ABSTRACT. The development of stone tools is thought to parallel the evolution of human cognitive and attentional capacities. One possibility is that sophisticated tool production techniques developed alongside brain systems that process action relevant features of natural objects. In a series of experiments, we investigated eye movements and visual attention during the viewing of Palaeolithic stone tools. Participants were presented with images of hand-axes, roughly worked pebbles, unworked pebbles and hi-bred images combining a hand-axe outline with an unworked stone texture. Images were also analysed using a graph based visual salience algorithm in order to identify which components of the image were most visually conspicuous as opposed to functionally relevant. Overall, the findings suggest that visual exploration is directed towards functional aspects of tools such as the grasping point at the base and the knapped cutting surface rather than salient internal and external features of the tool highlighted by the saliency model. A follow-up study used "real-world" mobile eye tracking during handling of the same stone objects and found an identical distribution of fixation dwell times across the tool regions. It is concluded that functional aspects of Palaeolithic tools strongly influence viewing patterns over and above the visual salience of images features.

ABSTRACT. Social anxiety disorder (SAD) is among the most common mental health disorders in children and SAD negatively affects quality of life. Cognitive models of anxiety propose that attentional biases toward threatening information play etiological and maintenance roles in anxiety. Threat processing is related to automatic attentional orienting to threat, difficulty disengaging from task-irrelevant threatening stimuli, and vigilance-avoidance of threat. The current study examined attentional processing in relation to SAD symptoms in N = 38 children (25 males) aged between 9-11 years old. We employed the Remote Distractor Paradigm (RDP) and presented faces that displayed different expressions (happy, angry, neutral) to act as distractors in a task where participants were required to ‘look at’ a non-face target presented at a lateral location. Distractors could appear at central, parafoveal and peripheral eccentricities, and we recorded and analysed eye movements for the task. Higher symptoms of SAD were associated with more saccadic errors for angry compared to happy face distractors, and slower latencies to the target in the presence of angry compared to happy and neutral face distractors, regardless of the distractor eccentricity. The findings indicate that social anxiety is characterised both by orienting to threat and difficulty disengaging from task-irrelevant threat.

ABSTRACT. In an earlier study, we found that the fusiform face area and parahippocampal place area were activated not only by viewing face and house images, but also by gaze paths corresponding to face and house viewing. While viewing a face, we look at specific regions of the face in a specific sequence. How these viewing patterns of faces differ from those of houses is not known. To study differences in gaze paths between categories, we compared viewing patterns of faces with houses and a third category, leaves, by recording eye movements while participants viewed face, house and leaf images. We found that locations, but not durations, of fixations contributed toward identifying the three categories using a classifier. Distinguishing face from house gaze paths had higher accuracy than distinguishing house from leaf gaze paths or face from leaf gaze paths. We also found that early fixations were better predictors of category compared to later fixations. The second and third fixations contributed most to category identification. The differences between the gaze paths associated with these three categories arise early and lie in the locations of the fixations.

ABSTRACT. In this study, we used a dual eye-tracking setup to investigate face scanning behavior during video-mediated interactions between parents and their preadolescent children (8-10 years). 81 parent-child dyads engaged in two brief conversations about cooperative and conflictive family topics. We used eye tracking, audio-, and video-recordings to assess what regions of the face are looked at during episodes of speaking and listening, and whether patterns of gaze and speech were influenced by the topic of conversation. Regarding speech, our results show that children spoke more in the cooperation-scenario whereas parents spoke more in the conflict-scenario. Regarding gaze, we found that both parents and children looked more at the other's mouth region while listening compared to while speaking, which converges with previous studies about face scanning during speech perception. Furthermore, we found that parents gazed slightly more at the eyes of their children in the conflict-scenario compared to the cooperation-scenario. Results are discussed in terms of the role that parents and children take during cooperative and conflictive interactions, and how gaze behavior may support and coordinate such interactions.

ABSTRACT. As humans move through parts of their environment, they meet others that may or may not try to interact with them. Where do people look when they meet others? Does one’s looking behavior depend on the other person’s behavior, and does one’s looking behavior predict subsequent interaction? We equipped 23 participants with a wearable eye tracker and had them walk through a university building. On the way, they encountered nine “walkers”. Walkers were instructed to e.g., ignore the participant, greet him or her, or attempt to hand out a flyer. Participants’ gaze behavior was analyzed using a fully automated area-of-interest analysis based on OpenPose pose estimation. The participant’s gaze was mostly directed to the currently relevant body parts of the walker. Thus, the participant’s gaze depended on the walker’s actions. Individual differences in participants’ gaze were consistent across walkers, regardless of the walkers’ behavior. Our findings are discussed in the light of existing theories of gaze allocation, which may be extended by taking social motivation into account.

ABSTRACT. Teachers need to distinguish between irrelevant and relevant classroom information in order to meaningfully interpret and react to this information (Wolff et al., 2021). To understand teachers’ information processing in real classrooms, mobile eye-tracking is used to assess teachers´ visual attention (Stürmer et al., 2017; McIntyre et al., 2019). Since the direct comparison of diverse classroom lessons in real life poses challenges, we have limited knowledge about teachers’ visual focus of attention during different didactic activities. As these activities, however, place different demands on the teachers’ instructive and learning-supporting actions (i.e. verbal and non-verbal, including eye-movements), we developed a standardized classroom teaching simulation to examine novice teachers´ focus of attention during teacher- and student-led activities. Gaze behaviour of N=14 novice teachers’ was recorded with Tobi Glasses 3 (50 Hz). Eye movement measures (i.e. fixation frequency, average fixation duration) for AOIs representing seven different learner profiles were analysed and compared for a teacher-, and student-led activity. Results show that novice teachers significantly differ in their fixation frequency on the student with active off-task behaviour during both activities. Additionally, we observe a slight tendency of novice teachers to distribute their attention more evenly while teaching teacher-led activities.

ABSTRACT. Reading multiline texts in logographic and alphabetic writing systems requires return sweeps. We examined presentation format on return sweeps in native Chinese (Experiment 1) and English readers (Experiment 2) to determine how text alignment and column setting influence reading behaviour across two different written orthographies. Participants read left-aligned and fully justified multiline texts presented in one and two columns. We reasoned that more return sweeps and greater variability in line length might be disruptive to reading at a global (text reading time) and local (sweep targeting/correction) level. Chinese readers fixated further to the extremes of the final character of a line for text presented in two than one column, and for left-aligned than fully justified text. These effects indicate less reliance on line-final parafoveal processing for text in two columns, and for left-aligned text. Participants were more likely to undershoot line-initial target words following a long than a short return sweep due to greater saccadic targeting error for longer return sweeps. Similarly, undersweep fixation durations were longer for text in two columns than one column suggesting costs associated with more return sweeps. We will assess the results for English readers (currently under analysis) against those of the Chinese readers.

ABSTRACT. The relationship between the eye-voice span (EVS) and eye movement control has been documented recently in single line reading studies (Inhoff et al., 2011; Laubrock & Kliegl, 2015), yet no recent studies have been carried out with multiline stimuli. Therefore, this relationship was examined across line boundaries by comparing fixations adjacent to the return-sweep and intra-line fixations in a group of developing readers. The EVS had a greater impact on line-final and undersweep fixations compared to intra-line fixations. However, the EVS impacted accurate line-initial and intra-line fixations similarly. Our findings replicate previous work (Laubrock & Kliegl, 2015) of longer fixation durations in response to large EVSs. We extend this work by showing that this effect is greater for line-final and undersweep fixations in children. Additionally, we found that accurate line-initial fixations were much longer than intra-line fixations. Taken together, these suggest that return-sweep saccades may be more costly during oral reading and may contribute significantly to longer reading times in children who primarily read aloud.

ABSTRACT. ******* THIS SUBMISSION IS FOR A SYMPOSIUM EYE MOVEMENTS DURING MULTILINE READING *******

A common problem in eye tracking research is vertical drift – the progressive displacement of fixation registrations on the vertical axis that results from a loss of calibration over time. This is esspecially problematic for experiments that involve the reading of multiline passages, where it is critical that fixations on one line of text are not erroneously recorded on another. Correction is often performed manually, but this is time-consuming and error-prone. Various methods have previously been proposed for the automated correction of vertical drift in the context of reading, but these have largely been developed in isolation with little attempt to systematically evaluate them, meaning that drift correction and line assignment techniques have been moving forward blindly. We document the major algorithms and evaluate them using both simulated and natural eye tracking data. Our results indicate that different algorithms are better suited to different types of drift phenomena, allowing us to offer evidence-based advice on algorithm selection.

ABSTRACT. In this talk, we will explore the consequences of assigning fixations to lines automatically when analysing data from multi-line experiments. Using data from the MECO corpus we computed correlations for various eye-movement measures based on manually annotated data and data automatically pre-processed using different line-assignment algorithms. Correspondence for first-fixation duration, gaze duration, and total reading time was generally high (correlations varying r = .81-.96 between the different algorithms). By contrast, correlations for skipping probability (r = .53-.85) and go-past time (r = .38-.72) were substantially lower and should thus be interpreted with caution. For all measures, there were consistent differences between different assignment algorithms. In a next step, we investigated whether the effects of word length, word frequency, and predictability differed between manually annotated data and the different line-assignment algorithms. Results show that effect sizes obtained using automatically processed data are qualitatively and quantitatively similar to those obtained using manually annotated data with little differences between algorithms. We discuss our findings with regard to different strategies how to (semi-)automatically pre-process data from multi-line reading experiments. In addition, we will elaborate on further methodological challenges involved in such experiments and provide practical recommendations for their analysis.

