional control in low-skilled deaf readers. The findings lead us to speculate that in deaf individuals, the ability to control what is processed or inhibited in the parafovea has the potential to impact in everyday reading in the deaf population. Supporting information S1 Appendix. (DOCX) S1 Dataset. Compiled raw data used for analysis the current study. (XLSX) S1 Materials. Materials used in the experiment. (ZIP) Author Contributions Conceptualization: Jiayu Tao, Zhao Qin, Li Zhang, Lu Liu, Guoli Yan, Valerie Benson. Formal analysis: Jiayu Tao. Funding acquisition: Guoli Yan. Methodology: Jiayu Tao, Zhao Qin, Zhu Meng, Li Zhang, Guoli Yan, Valerie Benson. Supervision: Guoli Yan, Valerie Benson. Writing – original draft: Jiayu Tao, Zhao Qin, Lu Liu, Guoli Yan, Valerie Benson. Writing – review & editing: Jiayu Tao, Zhu Meng, Li Zhang, Guoli Yan, Valerie Benson. References 1. Bavelier D, Neville HJ. Cross-modal plasticity:where and how? Nature Reviews Neuroscience 2002; 3 (6):443–52. Epub 2002/06/04. https://doi.org/10.1038/nrn848 PMID: 12042879. 2. Dye MW, Bavelier D. Attentional enhancements and deficits in deaf populations: An integrative review. Restorative Neurology and Neuroscience. 2010; 28(2):181–92. Epub 2010/04/21. https://doi.org/10. 3233/RNN-2010-0501 PMID: 20404407. 3. Bavelier D, Dye MW, Hauser PC. Do deaf individuals see better? Trends in Cognitive Sciences. 2006; 10(11):512–8. Epub 2006/10/04. https://doi.org/10.1016/j.tics.2006.09.006 PMID: 17015029. 4. Codina CJ, Pascalis O, Baseler HA, Levine AT, Buckley D. Peripheral visual reaction time is faster in deaf adults and British sign language interpreters than in hearing adults. Frontiers in Psychology. 2017; 8:50. Epub 2017/02/22. https://doi.org/10.3389/fpsyg.2017.00050 PMID: 28220085. 5. Hauthal N, Sandmann P, Debener S, Thorne JD. Visual movement perception in deaf and hearing indi- viduals. Advances in Cognitive Psychology. 2013; 9(2):53–61. Epub 2013/07/05. https://doi.org/10. 2478/v10053-008-0131-z PMID: 23826037. 6. Dye MW, Hauser PC, Bavelier D. Is visual selective attention in deaf individuals enhanced or deficient? The case of the useful field of view. PLoS One. 2009; 4(5):e5640. Epub 2009/05/23. https://doi.org/10. 1371/journal.pone.0005640 PMID: 19462009. 7. Bavelier D, Tomann A, Hutton C, Mitchell T, Corina D, Liu G, et al. Visual attention to the periphery is enhanced in congenitally deaf individuals. Journal of Neuroscience. 2000; 20(17):RC93. https://doi.org/ 10.1523/jneurosci.20-17-j0001.2000 PMID: 10952732. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 14 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students 8. Bavelier D, Brozinsky C, Tomann A, Mitchell T, Neville H, Liu G. Impact of early deafness and early exposure to sign language on the cerebral organization for motion processing. Journal of Neuroscience. 2001; 21(22):8931–42. Epub 2001/11/08. https://doi.org/10.1523/jneurosci.21-22-08931.2001 PMID: 9. Proksch J, Bavelier D. Changes in the spatial distribution of visual attention after early deafness. Journal of Cognitive Neuroscience. 2002; 14(5):687–701. Epub 2002/08/09. https://doi.org/10.1162/ 08989290260138591 PMID: 12167254. 10. Dye MW. Foveal processing under concurrent peripheral load in profoundly deaf adults. Journal of Deaf Studies and Deaf Education. 2016; 21(2):122–8. Epub 2015/12/15. https://doi.org/10.1093/deafed/ env054 PMID: 26657078. 11. Kelly LP, Barac-Cikoja D. The comprehension of skilled deaf readers: The roles of word recognition and other potentially critical aspects of competence. In: Cain K, Oakhill J, editors. Children’s comprehension problems in oral and written language: A cognitive perspective. New York: Guilford Press; 2007. p. 244–79. 12. Rayner K, Murphy LA, Henderson JM, Pollatsek A. Selective attentional dyslexia. Cognitive Neuropsy- chology. 1989; 6(4):357–78. https://doi.org/10.1080/02643298908253288 13. Casco C, Tressoldi PE, Dellantonio A. Visual selective attention and reading efficiency are related in children. Cortex. 1998; 34(4):531–46. Epub 1998/11/04. https://doi.org/10.1016/S0010-9452(08) 70512-4 PMID: 9800088. 14. Franceschini S, Gori S, Ruffino M, Pedrolli K, Facoetti A. A causal link between visual spatial attention and reading acquisition. Current Biology. 2012; 22(9):814–9. Epub 2012/04/10. https://doi.org/10.1016/ j.cub.2012.03.013 PMID: 22483940. 15. Rayner K. Eye movements and attention in reading, scene perception, and visual search. Quarterly Journal of Experimental Psychology. 2009; 62(8):1457–506. Epub 2009/05/19. https://doi.org/10.1080/ 17470210902816461 PMID: 19449261. 16. Liu L, Yan G. Effect of parafoveal visual attention enhancement in deaf reading:Evidence from disap- pearing text. Acta Psychologica Sinica. 2018; 50(7):715–26. https://doi.org/10.3724/SP.J.1041.2018. 17. Schotter ER, Angele B, Rayner K. Parafoveal processing in reading. Attention, Perception, & Psycho- physics. 