Visual Thresholds Cannot Be Reliably Measured Without Controlling for Mind-Wandering

Visual Thresholds Cannot Be Reliably Measured Without Controlling for Mind-Wandering | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Visual Thresholds Cannot Be Reliably Measured Without Controlling for Mind-Wandering Stefan Dürschmid, Annemarie Scholz, Christoph Reichert, Paul Schmid, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9412640/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Visual threshold estimation typically requires data from thousands of trials presented across extended periods of time. However, human attention fluctuates over time between focused and inattentive states, such as mind-wandering (MW). Failing to account for MW in perceptual research may produce systematically lower estimates of perceptual performance. We assessed how brief changes in brain state affect visual performance by investigating the impact of MW on visual target detection and discrimination in twelve behavioral tasks involving 239 subjects (156 female). In all tasks, subjects reported their task engagement (ON vs. OFF) in 20% of trials. We explored whether in the remaining 80% of unlabeled trials performance aligns with ON or OFF trials. MW led to threshold detection shifts and reduced perception in OFF compared to ON trials. MW primarily affected the perception of simple stimuli like gratings and curvature stimuli, with its influence diminishing as stimulus complexity increased. Importantly, visual performance in unlabeled trials aligned with the OFF-state. Our results show that mind wandering induces fluctuations in attentional capacity that render threshold measurements substantially less reliable without correction, underscoring the need to account for attentional state to obtain valid estimates of perceptual capacity. Biological sciences/Neuroscience/Visual system Biological sciences/Neuroscience/Cognitive neuroscience/Perception Biological sciences/Neuroscience/Cognitive neuroscience/Attention Biological sciences/Neuroscience/Sensory processing Biological sciences/Psychology/Human behaviour Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Significance Statement Perceptual thresholds are widely used to define the capacity of the human visual system, under the assumption that attention remains stable across trials. By combining experience sampling with behavioral performance across twelve tasks, we demonstrate that mind-wandering introduces measurable fluctuations in perceptual capacity. Our study shows that perceptual threshold estimation is compromised in inattentive states, particularly for simple stimuli. These findings highlight the necessity of accounting for attentional state in perceptual research to obtain valid measures of human vision. Introduction The visual perception threshold assesses the degree of efficient perception. Typically, the probability of correct performance is estimated across multiple experimental trials. We obtained a psychometric function that quantitatively characterizes the perceptual sensitivity of the visual system. Participants were instructed to maintain attention throughout the duration of the task, assuming full attentiveness across trials (ON state). However, in humans, monkeys and rodents brain states undergo fluctuations affecting visual perception. Mind wandering (MW; OFF state) is a cognitive phenomenon occurring up to 50% of our daily lives and is often described as attentional decoupling from the external world due to decreases in behavioral performance (Smallwood, 2011). Standard perceptual threshold testing is typically assumed to be dependent by ON-task behavior. However, this assumption has not yet been empirically tested. Given the high frequency of MW in everyday life, it is unrealistic to expect continuous full attention during visual detection and discrimination tasks. Consequently, threshold estimation in perceptual research may be biased by MW, resulting in a systematic underestimation of true visual performance under standard experimental conditions. This raises the critical question of whether threshold estimation is affected by MW. Feature-based attention improves perception of basic features accompanied by a shrinkage of receptive field size 2 – 5 . Neurons with larger receptive fields in the primate visual system have lower attentional demand and may exhibit a tolerance to variations in brain states like MW. To investigate the impact of MW on visual perception depending on the attended feature, we conducted a series of experiments using four types of visual stimuli – varying in stimulus complexity (orientation, curvature, color, and faces) - across three task conditions (detection, global discrimination, and local discrimination). Experience-sampling probes were interspersed throughout each experiment to dissociate perceptual performance across distinct brain states in a subset of trials (ON-task vs. OFF-task). The majority of trials without mind-wandering probes represented the ground truth against which ON- and OFF-task performance was assessed. Two critical questions arise. First, if attention enhances performance, then accuracy should be higher during ON-task relative to OFF-task trials. Second, if standard threshold estimation faithfully reflects the perceptual capacity of the visual system, then estimates derived from the ground truth should resemble ON-task performance. We hypothesized that visual threshold measurements for features processed early in the visual hierarchy primarily reflect performance under the influence of MW. Our findings demonstrate that averaging across all available trials—without accounting for attentional fluctuations—yields unreliable estimates of visual thresholds, with estimation errors increasing as stimulus complexity decreases. Materials and Methods Participants After obtaining written informed consent, we tested 239 subjects (156 female, mean age: 24.63y) across 12 experiments. All subjects were compensated with 8€/hour or course credit. The study was approved by the local ethics committee (“Ethical Committee of the Otto-von-Guericke University Magdeburg”). General Paradigm We investigated the impact of MW on visual perception of target stimuli defined by Orientation ( Paradigm 1 ), Curvature ( Paradigm 2 ), Color ( Paradigm 3 ), and viewing direction of Faces ( Paradigm 4 ). Each visual feature was tested in three different experiments defined by the instruction (Detection, Global, and Local Discrimination). To maintain consistency across all experiments and ensure comparability, 18 stimuli were presented in each trial, divided into two blocks of nine stimuli each (arranged in a 3 x 3 grid, see Fig. 1 B) positioned to the left and right of the fixation cross. The stimuli, as well as the experimental setup, were implemented using Matlab R2013a (Mathworks, Natick, USA) and the Psychophysics Toolbox extension 6 . All stimuli were presented on a gray background using a 14'' color monitor with a resolution of 1.920x1.200 pixels and a refresh rate of 60 Hz. Distance to the display was kept at 70 cm. Paradigm 1: Orientation Perception Participants were presented with simple grating stimuli (Gabor Patches; see Fig. 1 A), where the target stimulus differs in orientation from the surrounding distractor stimuli. Participants were required to either detect the presence of a target (Detection), determine its location (Global Discrimination), or identify the direction of its tilt (Local Discrimination). The distractors were uniformly vertically aligned, while the target's angle varied between 1° and 24.8° across different blocks, with 15 steps of 1.7°. Each stimulus consisted of four parallel black stripes on a white background and had a total size of 80 x 80 pixels. To minimize edge artifacts, a Gaussian filter with an 80 x 80 pixel size was applied to each grating. The visible portion of each Gabor patch occupied approximately 0.81° of visual angle. Paradigm 2: Curvature Perception Targets consisted of line stimuli with a stronger curvature compared to the distractors (see Fig. 1 A). Participants were required to indicate whether a target was presented (Detection), where the target was located (Global Discrimination), or in which direction it was curved (Local Discrimination). In order to generate the curved stimuli, black circles with varying diameters were created and then one half of each circle was covered by a rectangle matching the grey background. This way, circles with a smaller diameter resulted in stimuli with a stronger curvature. The diameters of target stimuli varied between 30 and 44 pixels in 15 equal steps, while the distractors had a constant diameter of 45 pixels. The resulting stimuli consisted of a black curved line taking up approximately 0.98° va. Paradigm 3: Color Perception Targets were displayed with one side appearing red and the other side appearing more pinkish, while the distractors were red on both sides (RGB: 152 0 15). For each block, one out of 15 color values was chosen for the target stimulus (see Fig. 1 A). Participants' tasks included identifying the presence of a target (Detection), the location of a target (Global Discrimination) or which side of the target was more pinkish (Local Discrimination). The targets and distractors were composed of two semicircles with a vertical gap measuring 0.1° va and a total height of 0.76° va. The gap was selected to eliminate discrepancies in spatial frequency caused by color gradients between the targets and distractors. To prevent edge artifacts, the semicircles were overlaid with a Gaussian filter (3 pixels). Paradigm 4: Face Perception Stimuli consisted of face pictures with a neutral facial expression and varying degrees of pixelation. Participants were required to either detect the target face among distractor stimuli (Detection), differentiate where it was presented (Global Discrimination) or decipher the gaze direction of the target face (Local Discrimination). To capture the face, we took a photograph of a mannequin's face from a 24.5° angle on the left side, resulting in the face looking towards the right. The image was then mirrored to make it appear as if the face was gazing at the left side of the screen. An oval was superimposed over the face to cover hair and ears. Finally, the pixels of targets were shuffled at varying degrees (30–86% of pixels, in 15 equal steps of 4 pixels), while the distractors had a higher degree of shuffling (90% of pixels; see Fig. 1 A). Each stimulus occupied around 1.3° va in height and 0.73° va in width. General Procedure At the start of each trial, a black fixation cross was displayed. Following a delay of 750 ms (± 250 ms), participants were presented with a visual search display below the fixation cross for 100 ms. This display consisted of 18 stimuli, with one target and 17 distractor stimuli, consistent across all experiments. In 50% of the trials, the target was positioned either to the left or right (as shown in Fig. 1 B) with equal likelihood of appearing at any of the 18 positions. The different experiments varied in their instructions. In the Detection experiments, the target was shown in 50% of pseudo-randomly selected trials. Participants were instructed to indicate whether a target was present (using the "F" key) or absent (using the "J" key) on a standard QWERTZ keyboard. In the Global and Local Discrimination experiments, the target was always presented, while its location and orientation were chosen pseudo-randomly (50% right, 50% left). Participants were asked to indicate whether the target appeared to the left (using the "F" key) or to the right (using the "J" key) of the fixation cross for Global Discrimination, or whether the left-ward oriented version (using the "F" key) or right-ward oriented version (using the "J" key) of the target was presented for Local Discrimination. Each experiment commenced with a practice block consisting of 22 trials for all participants. The orientation of the target was randomized across these practice trials. During the practice block, two focus queries were administered. The location and orientation of the target (Global / Local Discrimination) or whether a target was presented (Detection) was randomized within each trial of