ABSTRACT. In recent years, researchers have explored the possibility of using eye movements to measure reading comprehension. Findings suggest that eye-movement measures can predict performance on reading comprehension assessments, but that the relationship between eye-movement behaviour and reading comprehension is mediated by differences in task demands between assessments. We compared word-based eye-movement measures to measures of scanpath regularity as predictors of reading comprehension scores. Scanpaths go beyond word-based eye-movement measures and capture the global pattern of fixations during reading. Scanpath regularity is relative to other participants’ eye movements and factors in readers’ adaptation to task demands. It may therefore be less sensitive to differences in task demands between assessments. We used a dataset in which participants were administered three reading comprehension assessments while their eye movements were monitored: the York Assessment for Reading Comprehension (YARC), the Gray Oral Reading Test (GORT), and the sentence comprehension subset of the Wide Range Achievement Test (WRAT). Results showed that word-based and scanpath measures both made unique contributions as predictors of reading comprehension scores, suggesting that scanpaths capture effects that are not caught by word-based measures and vice versa. Nevertheless, both types of measures were influenced by differences in task demands between reading comprehension assessments.

ABSTRACT. Return-sweep saccades are eye movements that are made when readers move from the end of one line in a passage to the beginning of the next line. Readers often under-shoot a return-sweep saccade and require a subsequent, corrective saccade to fixate at the appropriate location at the beginning of a line. Given that prior work examining oculomotor deficits in readers with Schizophrenia (Sz) has primarily focused on single-line and single-word reading tasks, return-sweep saccades have not yet been examined in Sz. To assess return-sweep saccades in Sz, we analyzed an existing dataset (Dias et al., 2021, Schizophrenia Bulletin) in which participants read multi-line passages in four different line-spacing conditions. The probability of making a corrective saccade following a return-sweep was significantly higher in readers with Sz compared to healthy controls. Because visual acuity constraints do not permit detailed lexical processing of line-initial words when return-sweep saccades are initiated, our findings provide evidence for low-level oculomotor deficits in Sz during reading. Additionally, the probability of making a corrective saccade was significantly higher for condensed-spaced passages compared to both the double- and triple- spaced passages, which is consistent with prior research showing that increased visual crowding disrupts saccade targeting.

ABSTRACT. The perception and evaluation of art might be governed by both stimulus-driven aspects and expectations of the observer. By framing actual artwork and actual scientific depictions as being either artwork or scientific depictions, we were able to investigate the individual contributions of both factors on viewers’ gaze behavior and aesthetic judgments. With Experiment 1, we identified a set of images that were equally likely classified as being either artwork or scientific depiction by art laypersons. In Experiment 2, we measured art laypersons’ eye-movements as well as their aesthetic judgments regarding those images (half being actual artwork and half being actual scientific depictions). Importantly, we told one group of participants that all images were artwork, whereas we told another group of participants that all images were scientific depictions. Our results indicate that gaze behavior was influenced by the actual image type but not by framing. Actual artwork led to fewer and longer fixations, a shorter scanpath, a shorter distance of fixations to the screen center, and a lower similarity of viewing patterns across participants. Further, actual artwork was rated as being more complex but less meaningful than scientific depictions, with no differences regarding perceived structuredness and liking. Framing the images as artwork led to no changes in perceived meaningfulness, complexity, or structuredness but to lower ratings on liking. We conclude that while the dichotomy between art and non-art influenced art laypersons’ gaze behavior and aesthetic judgments, this influence was primarily driven by bottom-up factors rather than a specific cognitive art schema.

ABSTRACT. It is a common observation that visitors dedicate a relatively short time period to individual exhibits in relation to their total length of stay in the museum. However, some artworks are looked at significantly longer than others, while some are barely noticed at all. Likewise, some artworks tend to leave a lasting impression, whereas others are not remembered in particular. Intriguingly enough, eye-catching exhibits are often the same among different visitors. Based on an interdisciplinary mobile eye tracking study conducted by the Lab for Cognitive Research in Art History (CReA) of the University of Vienna in cooperation with the University of Tübingen, the EVAlab and the Austrian Gallery Belvedere in two successive years, this analysis combines quantitative and qualitative data in order to measure the attraction potential of artworks within an exhibition context. While previous studies have already pointed out the cumulative effect of exhibits, the focus thus shifts to the inherent potential of single artworks to capture the visitors’ gazes.

ABSTRACT. The effects of painting restoration on the visual perception of museum visitors

This proposal will present a method to understand the influence of painting conservation on museum visitors' perception, visual exploration, and aesthetic experience. The visual exploration of two-dimensional figurative representations activates universal vision abilities, which are culturally and historically implemented. By combining psychological and cognitive studies with historical and eye-tracking analysis, this interdisciplinary proposal explores how our visual experience of paintings simultaneously addresses three pictorial experience levels: material, mental and historical. To deepen our knowledge of the vision and interpretation of paintings, we will focus on an in-depth empirical case study: the restoration of the Isenheim Altarpiece (1512-1516), painted by Matthias Grünewald. This Germanic polyptych, a masterpiece of Western art, is studied in its natural environment of observation, the Unterlinden Museum. This approach will allow us to engage in a comparative study of how different observers view the same unvarnished painting before and after restoration and integrate the altarpiece's material aspects with the observers' visual perception. This research, conducted at Centre de Recherche et de Restauration des Musées de France (C2RMF) and the Center for Research on Arts and Language (CRAL) of the École des Hautes Etudes en Sciences Sociales (EHESS), adds to the traditional scientific analyzes of the materiality of the works mentioned above the eye-tracking method but also the cognitive and social science method. We aim to contribute to integrating "hard" and "human" sciences in the service of innovation in art research and technology. The central aspect of the restoration of the Isenheim Altarpiece retained in this proposal is the cleaning of the panels. One of the main questions posed by us is to determine how the colors, the shapes, and the details of the paintings, which become more intense after the lightening of the brown-yellow varnish, which currently protects the pictorial layer, change the visual perception and the impressions of observers of the artwork. Another question concerns the potential effect of restoring the images, weathered over time by the oxidation of the varnish, which has become opaque and darkens the panels, on the "historical" imagination of the altarpiece. We also wonder about the impact of conservation-restoration choices on the aesthetic experience of the people who look at the artwork—for example, leaving visible signs of the effects of time and even the marks of accidents on the paintings. This study will use two experimental methods that are potent instruments in interdisciplinary research. The first is the use of an eye-tracker. The eye-tracker is a versatile instrument that can be exploited at different levels of cognitive resolution. This study combines eye-tracking results with qualitative methodologies (self-comments, questionnaires). The second innovative experimental aspect of the presentation will be computer models to analyze visual perception. Such models are helpful in that objective measures can be compared to a person's subjective perception. We propose to use this type of model to study the perception of the ten panels of the Issenheim Altarpiece before and after the varnish cleaning. From this point of view, detecting AOI (Areas of Interest) in the images will contribute to understanding the visual exploration of the altarpiece panels and

ABSTRACT. Smartphones are quickly becoming the primary device that many use to access the internet and the billions of images it makes available. This new context provides an intimate experience of a series of small images in quick, and nearly endless, succession. As this becomes the default image viewing setting, how do our experience of visual information and habits of viewing change? The current study investigates whether art viewing behaviour changes in relation to average daily screen time. We present measurements of attention span and memory retention as well as a self-reported aesthetic experience. Results confirm some past research on the effects of smartphones on general cognition but also demonstrate some emerging viewing behaviours.

ABSTRACT. According to aesthetic philosopher Richer Wollheim, when looking at painting our attention oscillates constantly between two elements: its subject matter and its surface qualities, a perceptual phenomenon which he calls "twofoldness". The existence of two-fold perception has been theorized in art historical writing at least since the nineteenth century, in parallel with the development of modern art. Nonetheless, unlike its art historical precedents which focus on aesthetic phenomenology, Wolheim´s theory considers twofoldness to be also observable from a behavioural perspective, as assertion which calls for an empirical investigation.

This paper reports the results of two eye-tracking investigations on twofoldness: one conducted in a laboratory setting, one in the museum. We hypothesised that twofoldness, as described by Wollheim, could find a physiological correspondence in a viewer´s eye-movements and that it could be described as an the alternation of local and global aspects in a scanpath. In order to test this hypothesis, we examined the behaviour of viewers looking at modern art. Egon Schiele´s paintings, in which surface qualities are evident also to non-experts and play an active role in shaping identifiable subjects, lend themselves perfectly to this type of investigation.

ABSTRACT. Many models have been proposed to account for eye-movement behaviour during reading. Most share the assumption that readers’ saccades are controlled in a top-down manner by the needs of ongoing word-identification processes. However, it remains undetermined whether eye fixations during reading (on average 225 ms) are long enough to enable predominant top-down control. Here we assessed the time course of visuo-motor and linguistic processes during reading using Fixation-Related Potentials (FRPs). Twenty participants read 316 French sentences, as well as 316 pseudo-sentences (all letters replaced by the letter “z”), while their eye movements and their electroencephalogram (EEG) were co-registered. FRPs were analysed using Unfold, an EEG-deconvolution and (non-)linear-modelling toolbox. We entered sentence type, frequency and length of the fixated word, and incoming-saccade launch site, as predictors. Our preliminary results showed an effect of sentence type on the N1 component (~150 ms) but mainly in near launch-site cases, thus when the (z-)word could be previewed during the previous fixation. In contrast, there was no effect of word frequency before 200 ms from fixation onset, thus suggesting that the effect of sentence type was visual rather than linguistic in nature. Therefore, word-identification processes might be too slow to efficiently control saccades during reading.

