Electrophysiological correlates of detection and identification awareness for digits and letters

Electrophysiological correlates of detection and identification awareness for digits and letters | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 2 July 2025 V1 Latest version Share on Electrophysiological correlates of detection and identification awareness for digits and letters Authors : Stefan Wiens 0000-0003-4531-4313 [email protected] , Mingailė Greičiūtė , Billy Gerdfeldter 0000-0002-3222-8056 , and Annika Andersson Authors Info & Affiliations https://doi.org/10.22541/au.175143836.64282168/v1 Published Neuropsychologia Version of record Peer review timeline 251 views 115 downloads Contents Abstract Introduction Experiment 1 - Method Experiment 1 - Results Figure 2 Figure 3 Experiment 1 - Discussion Experiment 2 - Method Experiment 2 - Results Experiment 2 - Discussion Figure 8 General discussion Declarations References Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract A central feature of consciousness is the association between external events and subjective experiences of content. These experiences range from low level (detection) to high level (identification). For example, a visual experience may range from seeing something on a computer screen (detection) to seeing the digit 3 (identification). In research, neural processes that correlate with these experiences are called neural correlates of consciousness (NCCs). In vision, a prominent NCC is visual awareness negativity (VAN) that is derived from event-related potentials and occurs about 200 ms after stimulus onset. Previous research suggests that VAN is more sensitive to low-level experiences (detection) than high-level experiences (identification). Because previous results are limited, two preregistered experiments were conducted. Experiment 1 ( N = 30) showed that VAN was similarly sensitive to detection and identification. This was shown for a digit task and for a letter task. Experiment 2 ( N = 28) used the same digit task as in Experiment 1 with two stimulus sizes. Results found identification VAN for both digit sizes, and VAN was unaffected by stimulus size. These results confirm the sensitivity of VAN to both low-level and high-level experiences, consistent with recurrent processing theory. However, results emphasize the limited specificity of VAN in separating between low-level and high-level experiences, suggested by the similarity of VAN in both conditions. Introduction A central feature of consciousness is the association between external events and subjective experiences of content. For example, when something flashes on a computer screen, you may experience only that something was there without experiencing what it was. These experiences are also referred to as qualia or phenomenal consciousness (Block, 2005). Electroencephalography has been widely used to identify potential neural correlates of consciousness (NCC): neural activity that is consistently correlated with and sufficient for particular experiences of content (Frith et al., 1999; Koch et al., 2016). Because of its excellent time resolution (Biasiucci et al., 2019; Luck, 2014), electroencephalography has been used to characterize NCC in time. Two event-related potentials (ERPs) have received particular interest (Förster et al., 2020; Koivisto & Revonsuo, 2010). In vision, an early response at about 200 ms after stimulus onset is the visual awareness negativity (VAN), and a late response 300 ms after stimulus onset is the late positivity (LP). With regard to LP, research in the last decades has provided convincing evidence that it is not an NCC of phenomenal consciousness but reflects processes associated with task relevance, evidence accumulation, and response preparation (Cohen et al., 2020, 2024; Dellert et al., 2021, 2022; Dembski et al., 2021; Förster et al., 2020; Koivisto & Revonsuo, 2010; Kronemer et al., 2022; O’Connell & Kelly, 2021; Pitts et al., 2012, 2014; Schlossmacher et al., 2020). Accordingly, LP may correlate with consciousness, but it is not a true NCC. If it were a true NCC, it would correlate closely with the experience per se (Aru et al., 2012; de Graaf et al., 2012; Dembski et al., 2021). VAN has been observed in vision, but similar ERP responses have been found in other modalities such as hearing (Eklund & Wiens, 2019; Gerdfeldter et al., 2024) and touch (Auksztulewicz & Blankenburg, 2013; Schröder et al., 2021). Because these early ERPs are generated in sensory cortices (Meyer, 2011; Snyder et al., 2015), they are believed to capture the same process in different sensory modalities (Dembski et al., 2021). These findings are consistent with recurrent processing theory (V. A. F. Lamme, 2006, 2010). This theory emphasizes that experience is mediated by early, recurrent processing in sensory areas, and VAN has been proposed as an index of this processing in vision (Eklund & Wiens, 2018; Förster et al., 2020; Koivisto & Revonsuo, 2010; V. A. F. Lamme, 2018). Although it is debated whether these early ERPs may be confounded by preconscious processes such as attention (Bola & Doradzinska,´ 2021), VAN is currently considered the most promising ERP candidate for a true NCC in vision (Dembski et al., 