the practice block. In the subsequent experiment, participants were exposed to a total of 1620 trials (108 per block) divided into 15 blocks, with 330 focus questions (see Experience Sampling ) in total (22 per block). The order of the 15 different deviation strengths was randomly varied across the blocks. Within a given block, the deviance strength of the target from the distractors remained consistent. Between experimental blocks, participants took a break and could proceed to the next experimental block at their own pace. The entire experiment took about 1.5h for each subject. Experience Sampling Throughout the experiment, we incorporated thought probes in a pseudorandom manner, selecting them for 20% of all trials. These probes aimed to assess participants' level of MW in the trial immediately preceding the probe. Participants were asked to rate their MW experience on a five-point Likert-scale, ranging from 1 (indicating that their thoughts were elsewhere - OFF) to 5 (indicating that their thoughts were completely focused on the task - ON). Participants responded to the focus questions using the fingers of their left hand, assigning a rating using the following finger-to-rating mapping: thumb for 5, index finger for 4, middle finger for 3, ring finger for 2, and little finger for 1. The thought probes were presented after the visual search display. Furthermore, we ensured that there was a minimum of one intervening trial between two probes (however, this minimum distance was observed in only 7% of the focus queries). The probes were initiated by an auditory stimulus with a frequency of 500 Hz at approximately 85 dB for a duration of 200 ms. This experience sampling approach analyzing a subset of trials was employed since more frequent focus queries could interfere with MW 7 . This results in a limited number of trials that can be assigned with a certain focus rating. Consequently, a significant number of unlabeled trials, usually around 80%, are excluded or discarded (9). Data analyses First, we determined whether there were systematic differences in the distribution of brain state reports between the experiments and conditions ( I – Brain State Distribution ). In the next step, we assessed the reaction times across experiments and conditions (II – Reaction Times) . Then, we compared thresholds between the three different trial types ( III – Threshold Estimation ). Finally, we evaluated whether performance in Unlabeled trials aligns more closely with ON or OFF trials ( IV – Supra-threshold Performance ). Supporting data and scripts are accessible at Github: https://github.com/AnneScholz/Visual_Threshold_Mindwandering.git I - Brain state Distribution In the first step, we determined the probability of each individual rating (1–5) for each subject. We found that the subjects displayed a typical distribution of MW ratings, with a high probability for 3 and a decreasing frequency from 2 to 1 and from 4 to 5. Therefore, we fitted a quadratic function $$\:q\:=\:a*{x}^{2}\:+\:b*x\:+\:c$$ to the likelihood values of the ratings, with 'a' as the curvature parameter, 'b' as the shift parameter in the x-direction, and 'c' as the shift parameter in the y-direction. 'b' indicates whether the entire distribution is shifted more towards ON (positive values) or OFF (negative values). In each condition and under each instruction, we tested the distribution of shift parameter b against 0. Then, we compared the shift parameters in a 4 x 3 ANOVA with the factors Visual Stimulus Type (Grating, Curvature, Color, Faces) and Instruction (Detection, Global, and Local Discrimination) across subjects. II - Reaction times We calculated the average reaction times across all trials for each participant. Consequently, three threshold values were acquired for each individual, and these were subjected to a 4 x 3 x 3 ANOVA with the factors Visual Stimulus Type (Grating, Curvature, Color, Faces), Instruction (Detection, Global, and Local Discrimination), and Trial Type (OFF, ON, Unlabeled) for comparison. III - Threshold Estimation The stimuli in the experiments differed in terms of their physical characteristics and their scale. To make the threshold values comparable across the experiments, we transformed the scale of each target's characteristic (e.g., range of x in the Grating experiment: 1.5° to 15°) linearly into unit scale. To achieve this, we first subtracted the minimum value of x (x = x - x min ) and then standardized the resulting vector based on the highest value (x = x*(1/x max )). As a result, each experiment's target scale varies in the range [0, 1] (see Fig. 4 A for unity-based normalization of psychometric functions). In each experiment, we determined the probability of correctly detecting the target in each block. For each participant, we fitted a psychometric function $$\:p\:=\:a\:+\:\frac{(b-a)}{{(1+\:\frac{x}{c\:})}^{d}}$$ to the resulting values to model the relationship between stimulus variability ( x - target orientation strength) and perceptual accuracy p . We defined the visual threshold of each participant as the stimulus values of 75% response accuracy. This was done for both the OFF, ON, and Unlabeled trials. Since we compared the performance across a wide range of different stimuli, the thresholds, on average, did not fall exactly in the middle between the largest and smallest manifestations of the target's characteristics. To eliminate this effect, we centered the thresholds across experiments. To achieve this, we calculated the mean threshold across all participants and instructions for each experiment. We then subtracted this mean threshold from each participant's threshold in the respective experiment and added 0.5. Note that this procedure does not alter the differences between ON, OFF, and Unlabeled trials. As a result, three threshold values were obtained for each participant, which were compared with the 4 x 3 x 3 ANOVA with the factors Visual Stimulus Type (Grating, Curvature, Color, Faces), Instruction (Detection, Global, and Local Discrimination) and Trial Type (OFF, ON, Unlabeled). We conducted post-hoc analyses using paired t-tests. In addition to p-values, we computed Bayes Factors (BF 10 ) for each t-test to assess the strength of evidence in favor of any observed difference. IV - Supra-threshold Accuracy Subsequently, we analyzed whether there was an accuracy difference between the OFF- and ON- states in trials where the stimuli were above the perception threshold. In each experiment, we took the average perception thresholds across all subjects in the Unlabeled trials as calculated in the previous step. All stimulus intensities above this threshold were considered supra-threshold. The accuracy in trials with supra-threshold stimulus intensities was separately averaged for the ON-state, the OFF-state, and the Unlabeled trials. As a result, three performance values were obtained for each participant, which were compared with the 4 x 3 x 3 ANOVA with the factors Visual Stimulus Type (Grating, Curvature, Color, Faces), Instruction (Detection, Global, and Local Discrimination) and Trial Type (OFF, ON, Unlabeled). Post-hoc t-tests were performed, including Bayes Factor analyses for each test. Subsequently, we compared the performance for each experiment separately between the trial types using a one-way ANOVA. Results I - Brain State Distribution We first assessed whether brain state reports differed across experiments and conditions. Participants showed the expected distribution of MW ratings, with lowest likelihood of extreme labels (mean likelihood of labels across all experiments: 1–8.2%; 2–19.53%; 3–27.5%; 4–29.82%; 5–14.96%). We tested whether participants showed a preference for reporting ON or OFF depending on the stimulus set. We fitted a square function of the individual likelihood distributions and tested for difference in the shift parameter. In all stimulus types and under all instructions, a positive shift parameter was observed indicating a stronger tendency to ON reports (see Fig. 2 A; Grating: M b = .27, SD b = .25, t 59 = 8.50, p<.0001, Cohen's d = 1.09, 95% CI [0.78, 1.42]; Curvature: M b = .33, SD b = .21, t 57 = 12.09, p<.0001, Cohen’s d = 1.59, 95% CI [1.20, 1.98]; Color: M b = .23, SD b = .28, t 59 = 6.49, p<.0001, Cohen's d = 0.84, 95% CI [0.54, 1.13]; Faces: M b = .23, SD b = .25, t 61 = 7.25, p<.0001, Cohen’s d = 0.92, 95% CI [0.62, 1.22]; see Fig. 2 B; Detection: M b = .26, SD b = .24, t 78 = 9.62, p<.0001, Cohen's d = 1.08, 95% CI [0.80, 1.36]; Global: M b = .25, SD b = .26, t 81 = 8.62, p<.0001, Cohen's d = 0.95, 95% CI [0.69, 1.21]; Local: M b = .30, SD b = .25, t 78 = 10.39, p<.0001, Cohen's d = 1.17, 95% CI [0.88, 1.46]). The 2 x 3 ANOVA did not show a main effect for Visual Stimulus Type (F 3,228 = 1.98, p = .117, η p ² = 0.03, 95% CI [0.01, 0.06]), nor a main effect for Instruction (F 2,228 = .85; p = .426, η p ² = 0.007, 95% CI [0.001, 0.03]), nor an interaction effect (F 6,228 = .69; p = .651, η p ² = 0.02, 95% CI [0.01, 0.04]) indicating that subjects chose ratings comparably across experiments. II - Reaction times Next, we examined reaction times across the different stimulus types and conditions. We found a significant main effect for the factor Visual Stimulus Type (F 3,684 = 65.86; p < .0001, η p ² = 0.22, 95% CI [0.19, 0.26]; see Table S1 for an overview of reaction times) with participants exhibiting shorter reaction times in the Grating and Face experiments compared to the Curvature and Color experiments. Additionally, we observed a significant main effect for the factor Instruction (F 2,684 = 37.19, p < .0001, η p ² = 0.09, 95% CI [0.08, 0.12]), attributed to quicker reaction times in Global Discrimination (M global = 468.6 ms; SD global = 111.4 ms) compared to Detection (M detection = 532.4 ms; SD detection = 116.9 ms; t 481 = 6.14, p < .0001, Cohen's d = 0.56, 95% CI [0.38, 0.74]) and Local Discrimination (M local = 545.5 ms; SD local = 123.0 ms; t 481 = 7.21, p < .0001, Cohen's d = 0.66, 95% CI [0.47, 0.84]). We also found a significant main effect for Trial Type (F 2,684 = 3.1, p = .0458, η p ² = 0.01, 95% CI [0.004, 0.02]). However, when we compared the trial types separately for each experiment, we only identified a significant main effect in the Color experiments (F 2,177 = 3.06, p = .0490, η p ² = 0.03, 95% CI [0.02, 0.07]). This effect stems from a difference between ON and OFF trials (t 118 = 2.65, p = .0089, Cohen’s d = 0.49, 95% CI [0.12, 0.85]; see Table S2 and Fig. 3 ). III - Threshold Estimation When comparing the threshold values of ON, OFF, and Unlabeled trials, we found a main effect for the factor Instruction (F 2,684 = 28.28, p < .0001, η p ² = 0.08, 95% CI [0.06, 0.09]; see Table S3 for an overview of perceptual thresholds) due to a lower threshold in Global Discrimination (M global = .44; SD global = .11) compared to Detection (M detection = .52; SD detection = .14; t 481 = 5.96, p < .0001, Cohen's d = 0.54, 95% CI [0.36, 0.72]) and Local Discrimination (M local = .54; SD local = .18; t 481 = 6.62, p < .0001, Cohen's d = 0.60, 95% CI [0.42, 0.79]). The comparison between Detection and Local Discrimination thresholds revealed only a trend (t 472 = 1.7, p = .08, Cohen's d = -0.16, 95% CI [-0.34, 0.02]). Furthermore, there was a main effect for Trial Type (F 2,684 = 29.32; p < .0001, η p ² = 0.08, 95% CI [0.06, 0.10]) due to a lower threshold in ON trials (M ON = .44; SD ON = .14) compared to OFF trials (M OFF = .52; SD OFF = .17; t 478 = 5.44, p < .0001, Cohen's d = -0.49, 95% CI [-0.68, -0.32]) and Unlabeled trials (M Unlabeled = .53; SD Unlabeled = .12; t 478 = 7.13, p < .0001, Cohen’s d = -0.65, 95% CI [-0.84, -0.47]). The thresholds for OFF trials and Unlabeled trials did not differ (t 478 = .6, p = .52, Cohen’s d = -0.06, 95% CI [-0.24, 0.12]). In the next step, we compared the thresholds separately for each experiment using a one-way ANOVA with the factor Trial Type (ON, OFF, Unlabeled). We found a significant main effect for Trial Type in the Grating, Curvature, and Color experiments (Grating: F 2,177 = 9.15, p = .0002, η p ² = 0.09, 95% CI [0.06, 0.15] – Curvature: F 2,171 = 9.33, p = .0001, η p ² = 0.09, 95% CI [0.06, 0.15] – Color: F 2,177 = 7.45, p = .0008, η p ² = 0.08, 95% CI [0.05, 0.13]), but