ABSTRACT. We recorded eye movements while participants (N=36) read sentence pairs in German. The second sentence included a target word with high vs. low frequency, which was either of high or low predictability based on context provided in the first sentence. Using this methodology, we obtained a pattern of oculomotor data that is highly compatible with the existing literature, including parafoveal-on-foveal effects. Concurrently, we utilized the rapid sampling frequency of near-infrared spectroscopy (NIRS) to examine neural correlates of word frequency and predictability in the occipital and orbitofrontal cortex. The onset of fixations within target words served as trigger events for the recording of hemodynamic responses. We observed increased activation in the right ventral occipital cortex when the fixated target word N was of low frequency, attributable to increased processing load during saccade planning. Unpredictable low frequency words increased activity in the left dorsal occipital cortex upon the fixation of the preceding word N-1, presumably due to interference with top-down modulated expectation. There was no such interaction in the orbitofrontal cortex, implying that it is not involved in the resolution of predictions during natural reading. Implications of these data for theories of eye movement control in reading will be discussed.

ABSTRACT. Dyslexia is a learning disability and its etiology is still debated. Phonological deficit and visuo-attentional impairment are two of the most discussed hypotheses to explain the poor reading abilities reported in dyslexia population. In this study, both eye movements and event-related potentials (ERP) were simultaneously recorded in two groups of university students, with and without dyslexia, during a phonological lexical decision task. Five categories of stimuli implicating various levels of sublexical or lexical orthographic and phonological processing were used (i.e., French content words, pseudohomophones, pseudowords, consonant strings, and symbols). We examined the number and the duration of fixations, the amplitude of saccades as well as three reading related ERP components, namely the N170, N320 and N400. Preliminary data show more and longer fixations, and several saccades in the dyslexics compared to controls, as well as reduced ERP amplitudes in the three ERP markers of interest. Taken together, these findings lend support to our assumption that dyslexia is a multifactorial deficit implicating both visual and phonological impairments.

ABSTRACT. Eye tracking studies reveal that readers make shorter fixations on foveated target words when previously perceived parafoveal preview words were higher frequency or linguistically similar to the target. This parafoveal preview benefit in eye movements does not reveal the extent to which parafoveal word processing, and the integration of that word with foveally obtained information, continues after saccade plans have been initiated. Therefore, we coregistered EEG and eye movements during a gaze-contingent display change paradigm and measured fixation related potentials (FRPs) upon foveal fixation. Eye movements showed primarily an effect of preview frequency, suggesting that saccade planning is based on the familiarity of the parafoveal input. FRPs, on the other hand, demonstrated a disruption in downstream processing (i.e., the N400 component) when parafoveal and foveal input differed, but only when the parafoveal word was high frequency. This suggests that lexical processing continues after the eyes have moved away from a word and that eye movements and FRPs provide distinct but complementary information about lexical processing in reading. These findings put constraints on models of reading by suggesting that lexical processes that occur before an eye movement program is initiated are qualitatively different from those that occur afterward.

ABSTRACT. Eye movement and ERP studies report conflicting findings concerning age differences in the effects of sentential constraint on reading. Whereas eye movement studies report larger effects of sentential constraint by older relative to younger adults (see Zhang et al., 2022), ERP findings suggest that context effects are diminished or delayed for older readers (see Payne & Silcox, 2019). These contrary findings may reflect methodological differences, including the use of unnatural sentence displays in ERP research. To address these limitations, we used a co-registration technique to record eye movements (EMs) and fixation-related potentials (FRPs) simultaneously while young (18-30 years) and older (65+ years) adults read sentences naturally, and used sentential constraint to manipulate the predictability of specific target words. Eye movement analyses were conducted over all data (full EM dataset) and only data matching FRPs, while FRPs were analysed to capture both early (P1 and N1) and later (N400 and P600) components. Both EM datasets and early FRP components showed main effects of age-group and predictability, while the full EM dataset and later FRP components revealed larger predictability effects for older adults. We argue that this novel approach provides compelling evidence that older adults rely more on sentential constraint in reading.

References Zhang, J., Warrington, K.L., Li, L., Pagán, A., Paterson, K.B., White, S.J., & McGowan, V.A. (2022). Are older adults more risky readers? Evidence from meta-analysis. Psychology and Aging, 37(2), 239-259. http://dx.doi.org/10.1037/pag0000522 Payne, B.R., & Silcox, J. (2019). Aging, context processing, and comprehension. Psychology of Learning and Motivation, 71, 215-264. https://doi.org/10.1016/bs.plm.2019.07.001

ABSTRACT. Evidence of processing costs for unexpected words presented in place of a more expected completion remains elusive in the eye-movement literature. The current study investigated whether prediction error costs might depend on the source of constraint violation provided by the prior context. Participants’ eye movements were recorded as they read predictable words or unpredictable alternatives that were semantically related or unrelated in three-sentence passages that were matched on cloze probability but differed in whether the source of constraint originated solely from the local context provided by the final sentence of the passage, from the global context provided by the first two semantically rich sentences of the passage; or from the combined local and global contexts. Linear mixed models assessed the relative impact of the type of prior constraint on readers’ early and late processing of unexpected input. We interpret our results within a hierarchical generative framework of language comprehension and consider their implications for existing theoretical models of predictive processing.

ABSTRACT. Objectives: Eye tracking is a valuable in assessing individuals gaze behaviour but does not unveil why they are engaging in certain practices. To address limitations in sport eye tracking research the present paper aims to investigate gaze behaviours of soccer officials during successful and unsuccessful offside decisions, but also why.

Methods: 20 male active amateur qualified (Level 4-7) soccer officials (Mage 22.5 SD 4.61 yrs) with an average experience of 41-50 games wore the SMI eye tracker during an applied attack versus defence drill. While reviewing the eye tracking footage, retrospective semi-structured interviews were conducted (M=20.4 min; SD=6.2; Range 11.7 – 26.8 min). Inductive thematic analysis was then performed.

Findings and Discussion: Key constructs of offside, decision making, obstacles and emotions were summarised as the higher order themes while making offside decisions. Gaze anchoring was highlighted to be a successful technique to allow officials to see all relevant information, whereas the type of offside was emphasised to be a key factor in correct interpretation.

Conclusions: Key constructs have been identified and explained, which can be shared with soccer officials through training regimes. Eye tracking technology has also been shown to be a useful and innovative reflective tool to assist in the understanding of individuals gaze behaviours.

ABSTRACT. According to chunking and template theories of expertise, experts learn how to group together individual features into larger meaningful patterns called “chunks”. Here, we tested if expert musicians use “chunks” to process music scores by re-analyzing an eye tracking dataset from a music expertise study by Maturi & Sheridan (2020). In Maturi & Sheridan (2020)’s study, participants completed a music-related visual search task while their eye movements were monitored. We extended Maturi & Sheridan (2020) by analyzing fixation locations and durations within bars of music. The experts allocated more attention to the beginning and end of a bar than non-musicians, as shown by a higher proportion of fixations and longer fixation durations compared to non-musicians. Similar to this pattern of results, skilled chess players are more likely than less skilled players to fixate near the edges of squares on a chessboard (de Groot and Gobet, 1996). One possible explanation for why experts fixate closer to the edge of a bar of music is that they are using parafoveal processing to encode larger patterns (or chunks) that extend across multiple bars. Our findings suggest that there are qualitative differences in how experts and non-musicians allocate their attention while processing music scores.

ABSTRACT. Together with the growing use of instructional videos in educational settings, comes an increased demand for concrete research-based principles concerning the design of such videos. This study explored whether a video’s background affects students visual information processing, and learning outcomes. In a between-group design, participants (N=47, age 14-17, 35 female) were assigned to one of three conditions: video with a neutral, authentic, or off-topic background. We measured prior-knowledge and working memory capacity (WMC), recorded eye tracking data during the learning process, and measured learning outcomes and learning experience. The eye tracking data revealed that a neutral background distracted the viewers the least, allowing them to pay better attention to relevant parts of the video. Also, learners self-reported that the off-topic background was distracting, however, the negative effect on the learning outcomes did not reach statistical significance. Furthermore, the WMC level had a significant impact on visual information processing and learning outcomes. We did not find a positive effect of the authentic background. This study suggests that backgrounds in learning material can distract visual processes and be experienced as such. Hence, we conclude that educators should preferably use a neutral background for their learning video’s, especially for learners with lower WMC.

ABSTRACT. Background: Over 80% of those diagnosed with Parkinson’s Disease (PD) progress to PD with dementia (PDD) – with mild cognitive impairment an intermediary step (PD-MCI) - suggesting a complex spectrum of pathology concomitant with aging. Biomarkers sensitive and specific to all stages of disease are required if clinically relevant diagnostics are to be developed. Objectives: We used video-based eye tracking and machine learning to develop a simple, non-invasive screening tool sensitive to PD and the stages of PDD. Methods: From 121 PD (46 CN / 47 MCI / 20 Dementia / 8 Other) and 108 healthy controls we collected video-based eye tracking measures on an interleaved pro/anti-saccade task. Features of pupil, blink, and saccade behaviour were used to train a machine-learning classifier to predict confidence scores for PD/PD-MCI/PDD diagnosis. Results: The overall Receiver Operator Characteristic Area Under the Curve (ROC-AUC) of the classifier was 0.87, whereas the subtypes showed progressively increased AUC with progressing disease, with the AUC of PDD alone being 0.94. The classifier had a sensitivity of 83% and a specificity of 78%. Moreover, the confidence scores predicted PD motor and cognitive scores at baseline and longitudinally. Conclusion: We operationalized measures of saccade, pupil, and blink from video-based eye tracking to achieve high sensitivity to the landscape of PD heterogeneity. This work provides a framework for true clinical utility, where a simple eye tracking task could be used to screen for PD/PDD and predict outcomes.