2021; Förster et al., 2020). Even if VAN is an NCC, it is important to resolve its sensitivity and specificity for phenomenal consciousness (Bachmann & Aru, 2023; Koivisto & Revonsuo, 2010). For example, when looking at a screen, is VAN sensitive to the experience of something on the screen (low-level experience) rather than the experience of what it is (high-level experience)? If it is sensitive only to low-level experiences, it would be a useful NCC for these experiences, but it would not be useful in measuring high-level experiences. Thus, sensitivity and specificity with regard to phenomenal consciousness need to be resolved for VAN. Recently, Koivisto et al. (2017) proposed that VAN is sensitive to low-level experiences rather than high-level experiences. Low-level experiences include elementary experiences of something on the screen ( detection awareness ). High-level experiences include experiences of the identity of a stimulus ( identification awareness ). In the study (Koivisto et al., 2017), subjects were shown single digits (3, 4, 6, or 7) in the center of the screen while EEG was recorded. Subjects performed two tasks in separate blocks. During the detection task, they had to rate their experience of whether something was presented. During the identification task, they had to discriminate whether the digit was below 5 or above 5. Awareness was measured with a modified version of the perceptual awareness scale (PAS, Ramsøy & Overgaard, 2004; Sandberg & Overgaard, 2015). That is, experiencing the identity of the stimulus was added to the PAS: Nothing (PAS1) referred to ”I did not see any stimulus”, something (PAS2) referred to ”I saw something (but could not identify the stimulus)”, almost clear (PAS3) referred to ”I saw the stimulus almost clearly (and could identify it),” and clear (PAS4) referred to ”I saw the stimulus clearly (and could identify it).” Before the detection task, stimulus contrast and duration were adjusted to target the detection threshold : At this stimulus level, half of the trials are rated as nothing (unaware) and half as something (aware). In the analyses, the difference of something minus nothing trials was used to measure VAN at the detection threshold ( detection VAN ). Before the separate identification task, stimulus parameters were adjusted to target the identification threshold : At this stimulus level, half of the trials are rated as something (unaware) and half as almost clear (aware). In the analyses, the difference of almost clear minus something trials was used to measure VAN at the identification threshold ( identification VAN ). The data of the study (Koivisto et al., 2017) were analyzed with null hypothesis significance testing (NHST, Wasserstein & Lazar, 2016). Results showed a significant interaction between threshold and awareness: Identification VAN was smaller (less negative) than detection VAN. Follow-up analyses suggested a significant detection VAN but a non-significant identification VAN. These findings are consistent with the notion that VAN is sensitive to low-level experiences (detection) rather than high-level experiences (identification), suggesting that VAN may be useful to measure detection experiences but not identification experiences. Because identification of a digit is a relatively high-level process (Bachmann & Aru, 2023), we previously examined whether detection VAN and identification VAN differ for a more basic stimulus (Wiens et al., 2023). The stimulus was a ring (annulus) that was shown in the middle of the screen and comprised a Gabor patch tilting either left or right (±45 ◦ ). In this study, identification referred to experiencing the Gabor ring as tilting to the left or to the right. The study design improved on the design of the earlier study (Koivisto et al., 2017). The main improvement was that instead of testing detection and identification thresholds separately, both thresholds were assessed concurrently. This avoided potential influences from differences in task instructions. On each trial, subjects rated their awareness as nothing (PAS1), something (PAS2), or identify (PAS3); identify referred to experiencing the orientation of the grating (left or right). Afterward, they discriminated whether the ring tilted left or right by choosing one of two circles (tilting left or right). Unbeknownst to the subjects, two separate staircases adjusted stimulus opacity continuously to either detection threshold or identification threshold. For the detection threshold, nothing represented unaware and something represented aware, and for the identification threshold, something represented unaware and identify represented aware. Results of Bayesian analyses (i.e., Bayes Factors) provided extreme evidence ( BF10 > 100) that identification VAN was smaller (less negative) than detection VAN. Also, there was only ambiguous evidence (1 < BF10 < 3) for identification VAN; that is, results did not provide enough support for identification VAN. Thus, these findings fit nicely into the proposal by Koivisto et al. (2017): VAN is sensitive to low-level experiences (detection) rather than high-level experiences (identification). Notably, results of several older studies appear to be inconsistent with this