only a trend with Face stimuli (Faces: F 2,183 = 2.91, p = .0572, η p ² = 0.03, 95% CI [0.01, 0.07]). Next, we performed post-hoc comparisons of the thresholds between the different trial types. We observed that the thresholds in the Grating, Curvature, and Color experiments differed between ON trials and Unlabeled trials (see Table S4 and Fig. 5 B). However, this comparison in the Face experiments did not survive the Bonferroni Correction. Furthermore, we found that ON and OFF trial thresholds differed, but only in the Gratings and Curvature experiments. This comparison was not significant at the corrected level in the Color experiment. Additionally, there was no difference in the Face experiment at the normal, not-corrected level. Thus, ON and OFF states lead to differences in visual performance and differences diminish with increasing stimulus complexity. Importantly, no significant difference between OFF and Unlabeled trials was found in any of the four stimulus sets. This suggests that the threshold estimation in standard experimental settings more closely reflects an OFF-state, rather than an ON-state, reinforcing our assumption about the nature of typical perceptual measurements. IV - Suprathreshold Accuracy Finally, we assessed whether suprathreshold accuracy in Unlabeled trials aligns more with ON or OFF trial accuracy. We found a significant main effect for the factor Visual Stimulus Type (F 3,684 = 19.54, p < .0001, η p ² = 0.08, 95% CI [0.06, 0.10]), as participants exhibited lower accuracy in the Curvature experiment (M curvature = .88; SD curvature = .06) compared to Grating (M grating = .92; SD grating = .05; t 352 = 5.39; p < .0001, Cohen's d = 0.57, 95% CI [0.36, 0.79]), Color (M color = .93; SD color = .08; t 352 = 5.40, p < .0001, Cohen's d = -0.58, 95% CI [-0.79, -0.36]), and Faces (M faces = .91; SD faces = .06; t 358 = 3.68, p = .0003, Cohen’s d = -0.39, 95% CI [-0.59, -0.18]). None of the pairwise comparisons of Grating, Color, and Face experiments showed a significant difference (p Bonferroni = .0083; all p values > .01). We also found a main effect for Trial Type (F 2,684 = 33.17, p < .0001, η p ² = 0.09, 95% CI [0.07, 0.11]), and an interaction effect for the factors Visual Stimulus Type and Trial Type (F 6,684 = 2.37, p = .0283, η p ² = 0.02, 95% CI [0.01, 0.03]). When we compared the accuracy values between trial types separately for each experiment, we found a significant main effect for Trial Type in the Grating, Curvature, and Color experiments (Grating: F 2,177 = 23.54, p < .0001, η p ² = 0.21, 95% CI [0.16, 0.28] – Curvature: F 2,171 = 7.14, p = .0011, η p ² = 0.08, 95% CI [0.05, 0.13] – Color: F 2,177 = 3.61, p = .0289, η p ² = 0.04, 95% CI [0.02, 0.08]). However, in the Face experiments, only a trend was observed (Faces: F 2,183 = 2.89, p = .0581, η p ² = 0.03, 95% CI [0.01, 0.07]). Next, separately for each experiment we performed post-hoc comparisons of the suprathreshold accuracy between the different trial types. We observed that the thresholds in the Grating, Curvature, and Color experiments differed between ON trials and Unlabeled trials (see Table S5 for suprathreshold values, Table S6 for post hoc comparisons, and Fig. 6 B). This comparison in the Face experiment did not survive the Bonferroni correction. Post hoc comparisons between ON and OFF trials differed only in the Gratings and Curvature experiments (see Table S6 ). This comparison was not significant at the corrected level in the Color experiment. Additionally, there was no difference in the Face experiment at the normal, not-corrected level. The BF 10 -values for the comparison between Unlabeled and OFF trials (see Table S6 ) provide substantial evidence in favor of the null hypothesis in the Curvature and Faces experiments, indicating that accuracy in Unlabeled trials closely resembles that of OFF trials. Discussion We investigated whether threshold estimations in typical visual search paradigms are influenced by attentional brain states. To test this, we examined the influence of MW on the detection and discrimination of targets in a range of visual search paradigms. Given MW occurs in up to 50% of the waking time, we explored if MW epochs altered visual detection performance across an experimental session. We found that accuracy in the Unlabeled test trials resembled the OFF-state more closely than the ON-state. This discrepancy challenges the conventional assumption that subjects are fully attentive across an experimental setting and that measuring the visual threshold adequately captures visual perception capacity. The alteration in perception manifested in two main ways: a shift in thresholds and reduced perception in suprathreshold conditions. Importantly, we observed that MW did not uniformly deteriorate visual perception. Instead, it primarily impacted the perception of simple stimuli and the influence of MW diminished with increasing complexity. We found no differences in reaction times during MW episodes compared to ON trials, contrary to the slowing observed in most studies 8 with pop-out target stimuli 9 . We observed a trend towards slower reaction times in ON trials aligning with Andrillon et al 10 , who distinguished between MW and mind blanking. These authors found faster reaction times during MW compared to ON task trials suggesting increased impulsivity during MW. Motor performance tends to become more automatic and less precise during MW episodes 11 – 15 . Unlike engaged task states, MW is often associated with faster reaction times and increased error rates 16 , 17 as reflected by our results. These variations in reaction times likely depend on the specific task demands. In our experiments, participants were instructed to respond quickly and accurately, possibly reflecting a trade-off where MW prioritizes speed over accuracy, while ON-task states emphasize accuracy. This acceleration in reaction times suggests that MW not only degrades performance but also alters the dynamic control of motor outputs (Kam et al., 2012). The prevailing notion has been that MW involves a general decoupling from the external world to shield internal thoughts from external distractions. However, our data challenge this assumption for two reasons. First, in our study, participants exhibited altered performance during MW but maintained a clear supra-threshold response as reflected by high rates of accuracy. This finding suggests they remained connected to their surroundings rather than functioning at a random level as it would be predicted with the decoupling hypothesis. Second, participants displayed a threshold shift and reduced accuracy in supra-threshold trials only observed with simple stimuli. As stimulus complexity increased, the performance impairment associated with MW diminished. This suggests that attention modulation has a greater impact when processing simpler stimuli 2 – 5 . This finding suggests that the influence of MW is best understood as a gradual shift in perception, rather than an all-or-nothing effect. This observed MW gradient aligns with previous research. Classic Letter Sustained Attention to Response Task (SART) paradigms (similar to our gratings or curvature stimuli) reveal task-unrelated thought effects, leading to more commission errors during OFF-task trials 17 . This effect is not observed in Face SART paradigms 19 , where face detection is affected by MW. In sum, MW does not fully decouple perception from external stimulation. Our study highlights a methodological consideration. Numerous studies have focused on distinguishing between ON- and OFF-states. However, no study has directly compared the ON-state with an unlabeled state. Instead, it has been common practice to not only use the trials immediately preceding the focused inquiry but also to average over a variable number of preceding trials. Consequently, parameters in previous studies were often computed by averaging data from several preceding trials, assuming implicitly that participants remained in the same brain state. This averaging was achieved through methods utilizing a fixed number of preceding trials 8 , 20 or specifying a time duration, such as several seconds 21 . Another approach, as seen in Arnau et al. 22 , involves selecting different numbers of trials preceding ON- and OFF-ratings. In this method, trials were categorized as "mind wandering" when participants reported such experiences, along with the two preceding trials, while the remaining trials were classified as "on-task" trials. However, this approach is also susceptible to blending different brain states. Here, we demonstrate that this approach, particularly when considering different preceding trials, poses challenges, especially in the "ON" condition. Why did we observe that global performance aligns more with OFF-task processing? In fact, participants more frequently report being "ON the task" than "OFF the task" which should also be represented in the unlabeled trial. A stronger resemblance of unlabeled trials with OFF trials can only occur if the participants are in the OFF state more frequently than they report. This inconsistency may be attributed to participants more often selecting the MID category when they are uncertain about their brain state. This may occur for example due to social desirability biases which refers to the tendency of individuals to respond to self-report items in a manner that presents themselves in a more socially favorable light according to prevailing societal norms and standards. Here we obtained focus queries sporadically and randomly to minimize the risk of reducing the likelihood of MW 7 . Conversely, this implies that the absence of focus queries may lead to even more instances of MW. We posit that this affects performance, in the sense that without interruption, more MW occurs deteriorating performance. In conclusion, our study sheds light on the complex interplay between brain states, MW, and visual perception. We emphasize the importance of considering the task type, stimulus complexity, and the dynamic nature of brain states when investigating MW's effects. These findings have implications for understanding how MW impacts cognitive processes and the need to refine experimental methodologies to accurately capture its influence on task performance. Declarations Competing Interest Statement: All authors declare no competing interests. Author Contributions: S.D. conceived and designed the experiment. A.S., T.W., and P.S. collected the data. A.S., T.W., P.S., C.R., and S.D. analyzed the data, A.S., T.W., P.S., C.R., R.T.K., and S.D. interpreted the data. A.S., T.W., P.S., C.R., R.T.K., and S.D. wrote the manuscript. Acknowledgments: This work was supported by the Deutsche Forschungsgemeinschaft (DFG; Grant SFB 1436 A03). The authors thank Christiane Petzold and Nathalie Vogt for assisting in the data collection. References Smallwood, J. Mind-wandering While Reading: Attentional Decoupling, Mindless Reading and the Cascade Model of Inattention. Linguistics and Language Compass 5, 63–77 (2011). Luck, S. J. & Hillyard, S. A. Spatial filtering during visual search: Evidence from human electrophysiology. J. Exp. Psychol. Hum. Percept. Perform. 20, 1000–1014 (1994). Martinez-Trujillo, J. C. & Treue, S. Feature-Based Attention Increases the Selectivity of Population Responses in Primate Visual Cortex. Posner, M. I. & Gilbert, C. D. Attention and primary visual cortex. Proc. Natl. Acad. Sci. U. S. A. 96, 2585–2587 (1999). Steven J. Luck & Edward K. Vogel. The capacity of visual working memory for features and conjunctions. Nature 390, 279–281 (1997). Brainard, D. H. The Psychophysics Toolbox. Spat. Vis. 10, 433–436 (1997). Seli, P., Carriere, J. S. A., Levene, M. & Smilek, D. How few and far between? Examining the effects of probe rate on self-reported mind wandering. Front. Psychol. 4, 430 (2013). Dong, H. W., Mills, C., Knight, R. T. & Kam, J. W. Y. Detection of mind wandering using EEG: Within and across individuals. PLoS One 16, (2021). Wienke, C. et al. Mind-wandering Is Accompanied by Both Local Sleep and Enhanced Processes of Spatial Attention Allocation. Cereb. Cortex Commun. 