ABSTRACT. Domain experts and novices differ in the way they sample visual information. Information about experts' allocation of attention can be used for educational as well as for diagnostic purposes. However, eye tracking data on expertise in medical imaging is scarce, and in particular absent for echocardiography, he key imaging and diagnostic method in cardiology. How is expert diagnosticians' allocation of attention characterized? Here we used eye tracking on echocardiographic loops of patients with reduced ejection fraction and controls to compare 15 national or international board certified diagnostic experts with 15 beginning medical students. Participants viewed 21 one-second video loops of apical four-chamber views and were asked to assess the systolic function of the left ventricle (as a diagnostic task) and image quality (as a visual control task), while their eye movements were recorded at 1000Hz using an EyeLink Plus. Rating data show that experts had a higher sensitivity in discriminating between patient and control loops. AOI analyses show that experts fixated diagnostically relevant areas such as the left ventricle wall earlier and more often, but more importantly, novices spent a lot more time on visually salient, but diagnostically irrelevant regions for the current diagnostic task, like the mitral valve. It appears that experts have learnt to ignore the salient motion cue presented by the mitral valve, and to focus on diagnostically more relevant areas. In addition to enhancing the training of echocardiography, eye-tracking results might be useful to aid in characterizing echocardiographic expertise objectively.

ABSTRACT. Introduction. Ocular microtremor (OMT) is a natural oculomotor activity which originates in the oculomotor area of the brainstem. The amplitude of tremor oscillations is normally 20-40 arc seconds, the average oscillation frequency is 87 Hz. The working hypothesis - the parameters of eye tremor can be objective and adequate indicators of the mental state. Method. The parameters of OMT were recorded using an optical system elaborated by the authors and calculated using the author's program. Video recordings with the duration 0.3-1.5 were performed from the outer side of the sclera. Participants: 9 persons without psychiatric and neurological diagnoses, 7 patients with schizophrenia and 3 patients with delusional disorder. Results. Analysis of variance showed a significant difference between groups in certain ranges of frequency (ft) and amplitude (at) of tremor. Differences were observed between patients with schizophrenia and delusional disorder in ft71 (p=0.02), at67 (p=0.0009), at71 (p=0.00001), at110 (p=0.00006), mentally healthy and schizophrenic patients in ft71 (p=0.02) and ft110 (p=0.00001), healthy participants and patients with delusional disorder in at67 (p=0.001), at71 (p=0.001), and at110 (p=0.0001). Conclusion. OMT measurement could be a useful tool for clinical diagnosis of a number of conditions.

Supported by Saint Petersburg State University (ID 37590964)

ABSTRACT. Predictive remapping is considered to be a crucial mechanism of maintaining visual stability. In this study, we aimed to investigate the role of attention in predictive remapping while performing a dual task paradigm which included two conditions: with and without saccadic remapping. In the first task participants remembered the clock hand position either after a saccade to the clock face (saccade condition requiring remapping) or after the clock being displaced to the fixation point (fixation condition with no saccade). In the second task they reported the remembered location of a dot on the periphery appeared for 1 s. We hypothesized that performance in these two tasks would interfere in the saccade condition, but not in the fixation condition, due to the attentional demands. We revealed that for the clock task, answers in the saccadic condition tended to underestimate the actual clock hand position by approximately 37 ms while responses in the fixation condition were close to veridical. The results point at significant interaction between the two tasks with decreased predicted accuracy in the clock task for increased error in the localization task, but only for the saccadic condition. These results demonstrate the key role of attention in predictive remapping.

ABSTRACT. Atypical joint attention has been frequently reported in individuals with autism spectrum condition (ASC) but, empirical studies suggest intact gaze following in this population. The current study investigated gaze following under an emotional gaze cueing paradigm (GCP) context in young children with and without ASC. Real faces were presented centrally, and the task was to look at a parafoveal or peripheral target which was displayed following a gaze change, and on some trials an emotion change, in the central faces. The gaze change was either congruent (cued) or incongruent (mis-cued) with the location of the target. In an angry gaze condition the centrally-presented face switched expression following a shift of the eyes to the left or right, whereas in the neutral face condition there was no expression change. Typical children exhibited an increased gaze cueing effect (faster to look to the lateral target correctly cued by gaze relative to the mis-cued condition) for angry faces compared to neutral faces. This effect was absent in the ASC group who also showed disengagement difficulties from the central faces regardless of the face type and gaze direction. The findings provide evidence for atypical joint attention in an emotional GCP context in ASC.

ABSTRACT. The progressive neurodegeneration in Parkinson’s disease (PD) leads to impairments in autonomic, motor, and cognitive functions. Pupillometry is an easy-to-measure method that is increasingly used in clinical investigations to assess these functions. In addition to global luminance and arousal, pupil size is also modulated by converging bottom-up sensory and top-down cognitive signals. Therefore, the disruptions in neural circuitry due to neurodegeneration should affect pupil control, and disease progression and level of cognitive dysfunction should produce impaired pupil responses. We examined pupil dynamics in patients diagnosed with early-stage PD during an interleaved pro-/anti-saccade task annually over 4 years. Patients were categorized as cognitive normal (CN), mild cognitive impairment (MCI), and dementia based on composite neuropsychological scores. Compared to controls, PD patients showed significant reduction in pupil dilation suggesting changes in the top-down preparation signals. Pupil measures differentiated between CN, MCI, and dementia PD patients, and showed changes with disease progression. Our findings demonstrated links between pupil dynamics and PD neurodegeneration, showing that pupil measurements in visuomotor tasks have the potential to provide relevant early behavioural biomarker for diagnosis and tracking disease progression.

ABSTRACT. The current study aimed to investigate attentional processing differences for circumscribed interest (CI) and non-CI objects in young Chinese children with autism spectrum condition (ASC) and typically developing (TD) controls. In Experiment 1, a visual preference task explored attentional allocation to cartoon CI and non-CI materials, between the two groups. We found that ASC children exhibited a preference for CI-related objects compared to non-CI objects, and this effect was absent in the TD children. Experiment 2 utilised the traditional gap-overlap paradigm (GOP) to investigate attentional disengagement from CI or non-CI items in both groups and showed a similar disengagement performance for all children. Experiment 3 adopted a modified GOP (MGOP) to further explore disengagement in the two groups and results suggested that children with ASC exhibited ‘sticky attention’ compared to their TD peers, and that endogenous disengagement was influenced further in the presence of CI-related objects in the ASC children. However, exogenous disengagement performance was preserved in the ASC group. The current results have implications for understanding how the nature of engagement and disengagement processes can contribute to differences in the development of core cognitive skills in young children with ASC.

ABSTRACT. Frontotemporal dementia (FTD) is a clinical, pathological and genetically heterogeneous neurodegenerative disorder. Disease pathology in familial FTD already emerges years before symptom onset, which is associated with subtle cognitive changes. For upcoming medication trials, sensitive biomarkers to assess disease stage and track progression are crucial. Cognitive measures could serve as biomarkers, however, standard neuropsychological assessment lacks sensitivity. We hypothesize that eye movement features could serve as a sensitive biomarker for FTD. Our primary aim is to identify which eye movement features can be used as a biomarker for FTD. We will assess which (structured and spontaneous) eye movement features (1) dissociate best between patients with FTD, Alzheimer’s disease, presymptomatic FTD mutation carriers, and healthy controls, and (2) are sensitive to detect longitudinal cognitive decline in presymptomatic C9orf72, GRN and MAPT mutation carriers. A range of eye-tracking tests that have been related to FTD and Alzheimer’s disease in previous research will be administered, including pro-/anti-saccade, oculomotor capture, smooth pursuit, self-paced eye movements, Brixton spatial anticipation, visual search (Trail Making Test), and free viewing. In addition to features derived from structured eye-tracking tests, such as the anti-saccade test, spontaneous eye movements are evaluated. Frequentists and machine learning techniques will be used.

ABSTRACT. This study investigated the effect of the personal interest level (four levels) of an object (stimulus image) on gaze bias. Pictures of clothes and accessories that can easily attract attention in daily life were used as visual stimuli, and eye movements during preference decision making were recorded. The gaze bias (the bias of gaze toward the selected stimulus) was calculated and compared for each interest. The results showed that reaction time was longest for interest level 4, and characteristics of the gaze bias for the same were different from those for the other interest levels (1-3). The higher reaction time may be due to the liking effect combined with the time taken for selective encoding. The gaze likelihood curve showed that the curve started earlier with an increase in the interest level, which may have been due to differences in the reaction time. In the single dwell trials (looking at each object at once only), the opposite gaze bias was observed in the first look for interest level 4. When the personal interest level of the target was particularly high, the eye movement was characterized differently from that of targets with low interest level.

ABSTRACT. Background: Previous eye-tracking research has demonstrated that high-calorie food cues capture visual attention, particularly in individuals with overweight and weight concerns. The present experiment investigated whether this attentional bias can be influenced by context variables (i.e., the presence of pleasant low-calorie cues), and changes over time. This was studied via time-course analysis. Method: Ninety women (mean age = 25 years) were presented with 60 picture pairs (à 6 s) that either showed a combination of high-calorie food + low-calorie food, high-calorie food + non-food, or low-calorie food + non-food. The women were assigned to one of three groups (n = 30): overweight with weight concerns (OW+), normal-weight with weight concerns (NW+), and normal weight without weight concerns (NW-). The relative number of fixations on (high-calorie) food cues was examined across the course of the picture pair presentations for one-second intervals. Results: High-calorie food was fixated more often than low-calorie food and non-food only during the first second of the picture pair presentation. Subsequently, all participants (independent of the group assignment) showed the tendency to gaze at low-calorie cues when these cues were combined with high-calorie cues. Conclusion: This study revealed timing-related context effects on visual food cue reactivity. While the early attentional bias to high-calorie cues could not be changed by context, this was possible during the later processing stage. Future studies need to examine whether the observed context effect can influence food choices and eating behavior.