conclusion, as these studies reported VAN in the context of high-level experiences (Koivisto & Revonsuo, 2008; Koivisto et al., 2005, 2009; Wilenius & Revonsuo, 2007). For example, VAN was obtained for low-contrast line drawings when subjects reported to be very or absolutely sure about the content rather than unsure (Wilenius & Revonsuo, 2007). In these studies, however, it is unclear whether unaware trials may have included nothing as well as something experiences. If so, the difference between aware (identify) trials and unaware (nothing and something) trials reflects a combination of detection awareness and identification awareness. To illustrate with a hypothetical example that shows only evidence for detection VAN: Mean amplitudes are 0 µV for nothing, −1 µV for something, and −1 µV for identify responses. Aware (identify) trials minus unaware (nothing and something) trials would be −1 minus (−1 + 0)/2 = −0.5. Importantly, this would not be identification VAN but a confound from detection VAN. Similarly, if aware trials included something experiences, any difference between aware (identify and something) trials minus unaware (nothing) trials would be a combination of detection and identification VAN. To separate detection VAN (nothing vs something) from identification VAN (something vs identify), a critical design feature is to include the required response alternatives to separate between nothing, something, and identify experiences (Koivisto et al., 2017; Wiens et al., 2023). Whereas Koivisto et al. (2017) referred to low-level experiences such as detection and high-level experiences such as identification, this approach differs from that of levels of processing (Jimenez et al., 2020; Windey et al., 2014). In levels of processing, low-level experiences are believed to be gradual whereas high-level experiences are considered to be dichotomous. In support, ERP studies have examined claims of levels of processing (Derda et al., 2019; Jimenez et al., 2018, 2021). In a typical study, the stimuli combine low-level and high-level features (e.g., digits in different colors). In separate tasks, subjects are instructed to rate their awareness of low-level features (e.g., color discrimination) or high-level features (e.g., digit identification), using the original PAS (Ramsøy & Overgaard, 2004; Sandberg & Overgaard, 2015). Afterward, the distribution of PAS is examined for different combinations of stimulus and task. Critically, the notion of levels of processing does not map directly onto the proposal by Koivisto et al. (2017) because it is possible to assess both detection awareness and identification awareness separately for low-level and high-level features (Förster et al., 2020; Wiens et al., 2023). That is, detection awareness would refer to the simple experience of something on the screen whereas identification awareness would refer to the color of the digit (low-level feature) or the identity of the digit (high-level feature). When considering previous studies (Derda et al., 2019; Jimenez et al., 2018, 2021; Tagliabue et al., 2016) in terms of detection and identification awareness (Koivisto et al., 2017), it is unclear how to interpret the results (for details, see Wiens et al., 2023). The main issue is that most studies used the original PAS levels (nothing, weak, almost clear, and clear), but it is unclear how these map onto detection awareness and identification awareness. For example, subjects may differ in whether almost clear or clear captures identification awareness. Therefore, an important design feature is to provide clear response alternatives (nothing, something, identify) to distinguish between detection awareness and identification awareness (Koivisto et al., 2017; Wiens et al., 2023). In sum, two previous studies provide support for the claim that VAN is more sensitive to low-level (detection) experiences than high-level (identification) experiences (Koivisto et al., 2017; Wiens et al., 2023). Identification experiences referred to experiencing the orientation of a Gabor ring (Wiens et al., 2023) or experiencing the identity of digits (Koivisto et al., 2017). Although these previous results are promising, they are limited. First, experiencing the orientation of a Gabor ring is a rather atypical task. In research on levels of processing, a more common task is to identify individual digits or letters. Second, the study by Koivisto et al. (2017) used a typical digit task, but it is unclear whether differences between detection VAN and identification VAN may have been confounded by the fact that they were assessed separately. Because a supplementary analysis (Koivisto et al., 2017) showed that the detection LP differed between the detection task and the identification task, a confound by task differences cannot be ruled out (for details, see Wiens et al., 2023). Because previous results are limited with regard to whether VAN is sensitive mainly to low-level (detection) experiences rather than high-level (identification) experiences (Koivisto et al., 2017; Wiens et al., 2023), the goal of the present research was to measure and compare detection VAN and identification VAN concurrently. For convenience, we refer to low-level experiences as detection and