2012; 74(1):5–35. Epub 2011/11/02. https://doi.org/10.3758/s13414-011-0219-2 PMID: 18. Dye MW, Hauser PC, Bavelier D. Visual skills and cross-modal plasticity in deaf readers:Possible impli- cations for acquiring meaning from print. Annals of the New York Academy of Sciences. 2008; 1145:71–82. Epub 2008/12/17. https://doi.org/10.1196/annals.1416.013 PMID: 19076390. 19. Be ´ langer NN, Slattery TJ, Mayberry RI, Rayner K. Skilled deaf readers have an enhanced perceptual span in reading. Psychological Science. 2012; 23(7):816–23. Epub 2012/06/12. https://doi.org/10.1177/ 0956797611435130 PMID: 22683830. 20. Yan M, Pan J, Be ´ langer NN, Shu H. Chinese deaf readers have early access to parafoveal semantics. Journal of Experimental Psychology:Learning, Memory, and Cognition. 2015; 41(1):254–61. Epub 2014/07/08. https://doi.org/10.1037/xlm0000035 PMID: 24999711. 21. Eriksen BA, Eriksen CW. Effects of noise letters upon the identification of a target letter in a nonsearch task. Perception & Psychophysics. 1974; 16(1):143–9. https://doi.org/10.3758/BF03203267 22. Mayberry RI, del Giudice AA, Lieberman AM. Reading achievement in relation to phonological coding and awareness in deaf readers: A meta-analysis. Journal of Deaf Studies and Deaf Education. 2011; 16 (2):164–88. Epub 2010/11/13. https://doi.org/10.1093/deafed/enq049 PMID: 21071623. 23. Westfall J. PANGEA: Power ANalysis for GEneral Anova designs. 2015. Available from: http:// jakewestfall.org/publications/pangea.pdf. 24. Zang C, Fu Y, Bai X, Yan G, Liversedge SP. Investigating word length effects in Chinese reading. Jour- nal of Experimental Psychology: Human Perception and Performance. 2018; 44(12):1831–41. https:// doi.org/10.1037/xhp0000589 PMID: 30475051. 25. Zhang F, Li D. Deaf students’ syntactic awareness of the “shi. . .de” construction in sentence reading. Acta Psychologica Sinica. 2017; 49(11):1383–91. https://doi.org/10.3724/SP.J.1041.2017.01383 26. Zhang H, Lv H, Wu L. The development of Chinese curriculum for deaf college students. Beijing: Huaxia Press; 2009. 27. Li D, Hu K, Chen G, Jin Y, Li M. Combined Raven’s Test (CRT) Shanghai area test report. Journal of Psychological Science. 1988; 11(4):27–31. https://doi.org/10.16719/j.cnki.1671-6981.1988.04.007 28. Zhang H, Wang X. Standardization research on Raven’s standard progressive matrices in China. Acta Psychologica Sinica. 1989; 21(2):113–21. PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 15 / 16 Effects of parafoveal distractors on foveal lexical decision in deaf students 29. Dare N, Shillcock R. Serial and parallel processing in reading: Investigating the effects of parafoveal orthographic information on nonisolated word recognition. Quarterly Journal of Experimental Psychol- ogy. 2013; 66(3):487–504. Epub 2012/09/07. https://doi.org/10.1080/17470218.2012.703212 PMID: 30. Snell J, Grainger J. Parallel word processing in the flanker paradigm has a rightward bias. Attention Per- ception & Psychophysics. 2018; 80(6):1512–9. Epub 2018/06/03. https://doi.org/10.3758/s13414-018- 1547-2 PMID: 29858975. 31. Cai Q, Brysbaert M. SUBTLEX-CH: Chinese word and character frequencies based on film subtitles. PLoS One. 2010; 5(6):e10729. Epub 2010/06/10. https://doi.org/10.1371/journal.pone.0010729 PMID: 32. Liu L. The effect of parafoveal visual attention enhancement in deaf reading [dissertation]. Tianjin Nor- mal University; 2017. 33. Bates D, Maechler M, Bolker B, Walker S. lme4: Linear Mixed-Effects Models using ’Eigen’ and S4. 2017. Available from: https://CRAN.R-project.org/package = lme4 34. R Development Core Team. R: A language and environment for statistical computing, R Foundation for Statistical Computing, Vienna, Austria. 2017. Available from: http://www.R-project.org/. 35. Be ´ langer NN, Rayner K. Frequency and predictability effects in eye fixations for skilled and less-skilled deaf readers. Visual Cognition. 2013; 21(4):477–97. Epub 2013/08/27. https://doi.org/10.1080/ 13506285.2013.804016 PMID: 23976872. 36. Baayen RH, Davidson DJ, Bates DM. Mixed-effects modeling with crossed random effects for subjects and items. Journal of Memory and Language. 2008; 59(4):390–412. https://doi.org/10.1016/j.jml.2007. 12.005 37. Be ´ langer NN, Lee M, Schotter ER. Young skilled deaf readers have an enhanced perceptual span in reading. Quarterly Journal of Experimental Psychology. 2018; 71(1):291–301. Epub 2017/04/28. https://doi.org/10.1080/17470218.2017.1324498 PMID: 28447500. 38. Inhoff AW, Starr M, Shindler KL. Is the processing of words during eye fixations in reading strictly serial? Perception & Psychophysics. 2000; 62(7):1474–84. https://doi.org/10.3758/BF03212147 PMID: 39. Rayner K, Pollatsek A, Ashby J, Clifton C Jr. Psychology of reading. New York: Psychology Press; PLOS ONE | https://doi.org/10.1371/journal.pone.0221891 September 12, 2019 16 / 16

Journal

PLoS ONEPublic Library of Science (PLoS) Journal

Published: Sep 12, 2019

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create folders to
organize your research

Export folders, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off