ABSTRACT. Research on multimedia learning has shown that people learn better from words and pictures than from words alone (Multimedia Principle, Mayer, 2021). However, it is less clear which effects pictures have when they are added to step-by-step procedural instructions which users typically read not to learn, but to act on the information they receive. In order to test effects of pictures on how users work through procedural instructions and interleave reading and acting, we created a simple software tutorial describing eight sequences consisting of 2-5 tasks to perform with the GIMP image manipulation program. There were two versions of the tutorial, differing with respect to whether they included pictures or not. Eye movements were recorded while subjects (N=42) worked through a tutorial and executed the respective tasks. Results show that subjects solved tasks more accurately and faster when procedures contained pictures. Eye tracking data revealed that time to process the instructions and transitions between the tutorial and GIMP did not differ between conditions. However, pictures supported subjects in applying procedural information to the GIMP software and encouraged them to adopt an atomized instead of a delayed task execution strategy (Jannin et al., 2019) which we argue improved overall performance as well.

ABSTRACT. Eye behaviour differs between internally (IDC) and externally direct cognition (EDC) and thus is indicative of an internal versus external attention focus. Recent work proposed perceptual decoupling (i.e., eye behaviour becoming less determined by the sensory environment) as one of the key mechanisms involved in the observed eye movement differences. To investigate the effects of perceptual decoupling on the voluntary and reflexive eye behaviour during EDC and IDC we used a dual task approach, combining an external eye movement task with internal visuospatial and arithmetic tasks. Internal tasks were further manipulated by three workload conditions. We expected the effects to be moderated by task type (more pronounced for visuospatial internal activity), workload (higher for higher internal demands) and follow a characteristic time course relative to internal operations. Furthermore, we hypothesised that perceptual decoupling will be more pronounced in voluntary eye movements but to some degree still observable in the reflexive eye responses. Data from 50 participants was acquired and currently are being analysed with linear mixed models. The findings will and shed light on relevant conditions of perceptual decoupling.

ABSTRACT. To date, eye movement research has extensively utilised game environments to study the link between cognition and gaze behaviour in strategic settings. We demonstrate a machine learning (ML) classification method that accurately identifies participants’ cognitive strategy before they commit to an action while playing matrix games that are commonly found in the economic literature. By applying ML models trained on scanpaths derived from complete sequences of eye movement data, we are able to classify scanpath subsequences generated from a hold-out subset of the participants. Our proposed ML approach outperforms traditional methods of analysis applied to the same economic game environments. This research aims to illustrate how eye-tracking data combined with ML can bring about an information asymmetry in favour of those who collect and process the data. We highlight how eye-tracking technology may be used to produce undesirable outcomes for individuals operating in strategic settings in the not-too-distant future.

ABSTRACT. When individuals perform Mental Rotation tasks, individual differences emerge in solution strategies that could be described by the response patterns used to cope with the task. The purpose of this study was to manipulate instructions encouraging solving a Mental Rotation Task (MRT) with a holistic versus a piecemeal strategy. It is expected that both strategies could be elicited and measured through eye movements tracked when participants perform the task. Sex differences in response patterns and performance were also analyzed. 140 MR trials were presented using the Eyelink Portable Duo 2000 Hz. Participants completed two phases during the experiment. Each phase comprised 10 practice trials plus 60 experimental ones (30 same and 30 different). Additionally, five angular disparities were configured (0-45-90-135-180º). During the first phase, participants solved the task in their own style. During the second phase, half of the participants were encouraged to solve the task using the holistic strategy, while the other half were induced to apply a piecemeal strategy. Data analysis combined different eye movement parameters for indicating two distinct eye-patterns—fixating and switching patterns. Results were discussed highlighting the role that employed strategy plays when interpreting sex differences in Mental Rotation tasks.

ABSTRACT. In aesthetics research for visual art, a central question concerns the relationship between the complexity of the artwork and the appreciation of the viewer. It has been hypothesized that this relation follows a Goldilocks principle, meaning that the highest appreciation is expected somewhere in between a very low and very high complexity of the scene. (Berlyne, 1971; Cutting, 2020). Much of this aesthetics research is based on self-reports of the viewer, mostly single questions about liking and appreciation. Recent studies show that the concept of aesthetic appreciation is multi-dimensional, including both affective and cognitive involvement (Hager et al., 2012; Schindler et al., 2017). In this study, we relate the multi-dimensional self-report measures of aesthetic experience to measures derived from eye-tracking (Nayak & Karmakar, 2019) while free viewing an artwork. To this end, we conduct an eye-tracking study of adults looking at selected artworks from the database of paintings rated for complexity (The Vienna Art Picture System, VAPS). The aim of this study is to test whether the Goldilocks principle as found with self-report measures, is also present with eye-tracking measures as a function of perceived and calculated visual complexity.

ABSTRACT. Basic written addition and subtraction are the building blocks for more complex calculations but little is known about the cognitive processes involved. The current study examined eye movements during the processing of correct and incorrect arithmetic computations. Participants (n=24) were presented with three-digit addition and subtraction sums in a formal vertical presentation on a computer screen. They then had to make judgements as to whether the answer was correct or incorrect whilst their eye movements were recorded using an EyeLink 1000. Initial results show that eye movements are concentrated on the digits forming the second line, the central digits, for both addition and subtraction for correct sums. However, distinctly different eye movements were found for the error conditions, as the fixation count was greater on the central digit of the column that contained the error. Overall, the operator received fewer fixations than the digits indicating that, as it is necessary for the successful completion of the sum, it was being processed parafoveally without direct fixation. These eye movement data indicate that individuals follow distinct step-by-step cognitive processes when solving sums. Implications of the current results for understanding the time course of mathematical processing will be discussed.

ABSTRACT. Negators (e.g. “not”) often appear in health statements (e.g. “Do not take more than one dose”). Previous literature shows that sentences containing negators have longer sentence reading times, lower comprehension accuracy, and are less likely to be remembered accurately, relative to sentences without negators. Given their short length, negators are likely to have low fixation rates. However, no studies have examined whether skipping of negators may contribute to difficulties in processing, understanding, and remembering sentences containing negators. Furthermore, no studies have examined whether making negators visually more salient, to increase the likelihood of fixation, will reduce such difficulties. Accordingly, in this study, participants will read fictional health messages without negation, with salient negation (negator bold and underlined), or with standard negation (negator formatted identically to surrounding text). Eye movements for the negator and the surrounding region will be examined, as well as sentence reading times, and immediate and delayed comprehension accuracy. The findings will reveal whether eye movement behaviour for negators is linked with difficulties in processing, understanding, and remembering health messages containing negators, and whether this is mitigated by increasing negator salience. Such findings will have important practical implications for the wording and visual presentation of health messages.

ABSTRACT. Previous research showed that in case of interruption during reading participants re-read the part of the paragraph preceding the interruption when they are allowed to do so (Cane et al., 2012; Chevet et al., 2021) This behavior could prevent text comprehension from being degraded. We believe this is all the more true when the participant has no prior knowledge on the thematic of the interrupted text. 40 participants either with a Bachelor of psychology or physics took part in the experiment. They had to read 4 texts, 2 texts related to psychology and 2 texts related to physics. For each thematic, one of the texts was interrupted by a short distracting paragraph. Eye-movements were recorded during the full reading session. After each text, the comprehension was evaluated by means of questions of surface, targeting a proposition explicitly mentioned in the paragraph at the location of the interruption, and also by means of text-based inferential questions, requiring the subject to make the link between information of the paragraph located before and after the interruptions. We are expecting that participants who re-read the less the part preceding the interruption will have their performances in comprehension degraded especially for the text for which they have no prior knowledge.

ABSTRACT. Progressive power lenses (PPL) are the most popular solution for presbyopes, who cannot focus clearly at near-vision due to aging. PPLs have a continuous and smooth increase in spherical power from the upper to the bottom region. This geometrical change induces unwanted peripheral astigmatism, which negatively affects vision and therefore reading speed (RS). The goal of this study is to evaluate how RS is affected by 3 PPLs (PPL-D optimized for distance-vision, PPL-N optimized for near-vision, and PPL-B a balanced design). Pupil position was recorded using a wearable eye-tracker system (Tobii-Pro-glasses 3) on 21 PPL users while reading a text out-loud located at 5.25m and 0.37m. A randomized block design test was conducted to compare the effect of power distributions on RS. Statistical analysis was performed using Statgraphics Centurion XVI.II software. At distance-reading vision, PPL-D showed statistically significant higher RS (PPL-D: 160±30wpm, PPL-B: 150±30wpm, PPL-N: 150±30wpm, P=0.01) than PPL-B and PPL-N. At near-reading vision, PPL-N has statistically significantly higher RS (PPL-N: 170±30wpm, PPL-B: 150±40wpm, PPL-D: 140±30wpm, P<0.01) than PPL-B and PPL-D. In conclusion, the power distribution of PPLs significantly affects RS when doing vision-based tasks. To provide the best visual experience, PPL selection must be considered user needs.

ABSTRACT. Prior research has shown that removing inter-word spaces disrupts word recognition processes during reading in alphabetic languages. The present study examined whether this disruption to word recognition would negatively impact inferential processing. We examined eye movements while participants read two-sentence long passages that were either strongly constrained (e.g., “Amy Parked her car on a steep hill. She watched as her car slowly rolled down the street.”) or weakly constrained towards an inference (e.g., “Amy parked her car in her driveway. She watched as her car slowly rolled down the street.”). We also manipulated text segmentation by presenting the passages with either normal inter-word spacing, or with the spaces replaced by numbers to produce an unsegmented text condition (e.g., “Amy7parked3her2car8in5her9driveway”). Consistent with prior work, our preliminary findings showed significantly longer fixation durations, higher numbers of fixations, and shorter saccade amplitudes in the unsegmented compared to the normal spacing condition. Data collection is ongoing, and as a next step we plan to use eye tracking to test our hypothesis that removing spaces will be more disruptive to processing in the weakly constrained compared to the strongly constrained condition. Our future findings will inform how lower-level processing interacts with higher-level processing during reading.