high-level experiences as identification, even though other conceptualizations are possible depending on stimulus and task. For example, experiencing the location of a stimulus on the screen may be an alternative low-level experience, and experiencing whether a stimulus shows an object or an animal may be an alternative high-level experience. We conducted two experiments that were preregistered (Experiment 1 and Experiment 2). Experiment 1 was a between-subjects design: Subjects were assigned to either a digit task or a letter task. To replicate previous results (Koivisto et al., 2017), the digit task included the same stimuli (3, 4, 6, & 7) shown under similar conditions, a similar PAS rating scale, and the same discrimination task (below 5 or above 5). To extend previous results, the letter task included digits and letters and required subjects to discriminate whether they experienced a digit or letter. In line with previous results, we hypothesized that for both tasks, identification VAN would be smaller (less negative) than detection VAN. Furthermore, we predicted that there would be an identification VAN in each task. Although research has shown that LP is no longer considered a true NCC (as explained earlier), we included LP in the analyses because of its relevance to post-perceptual aspects of consciousness (Dembski et al., 2021; Förster et al., 2020; Koivisto & Revonsuo, 2010). Experiment 1 - Method The design of Experiment 1 was modeled after that of our previous study (Wiens et al., 2023). Experiment 1 was preregistered at OSF. Any deviations from the preregistration are noted below. All material, data, and scripts are shared via a public university repository (Wiens, 2025) to adhere to the recommendations of open science (Munafò et al., 2017). Participants Subjects were recruited through advertisements at the local university (physical and online). According to the ad, subjects should be between ages 18 and 40, with normal or corrected-to-normal vision, and without mental health disorders or neurological history. However, we did not screen subjects for these criteria except for age. In accordance with the Declaration of Helsinki, subjects provided their written informed consent (this included their consent that their data would be shared anonymized). The study was approved by the Swedish Ethical Review Authority (2022-06409-01). The experiment lasted one hour, and subjects received a 200-SEK gift voucher. We collected data from 31 subjects. Upon arrival at the lab, subjects were pseudo-randomly assigned to either a digit task or a letter task . Because of too few trials (< 25 trials) in the EEG analyses, one subject was excluded completely. One subject was excluded from the digit-detect condition, and one subject from the letter-detect condition. However, these two subjects were retained in the identify conditions. Thus, the final sample comprised 15 subjects in each task for a total of 30 subjects (10 males; 29 right-handed; age: M = 28, SD = 6). Stimuli and apparatus The visual stimuli were black digits (3, 4, 6, 7, and 8) and letters (A, E, and H) in a digital font on a gray background. Stimuli had a width of 5 to 7 mm and a height of 7 to 9 mm. Stimulus position was jittered so that the stimuli were shown within a horizontal range of 13 mm and a vertical range of 15 mm around fixation. Because viewing distance was 57 cm, 10 mm correspond to 1°. The stimuli were shown on a 24-inch BenQ XL2430T monitor at 144 Hz with a resolution of 1,920 by 1,080. The experiment was performed in a dimly lit room. PsychoPy (Peirce et al., 2019) was used to generate visual stimuli and to run the experiment. Procedure Participants performed either a digit task or a letter task. In the digit task, stimuli were the digits 3, 4, 6, and 7. On each trial, one of the digits was shown. Subjects rated their visual experience of the digit and then discriminated whether the digit was below 5 or above 5. In the letter task, the digits 4, 6, and 8, and the letters A, E, and H were used. On each trial, either a digit or letter was shown, and subjects rated their visual experience of the stimulus and then discriminated whether the stimulus was a digit or letter. For both tasks, a trial started with a small dot in the center of the screen. After 800 to 1000 ms (randomly in steps of 50 ms), the visual stimulus (digit or letter) was presented in the center of the screen for 21 ms (3 visual frames at a refresh rate of 144 Hz). Afterward, the screen was blank for 600 to 800 ms (randomly in steps of 50 ms). Subjects rated their visual experience of the stimulus on the perceptual awareness scale (PAS; Ramsøy & Overgaard, 2004; Sandberg & Overgaard, 2015) that was slightly modified (Koivisto et al., 2017; Wiens et al., 2023). Response options were nothing (PAS1), something (PAS2), and identify (PAS3). Nothing referred to no experience of the visual stimulus, not even a faint sensation. Something referred to a weak sensation of something (without seeing its identity). Identify referred to the experience of the identity of the stimulus. These labels were explained thoroughly to minimize potential biases from individual differences in response criteria (Fahrenfort et al., 2024). After the PAS