2, (2021). Andrillon, T., Burns, A., Mackay, T., Windt, J. & Tsuchiya, N. Predicting lapses of attention with sleep-like slow waves. Nature Communications 2021 12:1 12, 1–12 (2021). Carriere, J. S. A., Cheyne, J. A. & Smilek, D. Everyday attention lapses and memory failures: The affective consequences of mindlessness. Conscious. Cogn. 17, 835–847 (2008). Cheyne, J. A., Carriere, J. S. A. & Smilek, D. Absent-mindedness: Lapses of conscious awareness and everyday cognitive failures. Conscious. Cogn. 15, 578–592 (2006). Reichle, E. D., Reineberg, A. E. & Schooler, J. W. Eye movements during mindless reading. Psychol. Sci. 21, 1300–1310 (2010). Schooler, Jonathan W.; Reichle, Erik D.; Halpern, D. V. Zoning Out while Reading: Evidence for Dissociations between Experience and Metaconsciousness. Thinking and seeing: Visual metacognition in ad 203–226 Preprint at (2004). Weissman, D. H., Roberts, K. C., Visscher, K. M. & Woldorff, M. G. The neural bases of momentary lapses in attention. Nat. Neurosci. 9, 971–978 (2006). Franklin, M. S., Smallwood, J. & Schooler, J. W. Catching the mind in flight: Using behavioral indices to detect mindless reading in real time. Psychon. Bull. Rev. 18, 992–997 (2011). Smallwood, J. et al. Subjective experience and the attentional lapse: Task engagement and disengagement during sustained attention. Conscious. Cogn. 13, 657–690 (2004). Kam, J. W. Y. et al. Mind wandering and motor control: off-task thinking disrupts the online adjustment of behavior. 6, 1–9 (2012). Denkova, E., Brudner, E. G., Zayan, K., Dunn, J. & Jha, A. P. Attenuated face processing during mind wandering. J. Cogn. Neurosci. 30, 1691–1703 (2018). Jin, C. Y., Borst, J. P. & van Vugt, M. K. Decoding study-independent mind-wandering from EEG using convolutional neural networks. J. Neural Eng. 20, (2023). Wamsley, E. J. & Summer, T. Spontaneous entry into an “offline” state during wakefulness: A mechanism of memory consolidation? J. Cogn. Neurosci. 32, 1714–1734 (2020). Arnau, S. et al. Inter-trial alpha power indicates mind wandering. Psychophysiology 57, (2020). Additional Declarations There is NO Competing Interest. Supplementary Files SIVisualThresholdFINAL.docx Supporting Information Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9412640","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":631827908,"identity":"d7f71a84-53d7-4081-943d-11825b4b60e6","order_by":0,"name":"Stefan Dürschmid","email":"data:image/png;base64,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","orcid":"","institution":"Leibniz Institute for Neurobiology","correspondingAuthor":true,"prefix":"","firstName":"Stefan","middleName":"","lastName":"Dürschmid","suffix":""},{"id":631827909,"identity":"eeb26cbe-814a-4591-886c-eadf03aa8979","order_by":1,"name":"Annemarie Scholz","email":"","orcid":"https://orcid.org/0009-0002-0204-9556","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Annemarie","middleName":"","lastName":"Scholz","suffix":""},{"id":631827910,"identity":"114e517a-4bf9-44e3-b6ea-7da37755b39e","order_by":2,"name":"Christoph Reichert","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Christoph","middleName":"","lastName":"Reichert","suffix":""},{"id":631827911,"identity":"d6755840-60a1-46a0-bcf3-9deca97b49da","order_by":3,"name":"Paul Schmid","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Paul","middleName":"","lastName":"Schmid","suffix":""},{"id":631827912,"identity":"995364e7-2b3a-4826-bfb9-815f0b110f57","order_by":4,"name":"Tom Weischner","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Tom","middleName":"","lastName":"Weischner","suffix":""},{"id":631827913,"identity":"9be0b685-b2f4-4b79-8d79-69c5c7fd1470","order_by":5,"name":"Robert Knight","email":"","orcid":"https://orcid.org/0000-0001-8686-1685","institution":"University of California, Berkeley","correspondingAuthor":false,"prefix":"","firstName":"Robert","middleName":"","lastName":"Knight","suffix":""}],"badges":[],"createdAt":"2026-04-14 08:34:06","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9412640/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9412640/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108697981,"identity":"5c40dcc6-d490-4241-91bc-7e67885e5f18","added_by":"auto","created_at":"2026-05-07 12:12:19","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":229247,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eDepiction of experimental paradigm and frequency of Mind Wandering\u003c/em\u003e. \u003cem\u003e\u003cstrong\u003eA\u003c/strong\u003e\u003c/em\u003e Stimuli presented in the respective experiments. The first column shows the distractors used in each experiment, while the second and third columns show the targets with maximum left and right orientation, respectively. \u003cem\u003e\u003cstrong\u003eB\u003c/strong\u003e\u003c/em\u003e Spatial distribution of the positions where stimuli were presented in each experiment.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9412640/v1/e1384713fcbb61631911ae6f.png"},{"id":108698002,"identity":"eabcb44a-ac88-483d-a325-3100ffead8cd","added_by":"auto","created_at":"2026-05-07 12:12:25","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":201147,"visible":true,"origin":"","legend":"\u003cp\u003eFrequency distribution of the five different ratings, ranging from completely OFF (1) to completely ON (5) in the four different paradigms (\u003cem\u003e\u003cstrong\u003eA\u003c/strong\u003e\u003c/em\u003e) and three different instructions (\u003cem\u003e\u003cstrong\u003eB\u003c/strong\u003e\u003c/em\u003e). Colors refer to the outline colors in \u003cem\u003e\u003cstrong\u003eFigure 1 A\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e. \u003c/em\u003eThe scatter plot represents the shift in distribution for each participant and each experiment indicating that the participants were more likely to report being ON- than OFF-task.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9412640/v1/dd9399386ae1357ace2eece5.png"},{"id":108697994,"identity":"da023ae8-d3f7-4cb8-a5a8-ca9f37173c0a","added_by":"auto","created_at":"2026-05-07 12:12:21","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":60362,"visible":true,"origin":"","legend":"\u003cp\u003eDepicts reaction times for the trial types, separately for each paradigm. The colored bars represent the F-value for the main effect of Trial Type. The boxplots show the reaction time distribution for ON, Unlabeled, and OFF (from left to right, respectively). In sum there is only a trend towards faster reaction times for OFF than ON trials in the Color experiment.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-9412640/v1/2111e52d165c005e86b7b34a.png"},{"id":108697958,"identity":"30367ed5-d584-4165-a91e-67a8ee2616f0","added_by":"auto","created_at":"2026-05-07 12:12:10","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":256914,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eThreshold shifts due to brain state changes\u003c/em\u003e. \u003cem\u003e\u003cstrong\u003eA\u003c/strong\u003e\u003c/em\u003e Thresholds for each paradigm and instruction. The thick lines represent the average across all participants, and the shaded areas represent the standard error across subjects. \u003cem\u003e\u003cstrong\u003eB\u003c/strong\u003e\u003c/em\u003eEstimated threshold values for the paradigms, instructions, and trial types. Each triplet group shows the thresholds for OFF (top), unlabeled (middle), and ON trials (bottom). The black bars indicate the standard error across participants.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-9412640/v1/1eed502384d1320a63778c3f.png"},{"id":108697971,"identity":"218af727-2a9b-41eb-9021-25842eccc14b","added_by":"auto","created_at":"2026-05-07 12:12:17","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":157259,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eA\u003c/strong\u003e\u003c/em\u003e\u003cem\u003e \u003c/em\u003eDistribution of thresholds for the three trial types. In ON trials, participants exhibited lower thresholds compared to OFF and unlabeled trials. \u003cem\u003e\u003cstrong\u003eB\u003c/strong\u003e\u003c/em\u003e Distribution of thresholds within the trial types, separately for each paradigm. The bars represent the F-value for the main effect of trial type. The F-values decrease from simple to more complex stimuli.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-9412640/v1/46885718ec1e6a652c0c60c7.png"},{"id":108697972,"identity":"292ca975-9557-4d9c-b946-3ccb1544cf6e","added_by":"auto","created_at":"2026-05-07 12:12:17","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":132218,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eA\u003c/strong\u003e\u003c/em\u003eSuprathreshold performance for the different trial types. Performance in ON-trials is better than in unlabeled and OFF-trials. Conversely, the difference between OFF and unlabeled trials is smaller. \u003cem\u003e\u003cstrong\u003eB \u003c/strong\u003e\u003c/em\u003eSuprathreshold performance separately for the four different paradigms. In this case, it becomes evident that as the complexity increases, the difference between the trial types decreases.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-9412640/v1/3b89fe6df0fa1b391033648a.png"},{"id":108698047,"identity":"d0ca0495-5389-4278-9731-a10ff04d2adf","added_by":"auto","created_at":"2026-05-07 12:12:34","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1075961,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9412640/v1/fdb880f2-7d7f-49ab-a1fd-be1f6b777ab2.pdf"},{"id":108697903,"identity":"7315e954-4b50-43e4-a60e-d962bd462be8","added_by":"auto","created_at":"2026-05-07 12:12:03","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":34675,"visible":true,"origin":"","legend":"Supporting Information","description":"","filename":"SIVisualThresholdFINAL.docx","url":"https://assets-eu.researchsquare.com/files/rs-9412640/v1/22d4518c69f1ff2add27b3fa.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Visual Thresholds Cannot Be Reliably Measured Without Controlling for Mind-Wandering","fulltext":[{"header":"Significance Statement","content":"\u003cp\u003ePerceptual thresholds are widely used to define the capacity of the human visual system, under the assumption that attention remains stable across trials. By combining experience sampling with behavioral performance across twelve tasks, we demonstrate that mind-wandering introduces measurable fluctuations in perceptual capacity. Our study shows that perceptual threshold estimation is compromised in inattentive states, particularly for simple stimuli. These findings highlight the necessity of accounting for attentional state in perceptual research to obtain valid measures of human vision.\u003c/p\u003e"},{"header":"Introduction","content":"\u003cp\u003eThe visual perception threshold assesses the degree of efficient perception. Typically, the probability of correct performance is estimated across multiple experimental trials. We obtained a psychometric function that quantitatively characterizes the perceptual sensitivity of the visual system. Participants were instructed to maintain attention throughout the duration of the task, assuming full attentiveness across trials (ON state). However, in humans, monkeys and rodents brain states undergo fluctuations affecting visual perception. Mind wandering (MW; OFF state) is a cognitive phenomenon occurring up to 50% of our daily lives and is often described as attentional decoupling from the external world due to decreases in behavioral performance (Smallwood, 2011). Standard perceptual threshold testing is typically assumed to be dependent by ON-task behavior. However, this assumption has not yet been empirically tested. Given the high frequency of MW in everyday life, it is unrealistic to expect continuous full attention during visual detection and discrimination tasks. Consequently, threshold estimation in perceptual research may be biased by MW, resulting in a systematic underestimation of true visual performance under standard experimental conditions. This raises the critical question of whether threshold estimation is affected by MW.