ABSTRACT. Previous evidence demonstrates that Chinese three-character idioms with a 2-character modifier and 1-character noun structure (2+1 MN) are processed as Multi-Constituent Units (MCUs, Zang et al., 2021). However, it is unclear whether idioms with a 1-character modifier and 2-character noun structure (1+2 MN) are processed as MCUs, as there are less constraints from the first constituent over the second compared with 2+1 idioms. In Experiment 1, we monitored participants’ eye movements when they read sentences with embedded idioms. Idioms were either presented normally (without segmentation highlighting), or highlighted with segmentations for whole MCUs, or constituent parts that did or did not coincide with the structure. Reading times were shortest for whole unit segmentations and longer for non-structural segmentations. Experiment 2, used the boundary paradigm (Rayner, 1975) to orthogonally manipulate preview of the modifier and noun of 1+2 idioms (identity or pseudocharacter). A greater preview of the noun occurred on the modifier when the modifier was an identity than pseudocharacter preview. Overall, these results suggest that Chinese idioms with a 1+2 MN structure are processed as MCUs, both foveally and parafoveally.

ABSTRACT. Sight-reading of music is a task in which the musician must perform a score during the first reading or with very little preparation (Wolf, 1976). It is a widely studied task in the music reading literature as it generates issues that involve a high level of cognitive processing. The challenge facing the musician during sight-reading is to perform despite a conflict between, on the one hand, the need for time to decode the visual information and prepare a motor response, and on the other hand, a limited time determined by the rhythmic constraints to perform the score.

Eye movements variables are known to discriminate individuals according to their ability and difficulty to process visual information. Therefore, studying musicians' eye movements during a sight-reading task enables to discriminate musicians according to their level of expertise and to evaluate their ability to cope with the complexity generated under the temporal constraint linked to the production.

In this experiment, 68 musicians from 5 conservatory levels had to sight-read 34 excerpts of different level of complexity while their eye movements were recorded. Musical score consisted of both classical tonal of Western music and contemporary repertoire and eye-movements data were recorded using the SR Research Eyelink 2000Hz system.

The main results show that the study of eye movements makes it possible to discriminate the level of expertise of musicians, even those whose conservatory level is close. Furthermore, this experiment shows that complexity is a significant factor in the evolution of eye movements and should be taken further into account to explain the differences in eye movements between performers.

ABSTRACT. Eye movement studies show that auditory distraction from irrelevant but meaningful background speech can disrupt reading by increasing regression rates and overall reading times. Little research has examined how this effect changes with age, however, although older adults might experience greater disruption due to poorer attentional control in older age. Accordingly, we recorded the eye movements of young (18-30 years) and older (65+ years) adults who read sentences in counterbalanced blocks of silence, irrelevant but meaningful background speech, and meaningless background speech. In addition, to test the hypothesis that auditory distraction might disrupt reading by interfering with the semantic integration of words, participants were presented with sentences containing either a highly predictable or a less predictable word. Consistent with previous research, the meaningful background speech was most disruptive. However, while typical patterns of adult age differences in eye movement behaviour were observed, the older adults were not more affected by the presence of background speech. Finally, while the meaningful background speech produced increased effects of word predictability, consistent with this interfering with semantic integration, this effect did not differ across the two age groups. We consider these effects in relation to accounts of auditory distraction on eye movements in reading.

ABSTRACT. Reading behaviour in all studied written languages shows robust effects of “lexical benchmark” predictors: longer, less frequent and less predictable words take more effort to process. Yet little comparative evidence exists about cross-linguistic variability of these effects. This study compares lexical benchmark effects on eye-movements during text reading in 12 alphabetic and abjad (Hebrew) languages. We used eye-movement data from the Multilingual Eye-Movement COrpus (MECO L1), collected from 543 participants in 12 countries reading short expository texts in their L1. Word frequency and predictability (a computational measure of word’s similarity to previous context) were computed using fastText vector models pre-trained on Wikipedia corpora in respective languages. Linear-mixed models demonstrated a very high degree of cross-linguistic similarity between benchmark effect sizes across eye-movement measures. Despite cross-linguistic differences in respective distributions, word frequency, length and word predictability explained a similar amount of variance in all languages. The effects were also similar in whether they preferentially affected early (word frequency), late (word length) or cumulative (word predictability) oculomotor measures. Moreover, many observed instances of cross-linguistic variability – both in word length and frequency distributions and in behavioural effects sizes – can be traced back to morphological characteristics and script type of the language in question.

ABSTRACT. Finding predictors of news misinformation detection is one of the most relevant tasks in text processing in social media research. In this pilot study, we use the eye-tracker approach to check whether the news source visibility and source type affects the accuracy of fake news recognition and news credibility; we adopt a 10% Type-I error threshold. Russian native-speaking participants (N=9) were offered 17 political news messages via a unique web interface thus producing 153 observations. Each news item contained a headline, the news source type, the date of publication, and the news text. Participants were asked to evaluate each news item’s credibility on a six-point Likert scale ranging from “true” to “fake”. Psychometric data of the participants were also collected (e.g.,CRT). Accuracy (correctness of news recognition understood as the hit rate) and news credibility (measured on a six-point scale) were calculated as dependent variables. The source visibility (was it noticed or unnoticed by the participant) and source type (from domestic - i.e. Russian - media or from foreign media) variables were used as independent factors. Predictably, no statistically significant main effects or interactions were found for the accuracy. Therefore, we have no evidence that the source type or its visibility to the participant affect the accuracy of news categorization. However, the visibility of the source and its type affect the participants’ perception of message credibility. The analysis showed the main effect of the source visibility (t =-2.070, p<.05) and an interaction between source visibility and source type (t=1.941, p<.10). This means that foreign sources are significantly less trusted than domestic sources, but this effect only exists when the participants have actually seen the source.

ABSTRACT. Efficient cognitive functioning requires the ability to ignore task-irrelevant stimuli. However, it is important to detect unexpected changes in the environment, which could signal danger. It is well-established that the brain continuously monitors the auditory environment and can detect subtle changes to a repeated sound. Recently, it has been shown that the movement of the eyes' during reading is inhibited by unexpected sounds, referred to as novelty distraction. The current experiment investigated if novelty distraction is independent of higher-level linguistic processing by comparing a reading to a letter scanning task. Seventy-two Bournemouth University students either read sentences or scanned letter strings of the sentences for the letter ‘o’ with three auditory conditions: silence, standard, and novel. The results showed that first fixation durations were slightly longer in the standard sound condition compared to silence. Additionally, novel sounds led to a significant increase in first fixation durations compared to standard sounds (novelty distraction). Critically, the novelty distraction effect was identical between the two tasks, suggesting that it is independent of the type of task participants were doing. Because the letter scanning task eliminated higher-level language processing, the results suggest that novel sounds generally inhibit ongoing oculomotor processes independent of linguistic processing.

ABSTRACT. Unlike English, Chinese is a character-based, unspaced language. An intriguing question concerns how lexical processing in Chinese is operationalized across potentially ambiguous, multiple adjacent character sequences. We were particularly interested in processing of frequently occurring Multi-Constituent Units (MCUs), that is, linguistic units comprised of more than a single word, that might be represented lexically. Using the boundary paradigm (Rayner, 1975), we conducted six experiments and manipulated the preview of each constituent of target strings, including three-character famous people’s names (consisting of a one-character family name followed by a two-character given name, Experiment 1), place names (consisting of a two-character modifier and a one-character noun, Experiment 2), product names and modern phonetic borrowings (these structures are not transparent and thus constituents were defined and assessed comprehensively in Experiments 3a & b, and Experiments 4a & b). The results from all six experiments showed very consistent patterns with more pronounced preview effect from the second constituent when the first constituent was present in the parafovea, very strongly suggesting that these frequently occurring MCUs are lexicalized and processed parafoveally as single units during Chinese reading. The results provide compelling evidence in support of the MCU hypothesis.

ABSTRACT. popEye is an R package to analyse eye movement data from reading experiments (Schroeder, ECEM 2019; freely available at https://github.com/sascha2schroeder/popEye). A unique feature of popEye is that it allows semi-automatic analysis of text-level experiments. To this end, popEye implements a wide selection of state-of-the-art line assignment algorithms and sentence-level measures. In addition, popEye has been designed to be used in cross-linguistic studies and thus supports a wide variety of languages including Hebrew, Korean, and Chinese.

In this poster presentation, I will present some of the new features that have been implemented in popEye since its first release, including new options for data pre-processing, line-assignment, and automated cleaning that make the analysis process more flexible and transparent. I will also provide links to new tutorials that demonstrate how popEye is used and walk user through an example analysis. The poster presentation is intended to be open and interactive and is supposed to give (potential) users the opportunity to ask questions about popEye and discuss new feature requests.

ABSTRACT. Based on a 3D object’s geometry, some object viewpoints are easier to discriminate from their neighboring viewpoints than others. Humans use saccades to sample information in the world, but it is unknown whether object properties such as viewpoint discriminability can influence saccade target selection. This study tested whether participants can infer the discriminability of an object viewpoint, and how they use this inference when choosing saccade targets in a subsequent match-to-sample task.

In a saccade target selection experiment, subjects fixated a central cross and were presented with two different viewpoints of a 3D-rendered object on opposite sides of fixation. One viewpoint was always more discriminable (easier to match) than the other. Participants were instructed to saccade to one of the two objects. In one condition, subjects matched the object-viewpoint they saccaded to, by rotating an onscreen object. In a second condition they were asked to match either the saccaded or non-saccaded object-viewpoint.