ratings, subjects rated the stimulus in terms of below 5 or above 5 (digit task) or in terms of digit or letter (letter task). Subjects used a keyboard to respond. Awareness was rated with one hand, and stimulus discrimination was indicated with the other hand. Subjects rated their awareness with either left hand (keys S, D, and F) or right hand (keys J, K, and L). Subjects indicated their stimulus discrimination with either left hand (keys D and F) or right hand (keys J and K). On each trial, subjects had to finalize each choice by pressing the space bar. The initial mapping of the response hands to the questions was pseudorandomized. After a block of trials, response hands were reversed to minimize EEG confounds from consistent key-response mappings. When subjects were required to respond, the relevant key-response mapping was shown on the screen. In each task, stimulus opacity was continuously adjusted over trials in two separate but interleaved staircases (as in Wiens et al., 2023). One staircase adjusted opacity to the detection threshold , and the other staircase adjusted opacity to the identification threshold . The detection threshold is the opacity level at which 50% of trials are rated as something (and 50% as nothing). The identification threshold is the opacity level at which 50% of trials are rated as identify (and 50% as something). For each threshold, opacity was adjusted if the subject gave the same response in two consecutive trials (with the same opacity level). Trials for both staircases were shown in pseudo-randomized order, and on each trial, there was a 0.05 probability for a catch trial (i.e., transparent stimulus). Data collection occurred in blocks of approximately 5 min. For each subject, as many trials as possible were recorded within a one-hour session. On average, subjects in the digit task completed 421 trials ( SD = 57), and subjects in the letter task completed 417 trials ( SD = 50). Before the main task, subjects were presented with sample trials and asked to describe their experience. The instructions emphasized that the primary task was to rate their conscious experience accurately whereas identifying the digit or letter was secondary. They were instructed that when rating identify (PAS3), their subsequent response about the digit or letter should be correct. When rating nothing (PAS1) or something (PAS2), they should not worry about whether their subsequent response was correct. EEG recording EEG data were recorded from 64 electrodes at standard positions of the international 10–20 system with an Active Two BioSemi system (BioSemi, Amsterdam, Netherlands). Standard BioSemi CMS (Common Mode Sense) and DRL (Driven Right Leg) electrodes were used to establish a referencing and create a feedback circuit. An EEG cap (Electro-Cap International, Eaton, OH) was used to position these electrodes together with two system-specific electrodes. Data were sampled at 512 Hz and filtered with a hardware low-pass filter at 102.4 Hz. EEG preprocessing EEG data were preprocessed in MNE-python (Gramfort et al., 2013, 2014). A 1-Hz high-pass filter and a 40-Hz low-pass filter were applied to the raw data. Electrodes with excessive noise or no activity were interpolated. Stimulus onsets and offsets were detected with a photodiode (in the corner of the screen). For each stimulus, data were extracted between 500 ms before and 1000 ms after stimulus onset. Independent component analysis (fastica) was conducted, and eye blink components were identified on the basis of their topography. Then, the raw data were read in again, bad channels were interpolated, data were average referenced, eye blink components were excluded, and a 0.1 high-pass filter and a 30-Hz low-pass filter were applied. Epochs were extracted from 100 ms before to 800 ms after stimulus onset. Each epoch was baseline corrected to the mean of the 100-ms interval before stimulus onset (−100 to 0 ms). The epoched data were downsampled to 256 Hz. For each participant, maximum amplitude ranges were extracted for individual epochs, and the distribution of these amplitude ranges was inspected for outliers. The exclusion thresholds were set for each individual because subjects showed substantial variability in these amplitude ranges. Critically, to avoid bias, inspection of trials was blinded to PAS rating and discrimination response. VAN-relevant and LP-relevant intervals and electrodes were preregistered. As in our previous study (Wiens et al., 2023), VAN-relevant mean amplitudes were computed between 180 and 280 ms after stimulus onset across electrodes O1, O2, PO3, PO4, PO7, and PO8. LP-relevant mean amplitudes were computed between 350 and 550 ms after stimulus onset across electrodes Pz, P1, P2, CPz, CP1, and CP2. For each subject and trial, VAN-relevant mean amplitudes and LP-relevant mean amplitudes were extracted. Data analysis Individual behavioral trial data were processed in R using Quarto within R Studio ( Quarto: An OpenSource Scientific and Technical Publishing System , 2024; R core Team, 2016; RStudio Team, 2020). For each subject and task, the data were processed separately for each threshold (detection and identification), similar to our previous study (Wiens et al., 2023). For the