\u003c/p\u003e \u003cp\u003eFeature-based attention improves perception of basic features accompanied by a shrinkage of receptive field size \u003csup\u003e\u003cspan additionalcitationids=\"CR3 CR4\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Neurons with larger receptive fields in the primate visual system have lower attentional demand and may exhibit a tolerance to variations in brain states like MW. To investigate the impact of MW on visual perception depending on the attended feature, we conducted a series of experiments using four types of visual stimuli \u0026ndash; varying in stimulus complexity (orientation, curvature, color, and faces) - across three task conditions (detection, global discrimination, and local discrimination). Experience-sampling probes were interspersed throughout each experiment to dissociate perceptual performance across distinct brain states in a subset of trials (ON-task vs. OFF-task). The majority of trials without mind-wandering probes represented the ground truth against which ON- and OFF-task performance was assessed. Two critical questions arise. First, if attention enhances performance, then accuracy should be higher during ON-task relative to OFF-task trials. Second, if standard threshold estimation faithfully reflects the perceptual capacity of the visual system, then estimates derived from the ground truth should resemble ON-task performance. We hypothesized that visual threshold measurements for features processed early in the visual hierarchy primarily reflect performance under the influence of MW. Our findings demonstrate that averaging across all available trials\u0026mdash;without accounting for attentional fluctuations\u0026mdash;yields unreliable estimates of visual thresholds, with estimation errors increasing as stimulus complexity decreases.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eParticipants\u003c/h2\u003e \u003cp\u003e After obtaining written informed consent, we tested 239 subjects (156 female, mean age: 24.63y) across 12 experiments. All subjects were compensated with 8\u0026euro;/hour or course credit. The study was approved by the local ethics committee (\u0026ldquo;Ethical Committee of the Otto-von-Guericke University Magdeburg\u0026rdquo;).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eGeneral Paradigm\u003c/h3\u003e\n\u003cp\u003eWe investigated the impact of MW on visual perception of target stimuli defined by Orientation (\u003cem\u003eParadigm 1\u003c/em\u003e), Curvature (\u003cem\u003eParadigm 2\u003c/em\u003e), Color (\u003cem\u003eParadigm 3\u003c/em\u003e), and viewing direction of Faces (\u003cem\u003eParadigm 4\u003c/em\u003e). Each visual feature was tested in three different experiments defined by the instruction (Detection, Global, and Local Discrimination). To maintain consistency across all experiments and ensure comparability, 18 stimuli were presented in each trial, divided into two blocks of nine stimuli each (arranged in a 3 x 3 grid, see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB) positioned to the left and right of the fixation cross. The stimuli, as well as the experimental setup, were implemented using Matlab R2013a (Mathworks, Natick, USA) and the Psychophysics Toolbox extension \u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. All stimuli were presented on a gray background using a 14'' color monitor with a resolution of 1.920x1.200 pixels and a refresh rate of 60 Hz. Distance to the display was kept at 70 cm.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eParadigm 1: Orientation Perception\u003c/h3\u003e\n\u003cp\u003e Participants were presented with simple grating stimuli (Gabor Patches; see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA), where the target stimulus differs in orientation from the surrounding distractor stimuli. Participants were required to either detect the presence of a target (Detection), determine its location (Global Discrimination), or identify the direction of its tilt (Local Discrimination). The distractors were uniformly vertically aligned, while the target's angle varied between 1\u0026deg; and 24.8\u0026deg; across different blocks, with 15 steps of 1.7\u0026deg;. Each stimulus consisted of four parallel black stripes on a white background and had a total size of 80 x 80 pixels. To minimize edge artifacts, a Gaussian filter with an 80 x 80 pixel size was applied to each grating. The visible portion of each Gabor patch occupied approximately 0.81\u0026deg; of visual angle.\u003c/p\u003e\n\u003ch3\u003eParadigm 2: Curvature Perception\u003c/h3\u003e\n\u003cp\u003eTargets consisted of line stimuli with a stronger curvature compared to the distractors (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). Participants were required to indicate whether a target was presented (Detection), where the target was located (Global Discrimination), or in which direction it was curved (Local Discrimination). In order to generate the curved stimuli, black circles with varying diameters were created and then one half of each circle was covered by a rectangle matching the grey background. This way, circles with a smaller diameter resulted in stimuli with a stronger curvature. The diameters of target stimuli varied between 30 and 44 pixels in 15 equal steps, while the distractors had a constant diameter of 45 pixels. The resulting stimuli consisted of a black curved line taking up approximately 0.98\u0026deg; va.\u003c/p\u003e\n\u003ch3\u003eParadigm 3: Color Perception\u003c/h3\u003e\n\u003cp\u003eTargets were displayed with one side appearing red and the other side appearing more pinkish, while the distractors were red on both sides (RGB: 152 0 15). For each block, one out of 15 color values was chosen for the target stimulus (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). Participants' tasks included identifying the presence of a target (Detection), the location of a target (Global Discrimination) or which side of the target was more pinkish (Local Discrimination). The targets and distractors were composed of two semicircles with a vertical gap measuring 0.1\u0026deg; va and a total height of 0.76\u0026deg; va. The gap was selected to eliminate discrepancies in spatial frequency caused by color gradients between the targets and distractors. To prevent edge artifacts, the semicircles were overlaid with a Gaussian filter (3 pixels).\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eParadigm 4: Face Perception\u003c/h2\u003e \u003cp\u003eStimuli consisted of face pictures with a neutral facial expression and varying degrees of pixelation. Participants were required to either detect the target face among distractor stimuli (Detection), differentiate where it was presented (Global Discrimination) or decipher the gaze direction of the target face (Local Discrimination). To capture the face, we took a photograph of a mannequin's face from a 24.5\u0026deg; angle on the left side, resulting in the face looking towards the right. The image was then mirrored to make it appear as if the face was gazing at the left side of the screen. An oval was superimposed over the face to cover hair and ears. Finally, the pixels of targets were shuffled at varying degrees (30\u0026ndash;86% of pixels, in 15 equal steps of 4 pixels), while the distractors had a higher degree of shuffling (90% of pixels; see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). Each stimulus occupied around 1.3\u0026deg; va in height and 0.73\u0026deg; va in width.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eGeneral Procedure\u003c/h3\u003e\n\u003cp\u003eAt the start of each trial, a black fixation cross was displayed. Following a delay of 750 ms (\u0026plusmn;\u0026thinsp;250 ms), participants were presented with a visual search display below the fixation cross for 100 ms. This display consisted of 18 stimuli, with one target and 17 distractor stimuli, consistent across all experiments. In 50% of the trials, the target was positioned either to the left or right (as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB) with equal likelihood of appearing at any of the 18 positions. The different experiments varied in their instructions. In the Detection experiments, the target was shown in 50% of pseudo-randomly selected trials. Participants were instructed to indicate whether a target was present (using the \"F\" key) or absent (using the \"J\" key) on a standard QWERTZ keyboard. In the Global and Local Discrimination experiments, the target was always presented, while its location and orientation were chosen pseudo-randomly (50% right, 50% left). Participants were asked to indicate whether the target appeared to the left (using the \"F\" key) or to the right (using the \"J\" key) of the fixation cross for Global Discrimination, or whether the left-ward oriented version (using the \"F\" key) or right-ward oriented version (using the \"J\" key) of the target was presented for Local Discrimination.\u003c/p\u003e \u003cp\u003eEach experiment commenced with a practice block consisting of 22 trials for all participants. The orientation of the target was randomized across these practice trials. During the practice block, two focus queries were administered. The location and orientation of the target (Global / Local Discrimination) or whether a target was presented (Detection) was randomized within each trial of the practice block. In the subsequent experiment, participants were exposed to a total of 1620 trials (108 per block) divided into 15 blocks, with 330 focus questions (see \u003cb\u003eExperience Sampling\u003c/b\u003e) in total (22 per block). The order of the 15 different deviation strengths was randomly varied across the blocks. Within a given block, the deviance strength of the target from the distractors remained consistent. Between experimental blocks, participants took a break and could proceed to the next experimental block at their own pace. The entire experiment took about 1.5h for each subject.\u003c/p\u003e\n\u003ch3\u003eExperience Sampling\u003c/h3\u003e\n\u003cp\u003eThroughout the experiment, we incorporated thought probes in a pseudorandom manner, selecting them for 20% of all trials. These probes aimed to assess participants' level of MW in the trial immediately preceding the probe. Participants were asked to rate their MW experience on a five-point Likert-scale, ranging from 1 (indicating that their thoughts were elsewhere - OFF) to 5 (indicating that their thoughts were completely focused on the task - ON). Participants responded to the focus questions using the fingers of their left hand, assigning a rating using the following finger-to-rating mapping: thumb for 5, index finger for 4, middle finger for 3, ring finger for 2, and little finger for 1. The thought probes were presented after the visual search display. Furthermore, we ensured that there was a minimum of one intervening trial between two probes (however, this minimum distance was observed in only 7% of the focus queries). The probes were initiated by an auditory stimulus with a frequency of 500 Hz at approximately 85 dB for a duration of 200 ms. This experience sampling approach analyzing a subset of trials was employed since more frequent focus queries could interfere with MW \u003csup\u003e7\u003c/sup\u003e. This results in a limited number of trials that can be assigned with a certain focus rating. Consequently, a significant number of unlabeled trials, usually around 80%, are excluded or discarded (9).\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eData analyses\u003c/h2\u003e \u003cp\u003eFirst, we determined whether there were systematic differences in the distribution of brain state reports between the experiments and conditions (\u003cem\u003eI \u0026ndash; Brain State Distribution\u003c/em\u003e). In the next step, we assessed the reaction times across experiments and conditions \u003cem\u003e(II \u0026ndash; Reaction Times)\u003c/em\u003e. Then, we compared thresholds between the three different trial types (\u003cem\u003eIII \u0026ndash; Threshold Estimation\u003c/em\u003e). Finally, we evaluated whether performance in Unlabeled trials aligns more closely with ON or OFF trials (\u003cem\u003eIV \u0026ndash; Supra-threshold Performance\u003c/em\u003e). Supporting data and scripts are accessible at Github: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/AnneScholz/Visual_Threshold_Mindwandering.git\u003c/span\u003e\u003cspan address=\"https://github.com/AnneScholz/Visual_Threshold_Mindwandering.git\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eI - Brain state Distribution\u003c/h2\u003e \u003cp\u003eIn the first step, we determined the probability of each individual rating (1\u0026ndash;5) for each subject. We found that the subjects displayed a typical distribution of MW ratings, with a high probability for 3 and a decreasing frequency from 2 to 1 and from 4 to 5. Therefore, we fitted a quadratic function\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:q\\:=\\:a*{x}^{2}\\:+\\:b*x\\:+\\:c$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eto the likelihood values of the ratings, with 'a' as the curvature parameter, 'b' as the shift parameter in the x-direction, and 'c' as the shift parameter in the y-direction. 