Results show that participants changed saccade target selection depending on the report condition, swapping whether they fixated the most or least discriminable viewpoint between conditions. This suggests that complex inferences about 3D objects may be used to inform saccade target selection and information sampling to guide subsequent behaviour.

ABSTRACT. The sensation of having control over one's voluntary actions and their outcomes is referred to as sense of agency. We examined sense of agency for saccades with implicit and explicit measurements. Participants fixated different stimuli on a screen while eye movements were recorded. After a short delay following stimulus fixation the color of the stimulus changed. Participants then reproduced the interval between saccade-landing and color-change. We made participants believe that they had control over the color change in one condition and no control over it in the other condition. We analyzed differences in interval reproductions between both conditions as a correlate of temporal binding, which might be linked to an implicit sense of agency. Further, we collected explicit agency ratings from participants. We found no differences of temporal binding between the two conditions but higher agency ratings for conditions in which participants felt they caused the color change. These findings indicate that temporal binding might better be explained as a phenomenon of multisensory integration in our study. However, oculomotor actions may generate an explicit sense of agency comparable to manual actions. This has several implications for the application of eye-tracking technology and the design of eye-tracking interfaces.

ABSTRACT. Ballistic eye movement, saccade, disrupts temporal order judgment, which has been well investigated within a single modality such as vision while not being well investigated between one’s saccade and visual stimuli onset. Here we investigated the discrimination threshold of temporal order judgment before and after the saccade to see if the judgment can be illusory reversed. Subjects initiated saccade toward a target dot, while a central dot disappeared 200 ms before or after, or simultaneously with target dot onset(central dot offset condition). Two horizontal white bars at the bottom and top were displayed simultaneously, with onset asynchrony from 0 to 675 ms. Subjects had to answer whether the saccade or the bars onset occurred first. Psychometric functions on the proportion of “saccade” response against the relative temporal distance between the saccade and bars onset showed that Sigmoid-like function, suggesting the correct judgment when the saccade initiated from -500 to 100 ms relative to the bars onset, irrelevant to the central dot offset condition. Interestingly, proportion of “saccade” response increased when saccade was initiated 300 ms after the bars onset, which might suggest postdiction-like effect. Our results might be different phenomenon from what has been shown in the single modality study.

ABSTRACT. Ocular movements could help to understand better how visual performance is affected by defocus during Visual Acuity (VA) measurement. For that purpose, a clinical study was performed in 6 non-emmetropic subjects (age 34±8 years old). VA logMAR was measured at 2 distances (FV, 4m; NV, 37cm) under the induction of different amounts of defocus (±1, ±0.5, ±0.25, and 0D) on both eyes while an eye-tracker (Tobii Pro Glasses 3, Sweden) recorded pupil position. The following variables were analyzed using Statgraphics (Statgraphics Centurion XVII, v17.2.07): Reading Time (RT), Total Duration of Fixations (TDF), and Fixation Count (FC). At FV, positive defocuses present worse values on the variables, since they cannot be accommodated (p-value<0.05,RT&TD@1D). At NV, non-statistically significant differences were found on positive defocuses with 0D, since it helps to accommodation. The most demanding conditions ( 0.2VA and ±1D) were analyzed according to age. At FV, negative defocus requires extra accommodation, and the task is more difficult for elder subjects (R2=0.68&0.88 for RT&FC). At NV, positive defocus helps to read to elder people (R2=0.52 for FC). Eye-tracker is a useful tool to study the influence of defocus induction facing VA tasks. It was found a dependence on age in the most demanding conditions.

ABSTRACT. We tested the time course of saccadic inhibition of return and its relation with saccadic momentum. To this end, we extended a saccade sequencing paradigm, originally developed by Ludwig et al. (2009). Participants were required to make eight saccades in a display that consisted of 16 circles as possible fixation locations. The gaze of the participants was guided either by a saccade-contingent endogenous arrow cue at fixation pointing to the next fixation location (Experiment 1), or by an exogenous cue appearing at the next fixation location (Experiment 2). The direction of the subsequent saccade could deviate from the preceding saccade direction by four different angles. After the sixth fixation in the sequence, the gaze was either directed back to a location fixated up to five fixations earlier (fixation lag) or directed to a new location. We found longer saccade latencies for fixation lags of up to three fixations back compared to fixations of new locations for endogenous cues, indicating the presence of inhibition of return. However, forward saccades were consistently initiated faster than those with greater angular deviations, indicating saccadic momentum. Nonetheless, there seemed to be an effect of inhibition of return in addition to saccadic momentum.

ABSTRACT. Saccade adaptation corrects motor errors to maintain saccade accuracy. A recent study demonstrated that saccade adaptation is not only driven by accuracy feedback shortly after saccade offset, but also by temporarily delayed feedback signals if they are task-relevant (Wagner et al., 2021). Here we studied how delayed feedback from relevant stimuli interacts with conflicting feedback signals from temporally proximate, but irrelevant stimuli.

We instructed participants to make a saccade towards an unfilled circle. Inside the circle, two discrimination stimuli were shown successively at opposite locations. Participants were instructed to discriminate the location of a gap, either on the first (discriminate-first-condition) or the second stimulus (discriminate-second-condition).

We observed saccade adaptation in the discriminate-first-condition, but not the discriminate-second-condition. A second experiment showed that this was modulated by saccade target properties: using a small square, located in-between discrimination stimuli, instead of the circle, we observed saccade adaptation to the first stimulus in the discriminate-second-condition. The large saccade target in the first experiment might allow for a larger presaccadic attentional focus and lead to weaker object correspondence with the postsaccadic discrimination stimuli. Both factors might contribute to a greater flexibility in which feedback signals are selected and which are suppressed for saccade adaptation.

Acknowledgements: This project was funded by the SFB/TRR 135 and the International Research Training Group, IRTG 1901, “The Brain in Action”, from the German Research Foundation (DFG)

ABSTRACT. Kirsten Williams, Paul McGraw, Timothy Ledgeway and Denis Schluppeck Visual Neuroscience Group, School of Psychology, University of Nottingham, UK.

The visual system uses optokinesis to help stabilise the retinal image. Optokinetic nystagmus (OKN) displays a sawtooth time-course and results in after-nystagmus (OKAN) when all lights are extinguished during optokinetic stimulation. Here we compared OKN and OKAN in both cardinal and oblique directions for a range of stimulus velocities. We varied the instructions given to participants to elicit look OKN (actively following features) or stare OKN (passively viewing and keeping in focus) and examined their relationship with OKAN. Human participants observed a sinusoidal grating (0.5 c/deg at 10, 20 or 40 deg/s) drifting in one of eight directions, equally spaced in polar coordinates. Eye position was measured using the Eyelink 1000 eye tracker. At 10 deg/s, OKN gain was similar during both look and stare conditions (~1.0) and was largely independent of motion direction. At 20 deg/s, stare OKN gain was reduced relative to look OKN and both were attenuated for vertical motion. At 40 deg/s, gain was further reduced, and gain differences between look and stare OKN disappeared. OKAN was more reliably produced during stare conditions. Contrary to previous reports, OKAN was evident for vertical stimulation, but the direction was reversed with respect to stimulus motion.

ABSTRACT. Fluctuations in a person’s arousal accompany mental states such as drowsiness, mental effort, or motivation, and have a profound effect on task performance. Here, we investigated the link between two central instances affected by arousal levels, heart rate and eye movements. In contrast to heart rate, eye movements can be inferred remotely and unobtrusively, and there is evidence that oculomotor metrics (i.e., fixations and saccades) are indicators for aspects of arousal going hand in hand with changes in mental effort, motivation, or task type. Gaze data and heart rate of 14 participants during film viewing (https://studyforrest.org; Hanke et al., 2016) were used in Random Forest models, the results of which show that specifically the movement aspect of oculomotor metrics (i.e., velocities and amplitudes) links to heart rate – more so than the amount or duration of those features. We discuss that eye movements are not only linked to heart rate, but they may both be similarly influenced by the common underlying arousal system. These findings provide new pathways for the remote measurement of arousal, and its link to psychophysiological features.

ABSTRACT. Cyclovergence (rotations of the eyes about their visual axes in the opposite directions) normally occurs to correct torsional misalignments of the eyes produced during eye movements (e.g., gaze elevation). Previous studies have shown that cyclovergence can be elicited visually by vertical shear disparity (equal and opposite vertical shear between the two eyes' images). Vertical shear disparity provides a signal to control cyclovergence. However, studies have only used stimuli in the fixation plane, so it is not known whether vertical shear disparities are effective in surfaces at a different depth from fixation. Here we investigated whether cyclovergence is elicited by vertical shear disparities in surfaces at different depths from fixation. Stimuli were stereoscopic surfaces (approx. 70deg.) comprising of arrays of texture elements. The surfaces contained vertical shear disparities (+/-1.5deg.) and were presented at a range of depths (horizontal disparities, +/-40 arcmin) with respect to fixation. Cyclovergence was measured using a nonius technique. We also examined the perceived inclination in depth of the stimuli, since a frontal surface appears inclined when vertical shear disparities are introduced. We found that vertical shear disparities in surfaces away from fixation are effective in evoking cyclovergence and inducing apparent slant.