detection threshold, something (PAS2) represented aware and nothing (PAS1) unaware. For the identification threshold, identify (PAS3) represented aware and something (PAS2) unaware. The consecutive trials in each staircase were divided into blocks of 16 trials (starting from the last trial). A block was considered to be at threshold levels if the number of aware trials was between 6 and 10 (cumulative probability below .80). For each subject, only trials in valid blocks were processed further. The opacity level on each trial was mean centered within each block to remove drift; thus, only relative differences within a block were considered in the analyses. During data collection, the experiment software tried to adjust the opacity in 256 levels. After data collection, we realized that opacity could be manipulated only in 49 levels. Therefore, the opacity was recoded to vary from 0 (transparent) to 48 (opaque) before mean centering (for details, see Wiens, 2025). Individual trial behavioral data were merged with preprocessed individual trial EEG data, which contained mean VAN-relevant and LP-relevant amplitudes for each trial. Trials labeled as bad during EEG preprocessing were excluded (< 4% of all trials). If subjects had fewer than 25 trials in an unaware or aware condition for a threshold, their data were excluded for that threshold. On average, subjects had between 60 and 70 valid trials for individual combinations of task (digit, letter), threshold (detection, identification), and awareness (aware, unaware). Finally, individual trials were excluded separately for VAN and LP if the mean amplitudes deviated more than 3 SD s from the mean across all trials and subjects (< 0.7%). VAN-relevant and LP-relevant mean amplitudes were analyzed on trial level with Bayesian robust mixed effects regression models (Alday & van Paridon, 2021; Brown, 2021; Franke & Roettger, 2019; Kretzschmar & Alday, 2020). Amplitudes were not centered or standardized. Bayesian models were estimated with the R-package brms (Bürkner, 2017, 2018). In all models, vague priors were used for intercepts and slopes (i.e., normal distribution with M = 0 and SD = 2). Mixed effects models reduce the risk for biased effect size estimates because they include all available data (Matta et al., 2018). The main results are median estimates with Bayesian confidence intervals. For Bayesian hypothesis testing, the R-package bayestestR was used (Makowski, Ben-Shachar, & Lüdecke, 2019; Makowski, Ben-Shachar, et al., 2019). Bayes factors compare different models and provide evidence for or against a particular model (Dienes, 2016; Wagenmakers et al., 2016; Wiens & Nilsson, 2017). Thus, nonsignificance is not mistaken as evidence for no effect (Dienes, 2008; Makin & Orban de Xivry, 2019; Wasserstein & Lazar, 2016). The Bayes factor ( BF ) is a continuous measure of the relative evidence for one model versus another. For example, a BF10 > 3 supports the presence of an effect three times more than the absence of an effect (Dienes, 2016). The BF is a continuous measure of evidence, but we use verbal labels such as moderate, strong, very strong, and extreme (Wagenmakers et al., 2018). Although the label anecdotal was suggested for 1 < BF < 3, we prefer to use the label ambiguous . For behavioral analysis, performance on each trial was coded as 1 if subjects responded correctly whether the digit was below 5 or above 5 (digit task) or whether the stimulus was a digit or letter (letter task). Incorrect trials were coded as 0 . These trial-level data were analyzed with Bayesian logistic mixed models. The EEG data for the digit task and for the letter task were processed separately in two analyses. The goal of the first analysis was to show that VAN and LP can be observed at the detection threshold. The analysis considered only the EEG data for the detection threshold and examined effects of awareness and opacity on mean amplitudes. In a separate analysis of VAN-relevant and LP-relevant mean amplitudes ( EEG ), awareness (unaware vs. aware) and opacity (continuous variable) were modeled as fixed effects and were allowed to vary randomly across subjects (i.e., varying slopes and intercept over subjects). Awareness was dummy coded as 0 (unaware) and 1 (aware). Thus, awareness captured effects of awareness on mean amplitudes independent of opacity. The model formula in brms was as follows: EEG ∼ 1 + awareness + opacity + (1 + awareness + opacity | id). The main, second analysis examined effects of awareness for both thresholds (identification and detection). Fixed effects included the interaction of threshold and awareness, and the interaction of threshold and opacity (together with lower-order effects). These effects were allowed to vary randomly across subjects. Awareness was dummy-coded as 0 (unaware) and 1 (aware), and threshold was dummy-coded as 0 (identification) and 1 (detection). Thus, awareness by itself captured the awareness effect for the identification threshold, and the interaction of awareness with threshold captured the change of the awareness effect from identification to detection. The model formula in brms was as follows: EEG ∼ 1 + threshold * awareness + threshold * opacity + (1 + threshold * awareness + threshold * opacity | id). Note that this model automatically includes lower-order effects (i.e., threshold, awareness, and opacity). In two final models, the above analyses were repeated with task (digit/letter) as an additional predictor. In the first analysis of the detection threshold, the model formula was EEG ∼ 1 + task * awareness + task * opacity + (1 + awareness + opacity | id). In the second analysis with detection and identification thresholds, the model formula was EEG ∼ 1 + task * threshold * awareness + task * threshold * opacity + (1 + threshold * awareness + threshold * opacity | id). Task could not be modeled as a random effect because each subject participated only in one task. Note that when generating grand mean figures and topographies, trials for each subject and condition were simply averaged, whereas mixed models take into account the number of trials for each subject and condition. In theory, there may be differences between the figures and the results of the mixed models. In practice, no notable differences were observed. Although opacity could vary between 0 (transparent) and 48 (opaque), opacity levels were low in all conditions. For the digit task, mean opacity was 2.65 for unaware at detection, 2.93 for aware at detection, 4.07 for unaware at identification, and 4.37 for aware at identification. For the letter task, mean opacity was 2.26 for unaware at detection, 2.57 for aware at detection, 4.26 for unaware at identification, and 4.48 for aware at identification. Unsurprisingly, opacity levels tended to be higher for identification than detection. Despite the tiny differences between unaware and aware in each threshold, opacity was included as a predictor in the main analyses to remove its potentially confounding effect on mean amplitudes (Sassenhagen & Alday, 2016). In an exploratory analysis, we combined the present data with the individual trial data of a previous study (Wiens, 2025). Similarly as in the above analyses, we conducted Bayesian robust mixed models (see Wiens, 2025). Experiment 1 - Results Behavior Bayesian logistic mixed models showed that at the identification threshold, subjects could discriminate the stimuli more accurately when they responded identify rather than something. At the detection threshold, subjects performed close to the chance level (i.e., 0.50) regardless of whether they were aware or unaware. For the digit task, the estimated proportion correct was 0.96 for aware-identify, 0.71 for unaware-identify (aware minus unaware: BF10 > 100), 0.61 for aware-detect, and 0.50 for unaware-detect (aware minus unaware: 1 < BF 100), 0.56 for aware-detect and 0.52 for unaware-detect (aware minus unaware: BF01 > 3). For both tasks, the awareness effect was greater for identification than detection ( BF10 s > 100). The tasks did not differ in this pattern ( BF01 > 3). On average, subjects received 21.1 catch trials ( SD = 5.1). Subjects were unlikely to make a false alarm by responding something or identify ( M = 5.9%, SD = 7.4). ERP Figure 1 shows grand mean ERPs (Figure 1a) and topographies (Figure 1b) for VAN-relevant amplitudes, for combinations of task (digit/letter) and threshold (detection/identification). Figure 2 shows the same results for LP-relevant amplitudes. Figures 1 and 2 attest to the quality of the data. Further, they show that the preregistered electrodes and intervals were sensitive to VAN and LP. The difference of aware minus unaware yielded a negativity for VAN-relevant amplitudes (i.e., VAN; see di ff in 1a) and a positivity for LP-relevant amplitudes (i.e., LP; see di ff in 2a). The figures suggest that there were no noteworthy differences in topographies for the detection and identification tasks. In the text below, detection VAN and detection LP refers to the difference of something [PAS2] minus nothing [PAS1], that is, the detection threshold. Identification VAN and identification LP refers to the difference of identify [PAS3] minus something [PAS2], that is, the identification threshold. Figure 3 summarizes the estimated awareness effects from several Bayesian robust mixed models for VAN-relevant amplitudes (Figure 3a) and LP-relevant amplitudes (Figure 3b). For each effect, the figure shows the median effect, the credible interval (95%), and the strength of the Bayes factor in terms of evidence categories (Wagenmakers et al., 2018). Each effect refers to the awareness effect (aware−unaware) or its interaction with other variables. Specifically, digit_detect (or letter_detect ) refers to the difference of aware minus unaware at the detection threshold for the digit task (or letter task). This difference captures detection VAN (Figure 3a) and detection LP (Figure 3b). Similarly, digit_identify (or letter_identify ) refers to the difference of aware minus unaware at the identification threshold for the digit task (or letter task). This difference captures identification VAN (Figure 3a) and identification LP (Figure 3b). Figure 1 Experiment 1: Grand mean ERPs (a) and topographies (b) for VAN-relevant amplitudes. (a) Grand mean ERPs for VAN-relevant electrodes, for combinations of task (digit/letter) and threshold (detection/identification). (b) Topographies of differences (aware minus unaware) for