'b' indicates whether the entire distribution is shifted more towards ON (positive values) or OFF (negative values). In each condition and under each instruction, we tested the distribution of shift parameter b against 0. Then, we compared the shift parameters in a 4 x 3 ANOVA with the factors Visual Stimulus Type (Grating, Curvature, Color, Faces) and Instruction (Detection, Global, and Local Discrimination) across subjects.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eII - Reaction times\u003c/h2\u003e \u003cp\u003eWe calculated the average reaction times across all trials for each participant. Consequently, three threshold values were acquired for each individual, and these were subjected to a 4 x 3 x 3 ANOVA with the factors Visual Stimulus Type (Grating, Curvature, Color, Faces), Instruction (Detection, Global, and Local Discrimination), and Trial Type (OFF, ON, Unlabeled) for comparison.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eIII - Threshold Estimation\u003c/h2\u003e \u003cp\u003eThe stimuli in the experiments differed in terms of their physical characteristics and their scale. To make the threshold values comparable across the experiments, we transformed the scale of each target's characteristic (e.g., range of x in the Grating experiment: 1.5\u0026deg; to 15\u0026deg;) linearly into unit scale. To achieve this, we first subtracted the minimum value of x (x\u0026thinsp;=\u0026thinsp;x - x\u003csub\u003emin\u003c/sub\u003e) and then standardized the resulting vector based on the highest value (x\u0026thinsp;=\u0026thinsp;x*(1/x\u003csub\u003emax\u003c/sub\u003e)). As a result, each experiment's target scale varies in the range [0, 1] (see Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e4\u003c/span\u003eA for unity-based normalization of psychometric functions). In each experiment, we determined the probability of correctly detecting the target in each block. For each participant, we fitted a psychometric function\u003c/p\u003e \u003cp\u003e \u003cdiv id=\"Equb\" class=\"Equation\"\u003e \u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:p\\:=\\:a\\:+\\:\\frac{(b-a)}{{(1+\\:\\frac{x}{c\\:})}^{d}}$$\u003c/div\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003eto the resulting values to model the relationship between stimulus variability (\u003cem\u003ex\u003c/em\u003e - target orientation strength) and perceptual accuracy \u003cem\u003ep\u003c/em\u003e. We defined the visual threshold of each participant as the stimulus values of 75% response accuracy. This was done for both the OFF, ON, and Unlabeled trials. Since we compared the performance across a wide range of different stimuli, the thresholds, on average, did not fall exactly in the middle between the largest and smallest manifestations of the target's characteristics. To eliminate this effect, we centered the thresholds across experiments. To achieve this, we calculated the mean threshold across all participants and instructions for each experiment. We then subtracted this mean threshold from each participant's threshold in the respective experiment and added 0.5. Note that this procedure does not alter the differences between ON, OFF, and Unlabeled trials. As a result, three threshold values were obtained for each participant, which were compared with the 4 x 3 x 3 ANOVA with the factors Visual Stimulus Type (Grating, Curvature, Color, Faces), Instruction (Detection, Global, and Local Discrimination) and Trial Type (OFF, ON, Unlabeled). We conducted post-hoc analyses using paired t-tests. In addition to p-values, we computed Bayes Factors (BF\u003csub\u003e10\u003c/sub\u003e) for each t-test to assess the strength of evidence in favor of any observed difference.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eIV - Supra-threshold Accuracy\u003c/h2\u003e \u003cp\u003eSubsequently, we analyzed whether there was an accuracy difference between the OFF- and ON- states in trials where the stimuli were above the perception threshold. In each experiment, we took the average perception thresholds across all subjects in the Unlabeled trials as calculated in the previous step. All stimulus intensities above this threshold were considered supra-threshold. The accuracy in trials with supra-threshold stimulus intensities was separately averaged for the ON-state, the OFF-state, and the Unlabeled trials. As a result, three performance values were obtained for each participant, which were compared with the 4 x 3 x 3 ANOVA with the factors Visual Stimulus Type (Grating, Curvature, Color, Faces), Instruction (Detection, Global, and Local Discrimination) and Trial Type (OFF, ON, Unlabeled). Post-hoc t-tests were performed, including Bayes Factor analyses for each test. Subsequently, we compared the performance for each experiment separately between the trial types using a one-way ANOVA.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eI - Brain State Distribution\u003c/h2\u003e \u003cp\u003eWe first assessed whether brain state reports differed across experiments and conditions. Participants showed the expected distribution of MW ratings, with lowest likelihood of extreme labels (mean likelihood of labels across all experiments: 1\u0026ndash;8.2%; 2\u0026ndash;19.53%; 3\u0026ndash;27.5%; 4\u0026ndash;29.82%; 5\u0026ndash;14.96%). We tested whether participants showed a preference for reporting ON or OFF depending on the stimulus set. We fitted a square function of the individual likelihood distributions and tested for difference in the shift parameter. In all stimulus types and under all instructions, a positive shift parameter was observed indicating a stronger tendency to ON reports (see Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eA; Grating: M\u003csub\u003eb\u003c/sub\u003e = .27, SD\u003csub\u003eb\u003c/sub\u003e = .25, t\u003csub\u003e59\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;8.50, p\u0026lt;.0001, Cohen's d\u0026thinsp;=\u0026thinsp;1.09, 95% CI [0.78, 1.42]; Curvature: M\u003csub\u003eb\u003c/sub\u003e = .33, SD\u003csub\u003eb\u003c/sub\u003e = .21, t\u003csub\u003e57\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;12.09, p\u0026lt;.0001, Cohen\u0026rsquo;s d\u0026thinsp;=\u0026thinsp;1.59, 95% CI [1.20, 1.98]; Color: M\u003csub\u003eb\u003c/sub\u003e = .23, SD\u003csub\u003eb\u003c/sub\u003e = .28, t\u003csub\u003e59\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;6.49, p\u0026lt;.0001, Cohen's d\u0026thinsp;=\u0026thinsp;0.84, 95% CI [0.54, 1.13]; Faces: M\u003csub\u003eb\u003c/sub\u003e = .23, SD\u003csub\u003eb\u003c/sub\u003e = .25, t\u003csub\u003e61\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;7.25, p\u0026lt;.0001, Cohen\u0026rsquo;s d\u0026thinsp;=\u0026thinsp;0.92, 95% CI [0.62, 1.22]; see Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eB; Detection: M\u003csub\u003eb\u003c/sub\u003e = .26, SD\u003csub\u003eb\u003c/sub\u003e = .24, t\u003csub\u003e78\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;9.62, p\u0026lt;.0001, Cohen's d\u0026thinsp;=\u0026thinsp;1.08, 95% CI [0.80, 1.36]; Global: M\u003csub\u003eb\u003c/sub\u003e = .25, SD\u003csub\u003eb\u003c/sub\u003e = .26, t\u003csub\u003e81\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;8.62, p\u0026lt;.0001, Cohen's d\u0026thinsp;=\u0026thinsp;0.95, 95% CI [0.69, 1.21]; Local: M\u003csub\u003eb\u003c/sub\u003e = .30, SD\u003csub\u003eb\u003c/sub\u003e = .25, t\u003csub\u003e78\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;10.39, p\u0026lt;.0001, Cohen's d\u0026thinsp;=\u0026thinsp;1.17, 95% CI [0.88, 1.46]). The 2 x 3 ANOVA did not show a main effect for Visual Stimulus Type (F\u003csub\u003e3,228\u003c/sub\u003e = 1.98, p = .117, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.03, 95% CI [0.01, 0.06]), nor a main effect for Instruction (F\u003csub\u003e2,228\u003c/sub\u003e = .85; p = .426, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.007, 95% CI [0.001, 0.03]), nor an interaction effect (F\u003csub\u003e6,228\u003c/sub\u003e = .69; p = .651, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.02, 95% CI [0.01, 0.04]) indicating that subjects chose ratings comparably across experiments.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eII - Reaction times\u003c/h2\u003e \u003cp\u003eNext, we examined reaction times across the different stimulus types and conditions. We found a significant main effect for the factor Visual Stimulus Type (F\u003csub\u003e3,684\u003c/sub\u003e = 65.86; p \u0026lt; .0001, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.22, 95% CI [0.19, 0.26]; see \u003cb\u003eTable \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e\u003c/b\u003e for an overview of reaction times) with participants exhibiting shorter reaction times in the Grating and Face experiments compared to the Curvature and Color experiments. Additionally, we observed a significant main effect for the factor Instruction (F\u003csub\u003e2,684\u003c/sub\u003e = 37.19, p \u0026lt; .0001, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.09, 95% CI [0.08, 0.12]), attributed to quicker reaction times in Global Discrimination (M\u003csub\u003eglobal\u003c/sub\u003e = 468.6 ms; SD\u003csub\u003eglobal\u003c/sub\u003e = 111.4 ms) compared to Detection (M\u003csub\u003edetection\u003c/sub\u003e = 532.4 ms; SD\u003csub\u003edetection\u003c/sub\u003e = 116.9 ms; t\u003csub\u003e481\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;6.14, p \u0026lt; .0001, Cohen's d\u0026thinsp;=\u0026thinsp;0.56, 95% CI [0.38, 0.74]) and Local Discrimination (M\u003csub\u003elocal\u003c/sub\u003e = 545.5 ms; SD\u003csub\u003elocal\u003c/sub\u003e = 123.0 ms; t\u003csub\u003e481\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;7.21, p \u0026lt; .0001, Cohen's d\u0026thinsp;=\u0026thinsp;0.66, 95% CI [0.47, 0.84]). We also found a significant main effect for Trial Type (F\u003csub\u003e2,684\u003c/sub\u003e = 3.1, p = .0458, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.01, 95% CI [0.004, 0.02]). However, when we compared the trial types separately for each experiment, we only identified a significant main effect in the Color experiments (F\u003csub\u003e2,177\u003c/sub\u003e = 3.06, p = .0490, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.03, 95% CI [0.02, 0.07]). This effect stems from a difference between ON and OFF trials (t\u003csub\u003e118\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;2.65, p = .0089, Cohen\u0026rsquo;s d\u0026thinsp;=\u0026thinsp;0.49, 95% CI [0.12, 0.85]; see \u003cb\u003eTable S2\u003c/b\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eIII - Threshold Estimation\u003c/h2\u003e \u003cp\u003eWhen comparing the threshold values of ON, OFF, and Unlabeled trials, we found a main effect for the factor Instruction (F \u003csub\u003e2,684\u003c/sub\u003e = 28.28, p \u0026lt; .0001, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.08, 95% CI [0.06, 0.09]; see \u003cb\u003eTable S3\u003c/b\u003e for an overview of perceptual thresholds) due to a lower threshold in Global Discrimination (M\u003csub\u003eglobal\u003c/sub\u003e = .44; SD\u003csub\u003eglobal\u003c/sub\u003e = .11) compared to Detection (M\u003csub\u003edetection\u003c/sub\u003e = .52; SD\u003csub\u003edetection\u003c/sub\u003e = .14; t\u003csub\u003e481\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;5.96, p \u0026lt; .0001, Cohen's d\u0026thinsp;=\u0026thinsp;0.54, 95% CI [0.36, 0.72]) and Local Discrimination (M\u003csub\u003elocal\u003c/sub\u003e = .54; SD\u003csub\u003elocal\u003c/sub\u003e = .18; t\u003csub\u003e481\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;6.62, p \u0026lt; .0001, Cohen's d\u0026thinsp;=\u0026thinsp;0.60, 95% CI [0.42, 0.79]). The comparison between Detection and Local Discrimination thresholds revealed only a trend (t\u003csub\u003e472\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;1.7, p = .08, Cohen's d = -0.16, 95% CI [-0.34, 0.02]). Furthermore, there was a main effect for Trial Type (F \u003csub\u003e2,684\u003c/sub\u003e = 29.32; p \u0026lt; .0001, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.08, 95% CI [0.06, 0.10]) due to a lower threshold in ON trials (M\u003csub\u003eON\u003c/sub\u003e = .44; SD\u003csub\u003eON\u003c/sub\u003e = .14) compared to OFF trials (M\u003csub\u003eOFF\u003c/sub\u003e = .52; SD\u003csub\u003eOFF\u003c/sub\u003e = .17; t\u003csub\u003e478\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;5.44, p \u0026lt; .0001, Cohen's d = -0.49, 95% CI [-0.68, -0.32]) and Unlabeled trials (M\u003csub\u003eUnlabeled\u003c/sub\u003e = .53; SD\u003csub\u003eUnlabeled\u003c/sub\u003e = .12; t\u003csub\u003e478\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;7.13, p \u0026lt; .0001, Cohen\u0026rsquo;s d = -0.65, 95% CI [-0.84, -0.47]). The thresholds for OFF trials and Unlabeled trials did not differ (t\u003csub\u003e478\u003c/sub\u003e = .6, p = .52, Cohen\u0026rsquo;s d = -0.06, 95% CI [-0.24, 0.12]). In the next step, we compared the thresholds separately for each experiment using a one-way ANOVA with the factor Trial Type (ON, OFF, Unlabeled). We found a significant main effect for Trial Type in the Grating, Curvature, and Color experiments (Grating: F\u003csub\u003e2,177\u003c/sub\u003e = 9.15, p = .0002, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.09, 95% CI [0.06, 0.15] \u0026ndash; Curvature: F\u003csub\u003e2,171\u003c/sub\u003e = 9.33, p = .0001, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.09, 95% CI [0.06, 0.15] \u0026ndash; Color: F\u003csub\u003e2,177\u003c/sub\u003e = 7.45, p = .0008, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.08, 95% CI [0.05, 0.13]), but only a trend with Face stimuli (Faces: F\u003csub\u003e2,183\u003c/sub\u003e = 2.91, p = .0572, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.03, 95% CI [0.01, 0.07]). Next, we performed post-hoc comparisons of the thresholds between the different trial types. We observed that the thresholds in the Grating, Curvature, and Color experiments differed between ON trials and Unlabeled trials (see \u003cb\u003eTable S4\u003c/b\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). However, this comparison in the Face experiments did not survive the Bonferroni Correction. Furthermore, we found that ON and OFF trial thresholds differed, but only in the Gratings and Curvature experiments. This comparison was not significant at the corrected level in the Color experiment. Additionally, there was no difference in the Face experiment at the normal, not-corrected level. Thus, ON and OFF states lead to differences in visual performance and differences diminish with increasing stimulus complexity. Importantly, no significant difference between OFF and Unlabeled trials was found in any of the four stimulus sets. This suggests that the threshold estimation in standard experimental settings more closely reflects an OFF-state, rather than an ON-state, reinforcing our assumption about the nature of typical perceptual measurements.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eIV - Suprathreshold Accuracy\u003c/h2\u003e \u003cp\u003eFinally, we assessed whether suprathreshold accuracy in Unlabeled trials aligns more with ON or OFF trial accuracy. We found a significant main effect for the factor Visual Stimulus Type (F\u003csub\u003e3,684\u003c/sub\u003e = 19.54, p \u0026lt; .0001, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.08, 95% CI [0.06, 0.10]), as participants exhibited lower accuracy in the Curvature experiment (M\u003csub\u003ecurvature\u003c/sub\u003e = .88; SD\u003csub\u003ecurvature\u003c/sub\u003e = .06) compared to Grating (M\u003csub\u003egrating\u003c/sub\u003e = .92; SD\u003csub\u003egrating =\u003c/sub\u003e .05; t\u003csub\u003e352\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;5.39; p \u0026lt; .0001, Cohen's d\u0026thinsp;=\u0026thinsp;0.57, 95% CI [0.36, 0.79]), Color (M\u003csub\u003ecolor\u003c/sub\u003e = .93; SD\u003csub\u003ecolor\u003c/sub\u003e = .08; t\u003csub\u003e352\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;5.40, p \u0026lt; .0001, Cohen's d = -0.58, 95% CI [-0.79, -0.36]), and Faces (M\u003csub\u003efaces\u003c/sub\u003e = .91; SD\u003csub\u003efaces\u003c/sub\u003e = .06; t\u003csub\u003e358\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;3.68, p = .0003, Cohen\u0026rsquo;s d = -0.39, 95% CI [-0.59, -0.18]). None of the pairwise comparisons of Grating, Color, and Face experiments showed a significant difference (p\u003csub\u003eBonferroni\u003c/sub\u003e = .0083; all p values \u0026gt; .01). We also found a main effect for Trial Type (F\u003csub\u003e2,684\u003c/sub\u003e = 33.17, p \u0026lt; .0001, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.09, 95% CI [0.07, 0.11]), and an interaction effect for the factors Visual Stimulus Type and Trial Type (F\u003csub\u003e6,684\u003c/sub\u003e = 2.37, p = .0283, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.02, 95% CI [0.01, 0.03]). When we compared the accuracy values between trial types separately for each experiment, we found a significant main effect for Trial Type in the Grating, Curvature, and Color experiments (Grating: F\u003csub\u003e2,177\u003c/sub\u003e = 23.54, p \u0026lt; .0001, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.21, 95% CI [0.16, 0.28] \u0026ndash; Curvature: F\u003csub\u003e2,171\u003c/sub\u003e = 7.14, p = .0011, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.08, 95% CI [0.05, 0.13] \u0026ndash; Color: F\u003csub\u003e2,177\u003c/sub\u003e = 3.61, p = .0289, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.04, 95% CI [0.02, 0.08]). However, in the Face experiments, only a trend was observed (Faces: F\u003csub\u003e2,183\u003c/sub\u003e = 2.89, p = .0581, \u003cem\u003eη\u003c/em\u003e\u003csub\u003e\u003cem\u003ep\u003c/em\u003e\u003c/sub\u003e\u003cem\u003e\u0026sup2;\u003c/em\u003e = 0.03, 95% CI [0.01, 0.07]). Next, separately for each experiment we performed post-hoc comparisons of the suprathreshold accuracy between the different trial types. We observed that the thresholds in the Grating, Curvature, and Color experiments differed between ON trials and Unlabeled trials (see \u003cb\u003eTable S5\u003c/b\u003e for suprathreshold values, \u003cb\u003eTable S6\u003c/b\u003e for post hoc comparisons, and Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB). This comparison in the Face experiment did not survive the Bonferroni correction. Post hoc comparisons between ON and OFF trials differed only in the Gratings and Curvature experiments (see \u003cb\u003eTable S6\u003c/b\u003e). This comparison was not significant at the corrected level in the Color experiment. Additionally, there was no difference in the Face experiment at the normal, not-corrected level. The BF\u003csub\u003e10\u003c/sub\u003e-values for the comparison between Unlabeled and OFF trials (see \u003cb\u003eTable S6\u003c/b\u003e) provide substantial evidence in favor of the null hypothesis in the Curvature and Faces experiments, indicating that accuracy in Unlabeled trials closely resembles that of OFF trials.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eWe investigated whether threshold estimations in typical visual search paradigms are influenced by attentional brain states. To test this, we examined the influence of MW on the detection and discrimination of targets in a range of visual search paradigms. Given MW occurs in up to 50% of the waking time, we explored if MW epochs altered visual detection performance across an experimental session. We found that accuracy in the Unlabeled test trials resembled the OFF-state more closely than the ON-state. This discrepancy challenges the conventional assumption that subjects are fully attentive across an experimental setting and that measuring the visual threshold adequately captures visual perception capacity. The alteration in perception manifested in two main ways: a shift in thresholds and reduced perception in suprathreshold conditions. Importantly, we observed that MW did not uniformly deteriorate visual perception. Instead, it primarily impacted the perception of simple stimuli and the influence of MW diminished with increasing complexity.\u003c/p\u003e \u003cp\u003eWe found no differences in reaction times during MW episodes compared to ON trials, contrary to the slowing observed in most studies \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e with pop-out target stimuli \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. We observed a trend towards slower reaction times in ON trials aligning with Andrillon et al \u003csup\u003e10\u003c/sup\u003e, who distinguished between MW and mind blanking. These authors found faster reaction times during MW compared to ON task trials suggesting increased impulsivity during MW. Motor performance tends to become more automatic and less precise during MW episodes \u003csup\u003e\u003cspan additionalcitationids=\"CR12 CR13 CR14\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Unlike engaged task states, MW is often associated with faster reaction times and increased error rates \u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e as reflected by our results. These variations in reaction times likely depend on the specific task demands. In our experiments, participants were instructed to respond quickly and accurately, possibly reflecting a trade-off where MW prioritizes speed over accuracy, while ON-task states emphasize accuracy. This acceleration in reaction times suggests that MW not only degrades performance but also alters the dynamic control of motor outputs (Kam et al., 2012).\u003c/p\u003e \u003cp\u003eThe prevailing notion has been that MW involves a general decoupling from the external world to shield internal thoughts from external distractions. However, our data challenge this assumption for two reasons. First, in our study, participants exhibited altered performance during MW but maintained a clear supra-threshold response as reflected by high rates of accuracy. This finding suggests they remained connected to their surroundings rather than functioning at a random level as it would be predicted with the decoupling hypothesis. Second, participants displayed a threshold shift and reduced accuracy in supra-threshold trials only observed with simple stimuli. As stimulus complexity increased, the performance impairment associated with MW diminished. This suggests that attention modulation has a greater impact when processing simpler stimuli \u003csup\u003e\u003cspan additionalcitationids=\"CR3 CR4\" citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. This finding suggests that the influence of MW is best understood as a gradual shift in perception, rather than an all-or-nothing effect. This observed MW gradient aligns with previous research. Classic Letter Sustained Attention to Response Task (SART) paradigms (similar to our gratings or curvature stimuli) reveal task-unrelated thought effects, leading to more commission errors during OFF-task trials \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. This effect is not observed in Face SART paradigms \u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, where face detection is affected by MW. In sum, MW does not fully decouple perception from external stimulation.