ABSTRACT. Asthenopic symptoms are related to heterophoria and problems of binocular vision. In a recent paper, we showed that vergence drift and fixation durations are related to symptoms (CISS-questionnaire), but optometric measures such as heterophoria, vergence or accommodative facility, AC/A-ratio or NPC did not significantly add to the explained variance of asthenopia. We re-analysed our data, in which binocular eye movements were recorded (EyeLink II) for 64 participants, and linear regression analyses related all parameters of binocular coordination (objective heterophoria, vergence drift, saccade disconjugacy, fixation disparity and fixation duration), and the above-mentioned optometric tests to 4 symptoms factors (eye comfort, reading process, image quality and fatigue), which we identified by a factor analysis of the CISS-questionnaire. Objective heterophoria and fixation duration predicted 20% of the variance in symptoms concerning the reading process (factor 2). Furthermore, fixation duration seems to be slightly, but not significantly, related to symptoms addressing fatigue (factor 4), whereas optometric tests are not significantly related to any asthenopia factor. Overall, objective measures during reading relate to asthenopic symptoms, however the lack of asthenopic symptoms prediction by daily optometric parameters is still unexplained. Therefore, further research is needed to find out suited optometric parameters to infer asthenopic symptoms.

ABSTRACT. Saccades are often regarded as mere by-products of input-related cognitive processes, not as controlled actions that potentially interfere with other actions. However, robust performance costs associated with executing a saccade in conjunction with another concurrent action have repeatedly been demonstrated. These costs are typically smaller for saccades than for actions of other effectors given the high prepotency of the former and the resulting preferential execution. Here, we utilised these features of saccades (high inherent prepotency while still requiring active control) to study how inhibition of a prepotent action (while initiating another action) can prove more difficult than the concurrent execution of both actions, thereby yielding relative dual-action performance benefits. Participants responded to a single peripheral visual target with either a dual action consisting of a saccade and a manual button press, a single saccade, or a single button press. The latter response demand required saccade inhibition. The results revealed high rates of inhibition failures (false-positive saccade executions) in single manual trials. While similar effects were also demonstrated with other action modality pairings (e.g., manual-vocal), saccades can be considered ideal for eliciting such dual-action action benefits and thus for studying inhibitory control in general.

ABSTRACT. ilateral eye movements are the process of shifting your gaze from left to right in quick succession. Bilateral eye movements can lead to enhancements on subsequent memory and recall tasks. Previous research has prominently assessed healthy younger adult populations and has failed to assess the enhancement effect in clinical populations. Bilateral eye movements could offer therapeutic benefits to older adults with age-related cognitive decline and individuals with cognitive impairment. Across two experiments we assessed the effect of bilateral eye movements on memory and recall task performance. Experiment 1 aimed to replicate the effect in younger and older adults and assess whether the enhancement effect was exclusive to bilateral eye movements or elicited via top-down control tasks such as anti-saccade tasks. Participants were presented with 36 words to recall in a subsequent task and randomly allocated to an eye movement condition: bilateral, anti-saccade or fixed. Experiment 2 employed an online study methodology assessing the effect in individuals diagnosed with mild cognitive impairment, Alzheimer’s and Parkinson’s. Both experiments failed to replicate the enhancement effect across all participant groups with no significant differences in the number of words recalled based on eye movement condition. The results do not provide support for the use of bilateral eye movements to provide therapeutic benefits in clinical populations. This study adds to the growing number of failures to replicate and indicates an unstable effect lacking robustness. Future research should aim to establish clear and precise boundary conditions in which the effect is present, robust, and replicable.

ABSTRACT. This study deals with the processing of verbal and non-verbal information in Russian multimodal advertising posters. The effect of lexical ambiguity in a slogan is checked out. 36 Russian participants (female = 28, mean age = 25) read 14 advertisements with lexically ambiguous slogans (the first meaning is supported by the picture and the second one – by the text) and 14 unambiguous advertisements. All the posters were edited and equalized by the size, font, background colour, the layout of the elements and the number of letters in the text. We used EyeLink 1000 Plus (SR Research) apparatus, monocular reading, head fixed. The results show that total dwell time does not differ for ambiguous and unambiguous posters. Readers made almost the same number of fixations: Me = 32,39 for ambiguous posters and Me = 32,73 for unambiguous ones. Surprisingly, the first fixation was made on the verbal part of the poster significantly more often than on the picture (761 VS 247). Also, unambiguous slogans were read significantly slower (more fixations and more regressions) than ambiguous slogans. These findings contradict the previous results and assumptions that ambiguity in a text slows down the reading process, since the reader needs time to resolve it [Levelt, 1978; Sereno, O'Donnell, Rayner, 2006].

ABSTRACT. This study presents a cross-sectional comparison of eye movements during silent reading in Russian schoolchildren. We aimed to establish (1) the development of basic eye-movement characteristics with age, (2) the influence of psycholinguistic word properties (length, frequency, predictability) on eye movements at different ages, and (3) the association between phonological processing and reading. We tested 222 children from grades 1-6 (age range 6-13; 105 girls). All children’s non-verbal IQ (Raven, 2004), oral reading fluency, and comprehension (Kornev, 1997) were within the normative range. The children read 30 sentences (6-9-word long) adapted from the Russian Sentence Corpus (Laurinavichyute et al., 2019) while their eyes were tracked with Eyelink PortableDuo or 1000+. They also completed a complex phonological task ‘Changing the Sound in a Pseudoword’ (Dorofeeva et al., 2020). The results indicate that progress in reading manifests in eye-movement characteristics from grades 1 through 4 and then slows down, although it does not reach the adults’ level in grade 6 (Laurinavichyute et al., 2019). The effects of word length, frequency, and predictability become smaller with age, with a significant reduction between grades 1-2 and between grades 3-4. Better phonological processing is associated with shorter fixation durations, especially in grades 1 and 2.

ABSTRACT. The growing interest to reader's comprehension estimation leads to searching for new methods enabling such estimation in real-time (or pseudo real-time). It can be used not only for more effective educational processes, but also in adaptation of text content for different purposes. In the study itself, Oken Reader application with eye-tracking tech-nology was used to assess perception and comprehension processes in reading. Twen-ty-three (23) respondents aged between 19 and 31 years (mean = 24.5, SE = 1.4, 65% women) took part in the study. As a result, fixation duration analysis on the base of proportion of correct answers to the item showed significant difference for mu, sigma, tau ex-Gaussian parameters and mean all between groups, where all answers were correct or false. For the medium groups demonstrated statistically significant difference only in tau parameter. Using the preliminary video supplement led to decreased num-ber of long fixations. The results of the study revealed the fundamental possibility of searching the correlations of physiological indicators such as eye movements with the process of comprehension the read text.

ABSTRACT. Despite extensive research on extrafoveal linguistic processing during reading, there is a surprising gap regarding the role of syntax. It is still unclear whether syntactic information in the form of grammatical gender can be acquired in the parafovea and used to facilitate the processing of subsequent words. The issue is related to the question of whether small word groups like minimal noun phrases can be the functional target in the programing of reading saccades. Two experiments used a gaze-contingent display change paradigm, manipulating syntactic fit within nominal phrases in German. Either the article (experiment 1) or the noun (experiment 2) was manipulated, resulting in a mismatch within the nominal phrase, depending on the condition (“den/die Fisch” vs. “die Fisch/Hecht”; the fish/pike). When the eyes crossed a boundary to the right of the preceding word, the mask was replaced to provide a syntactically fitting noun phrase. Results indicated a substantial elevation of viewing times on both parts of the noun phrase. Contrary to expectation, the effect was independent of noun frequency. This work provides direct evidence for parafoveal syntactic processing, suggesting that that grammatical gender is used to generate constraints on subsequent nouns. Consequences for eye movement control will be discussed.

ABSTRACT. Research using the masked priming same-different matching task in Arabic strongly suggests that allographic variation (changes in a letter’s shape as a function of its position in the word) can constrain lexical processing. Specifically, response times to target words (e.g., يسترون [ySTRw] “they cover”) were shorter when preceded by a transposed-letter (TL) nonword prime in which transposed root letters maintained their shape (e.g., يتسرون [yTSRwn]) than when they did not (e.g., يسرتون [ySRTwn]). The present study investigated whether allography modulates TL priming in more naturalistic viewing conditions: when people move their eyes while reading. Accordingly, we used the boundary-change technique in which parafoveal preview of the target word was (a) identical (e.g., أخلقها –أخلقها ‘I create her’), (b) a TL-nonword without allographic changes (TL-Allog: e.g., أخقلها -أخلقها), and (c) a TL-nonword with allographic changes (TL+Allog: e.g., ألخقها –أخلقها). Fixation times on targets were shorter only when the preview was identical. Thus, visual similarity alone does not modulate the integration of parafoveal and foveal information during sentence reading when the manipulated letters are root letters. These findings support the view that the order of the root letters in Semitic languages allows little or no perceptual noise in a normal reading scenario.

ABSTRACT. We examined the relationship between foveal and parafoveal word processing efficiency and word-level eye-movement parameters in reading among 74 adult readers (age range 17-66 years). Foveal processing efficiency was assessed with performance accuracy in the lexical decision and naming task, where 6-8–letter words and pseudowords were presented backward-masked for identification using different exposure times (20-60ms). Parafoveal processing efficiency was assessed by presenting 5-letter words and pseudowords for 150 ms to the right and left of the fixation point (1.33-2.24 degrees of visual angle). Mean overall performance measures of word naming and pseudoword naming were calculated for the naming tasks and a d' sensitivity index for the lexical decision tasks. Measures of foveal and parafoveal processing efficiency were used to predict eye-behaviour in reading 8 expository texts. Word-based LMM–analyses revealed that individual differences in foveal prossessing efficiency predict gaze durations especially in long and infrequent words. Foveal processing effiency also affected skipping probability, which was revealed as a main effect. Parafoveal processing efficiency predicts probabilities of skipping especially shorter words that are parafoveally previewed from shorter distances, while showing less relation to first-pass reading times. Even among adult readers word recognition skills predict word reading times and word skipping.