\u003c/p\u003e \u003cp\u003eOur study highlights a methodological consideration. Numerous studies have focused on distinguishing between ON- and OFF-states. However, no study has directly compared the ON-state with an unlabeled state. Instead, it has been common practice to not only use the trials immediately preceding the focused inquiry but also to average over a variable number of preceding trials. Consequently, parameters in previous studies were often computed by averaging data from several preceding trials, assuming implicitly that participants remained in the same brain state. This averaging was achieved through methods utilizing a fixed number of preceding trials \u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e or specifying a time duration, such as several seconds \u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Another approach, as seen in Arnau et al. \u003csup\u003e22\u003c/sup\u003e, involves selecting different numbers of trials preceding ON- and OFF-ratings. In this method, trials were categorized as \"mind wandering\" when participants reported such experiences, along with the two preceding trials, while the remaining trials were classified as \"on-task\" trials. However, this approach is also susceptible to blending different brain states. Here, we demonstrate that this approach, particularly when considering different preceding trials, poses challenges, especially in the \"ON\" condition.\u003c/p\u003e \u003cp\u003eWhy did we observe that global performance aligns more with OFF-task processing? In fact, participants more frequently report being \"ON the task\" than \"OFF the task\" which should also be represented in the unlabeled trial. A stronger resemblance of unlabeled trials with OFF trials can only occur if the participants are in the OFF state more frequently than they report. This inconsistency may be attributed to participants more often selecting the MID category when they are uncertain about their brain state. This may occur for example due to social desirability biases which refers to the tendency of individuals to respond to self-report items in a manner that presents themselves in a more socially favorable light according to prevailing societal norms and standards. Here we obtained focus queries sporadically and randomly to minimize the risk of reducing the likelihood of MW \u003csup\u003e7\u003c/sup\u003e. Conversely, this implies that the absence of focus queries may lead to even more instances of MW. We posit that this affects performance, in the sense that without interruption, more MW occurs deteriorating performance.\u003c/p\u003e \u003cp\u003eIn conclusion, our study sheds light on the complex interplay between brain states, MW, and visual perception. We emphasize the importance of considering the task type, stimulus complexity, and the dynamic nature of brain states when investigating MW's effects. These findings have implications for understanding how MW impacts cognitive processes and the need to refine experimental methodologies to accurately capture its influence on task performance.\u003c/p\u003e"},{"header":"Declarations","content":" \u003ch2\u003eCompeting Interest Statement:\u003c/h2\u003e \u003cp\u003eAll authors declare no competing interests.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor Contributions:\u003c/h2\u003e \u003cp\u003eS.D. conceived and designed the experiment. A.S., T.W., and P.S. collected the data. A.S., T.W., P.S., C.R., and S.D. analyzed the data, A.S., T.W., P.S., C.R., R.T.K., and S.D. interpreted the data. A.S., T.W., P.S., C.R., R.T.K., and S.D. wrote the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgments:\u003c/h2\u003e \u003cp\u003eThis work was supported by the Deutsche Forschungsgemeinschaft (DFG; Grant SFB 1436 A03). The authors thank Christiane Petzold and Nathalie Vogt for assisting in the data collection.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSmallwood, J. Mind-wandering While Reading: Attentional Decoupling, Mindless Reading and the Cascade Model of Inattention. \u003cem\u003eLinguistics and Language Compass\u003c/em\u003e 5, 63\u0026ndash;77 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuck, S. J. \u0026amp; Hillyard, S. A. Spatial filtering during visual search: Evidence from human electrophysiology. \u003cem\u003eJ. Exp. Psychol. Hum. Percept. Perform.\u003c/em\u003e 20, 1000\u0026ndash;1014 (1994).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartinez-Trujillo, J. C. \u0026amp; Treue, S. Feature-Based Attention Increases the Selectivity of Population Responses in Primate Visual Cortex.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePosner, M. I. \u0026amp; Gilbert, C. D. Attention and primary visual cortex. \u003cem\u003eProc. Natl. Acad. Sci. U. S. A.\u003c/em\u003e 96, 2585\u0026ndash;2587 (1999).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSteven J. Luck \u0026amp; Edward K. Vogel. The capacity of visual working memory for features and conjunctions. \u003cem\u003eNature\u003c/em\u003e 390, 279\u0026ndash;281 (1997).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrainard, D. H. The Psychophysics Toolbox. \u003cem\u003eSpat. Vis.\u003c/em\u003e 10, 433\u0026ndash;436 (1997).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeli, P., Carriere, J. S. A., Levene, M. \u0026amp; Smilek, D. How few and far between? Examining the effects of probe rate on self-reported mind wandering. \u003cem\u003eFront. Psychol.\u003c/em\u003e 4, 430 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong, H. W., Mills, C., Knight, R. T. \u0026amp; Kam, J. W. Y. Detection of mind wandering using EEG: Within and across individuals. \u003cem\u003ePLoS One\u003c/em\u003e 16, (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWienke, C. \u003cem\u003eet al.\u003c/em\u003e Mind-wandering Is Accompanied by Both Local Sleep and Enhanced Processes of Spatial Attention Allocation. \u003cem\u003eCereb. Cortex Commun.\u003c/em\u003e 2, (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndrillon, T., Burns, A., Mackay, T., Windt, J. \u0026amp; Tsuchiya, N. Predicting lapses of attention with sleep-like slow waves. \u003cem\u003eNature Communications 2021 12:1\u003c/em\u003e 12, 1\u0026ndash;12 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarriere, J. S. A., Cheyne, J. A. \u0026amp; Smilek, D. Everyday attention lapses and memory failures: The affective consequences of mindlessness. \u003cem\u003eConscious. Cogn.\u003c/em\u003e 17, 835\u0026ndash;847 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheyne, J. A., Carriere, J. S. A. \u0026amp; Smilek, D. Absent-mindedness: Lapses of conscious awareness and everyday cognitive failures. \u003cem\u003eConscious. Cogn.\u003c/em\u003e 15, 578\u0026ndash;592 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReichle, E. D., Reineberg, A. E. \u0026amp; Schooler, J. W. Eye movements during mindless reading. \u003cem\u003ePsychol. Sci.\u003c/em\u003e 21, 1300\u0026ndash;1310 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchooler, Jonathan W.; Reichle, Erik D.; Halpern, D. V. Zoning Out while Reading: Evidence for Dissociations between Experience and Metaconsciousness. \u003cem\u003eThinking and seeing: Visual metacognition in ad\u003c/em\u003e 203\u0026ndash;226 Preprint at (2004).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeissman, D. H., Roberts, K. C., Visscher, K. M. \u0026amp; Woldorff, M. G. The neural bases of momentary lapses in attention. \u003cem\u003eNat. Neurosci.\u003c/em\u003e 9, 971\u0026ndash;978 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFranklin, M. S., Smallwood, J. \u0026amp; Schooler, J. W. Catching the mind in flight: Using behavioral indices to detect mindless reading in real time. \u003cem\u003ePsychon. Bull. Rev.\u003c/em\u003e 18, 992\u0026ndash;997 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmallwood, J. \u003cem\u003eet al.\u003c/em\u003e Subjective experience and the attentional lapse: Task engagement and disengagement during sustained attention. \u003cem\u003eConscious. Cogn.\u003c/em\u003e 13, 657\u0026ndash;690 (2004).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKam, J. W. Y. \u003cem\u003eet al.\u003c/em\u003e Mind wandering and motor control: off-task thinking disrupts the online adjustment of behavior. 6, 1\u0026ndash;9 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDenkova, E., Brudner, E. G., Zayan, K., Dunn, J. \u0026amp; Jha, A. P. Attenuated face processing during mind wandering. \u003cem\u003eJ. Cogn. Neurosci.\u003c/em\u003e 30, 1691\u0026ndash;1703 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJin, C. Y., Borst, J. P. \u0026amp; van Vugt, M. K. Decoding study-independent mind-wandering from EEG using convolutional neural networks. \u003cem\u003eJ. Neural Eng.\u003c/em\u003e 20, (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWamsley, E. J. \u0026amp; Summer, T. Spontaneous entry into an \u0026ldquo;offline\u0026rdquo; state during wakefulness: A mechanism of memory consolidation? \u003cem\u003eJ. Cogn. Neurosci.\u003c/em\u003e 32, 1714\u0026ndash;1734 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArnau, S. \u003cem\u003eet al.\u003c/em\u003e Inter-trial alpha power indicates mind wandering. \u003cem\u003ePsychophysiology\u003c/em\u003e 57, (2020).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-9412640/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9412640/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eVisual threshold estimation typically requires data from thousands of trials presented across extended periods of time. However, human attention fluctuates over time between focused and inattentive states, such as mind-wandering (MW). Failing to account for MW in perceptual research may produce systematically lower estimates of perceptual performance. We assessed how brief changes in brain state affect visual performance by investigating the impact of MW on visual target detection and discrimination in twelve behavioral tasks involving 239 subjects (156 female). In all tasks, subjects reported their task engagement (ON vs. OFF) in 20% of trials. We explored whether in the remaining 80% of unlabeled trials performance aligns with ON or OFF trials. MW led to threshold detection shifts and reduced perception in OFF compared to ON trials. MW primarily affected the perception of simple stimuli like gratings and curvature stimuli, with its influence diminishing as stimulus complexity increased. Importantly, visual performance in unlabeled trials aligned with the OFF-state. Our results show that mind wandering induces fluctuations in attentional capacity that render threshold measurements substantially less reliable without correction, underscoring the need to account for attentional state to obtain valid estimates of perceptual capacity.\u003c/p\u003e","manuscriptTitle":"Visual Thresholds Cannot Be Reliably Measured Without Controlling for Mind-Wandering","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-07 12:11:09","doi":"10.21203/rs.3.rs-9412640/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"communications-psychology","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"commspsychol","sideBox":"Learn more about [Communications Psychology](http://www.nature.com/commspsychol/)","snPcode":"44271","submissionUrl":"https://mts-commspsychol.nature.com/cgi-bin/main.plex","title":"Communications Psychology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Communications Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"865676ee-517a-4ab6-9ff9-04798917a20e","owner":[],"postedDate":"May 7th, 2026","published":true,"recentEditorialEvents":[{"type":"reviewerAgreed","content":"This content is not available.","date":"2026-05-13T02:53:07+00:00","index":2,"fulltext":"This content is not available."},{"type":"editorInvitedReview","content":"This content is not available.","date":"2026-05-11T13:48:21+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2026-05-11T02:05:53+00:00","index":1,"fulltext":"This content is not available."}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":67252104,"name":"Biological sciences/Neuroscience/Visual system"},{"id":67252105,"name":"Biological sciences/Neuroscience/Cognitive neuroscience/Perception"},{"id":67252106,"name":"Biological sciences/Neuroscience/Cognitive neuroscience/Attention"},{"id":67252107,"name":"Biological sciences/Neuroscience/Sensory processing"},{"id":67252108,"name":"Biological sciences/Psychology/Human behaviour"}],"tags":[],"updatedAt":"2026-05-07T12:11:09+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-07 12:11:09","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9412640","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9412640","identity":"rs-9412640","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}