VAN. Figure 2 Experiment 1: Grand mean ERPs (a) and topographies (b) for LP-relevant amplitudes (a) Grand mean ERPs for LP-relevant electrodes, for combinations of task (digit/letter) and threshold (detection/identification). (b) Topographies of differences (aware minus unaware) for LP. With regard to VAN (Figure 3a), all amplitude differences of aware minus unaware were negative. These findings provide evidence for detection VAN and identification VAN for both digit task and letter task ( BF10 s > 3). Critically, the results provided moderate evidence ( BF01 s > 3) that detection VAN and identification VAN did not differ for either the digit task ( digit_identify for detect − identify ) or letter task ( letter_identify for detect − identify ). It was ambiguous whether there was a three-way interaction of task, threshold, and awareness ( task_identify for detect − identify for L − D task , 1 < BF 3). It was ambiguous whether identification VAN differed between the digit task and the letter task ( task_identify for L − D task , 1 < BF 3); however, it was ambiguous whether there was identification LP for the letter task (1 < BF 10) that detection LP was larger than the identification LP for the digit task ( digit_identify for detect − identify ) and the letter task ( letter_identify for detect − identify ). The results suggested that there was no three-way interaction of task, threshold, and awareness ( task_identify for detect − identify for L − D task , BF01 > 3). The results further suggested that the detection LP was similar for the digit task and the letter task ( task_detect for L − D task , BF01 > 3). In addition, identification LP was similar for the digit task and the letter task ( task_identify for L − D task , BF01 > 3). Figure 3 Experiment 1: Estimated awareness e ff ects from Bayesian robust mixed models for VAN-relevant amplitudes (a) and LP-relevant amplitudes (b). BFlvl refers to the evidence category of the Bayes Factor, D refers to digit task, and L refers to letter task. For more details, see main text. (a) Estimated awareness e ff ects for VAN-relevant electrodes. (b) Estimated awareness e ff ects for LP-relevant electrodes. In an exploratory analysis, we compared the present individual trial data with the trial data from our previous study with a Gabor ring (Wiens et al., 2023). Figure 4 summarizes the estimated task effects for VAN (Figure 4a) and LP (Figure 4b). The label ring_digit (or ring_letter ) refers to task differences between the present digit task (or letter task) and the ring task (which was coded as the baseline). As shown in Figure 4a, detection VAN did not differ between digit and ring ( BF01 > 3). In contrast, identification VAN was larger (more negative) for digit than ring ( BF10 > 30). There was moderate evidence for an interaction between task (ring, digit) and awareness ( BF10 > 3). There was only ambiguous evidence for task differences between letter and ring (1 < BF < 3). For LP (Figure 4b), comparisons provided ambiguous or moderate evidence against any task effects. With regard to task performance, there was ambiguous to moderate evidence against any task effects (see Wiens, 2025). In sum, the main finding was that identification VAN was larger for the digit task than the ring task. Experiment 1 - Discussion With regard to VAN, the main results were moderate to extreme evidence for detection VAN and for identification VAN in both digit and letter tasks. For each task, moderate evidence suggested no difference between detection VAN and identification VAN. There was ambiguous evidence for an interaction of task, threshold, and awareness. With regard to LP, the main results showed strong to extreme evidence for detection LP in both tasks and for identification LP in the digit task. There was only ambiguous evidence for identification LP in the letter task. For each task, strong to extreme evidence showed that identification LP was smaller (less positive) than detection LP. Moderate evidence supported no interaction of task, threshold, and awareness. Together, these results suggest that VAN is Figure 4 Estimated task e ff ects (ring vs. Exp. 1) from Bayesian robust mixed models for VAN (a) and LP (b). BFlvl refers to the evidence category of the Bayes Factor. (a) Estimated task e ff ects for VAN. (b) Estimated task e ff ects for LP. similarly sensitive to detection awareness and identification awareness, whereas LP is less sensitive to identification awareness than detection awareness. In contrast to our hypotheses, VAN was equally sensitive to detection awareness and identification awareness. This pattern was observed in two separate tasks with different subject samples (Figure 3a). In the digit task, subjects rated their awareness and discriminated whether a stimulus was below 5 or above 5. In the letter task, subjects rated their awareness and discriminated whether a stimulus was a digit or a letter. The present findings seem contrary to previous reports that identification VAN was smaller (less negative) than detection VAN (Koivisto et al., 2017; Wiens et al., 2023). The findings are unexpected, particularly so because the present digit task closely resembled that of a previous study (Koivisto et al., 2017): Both studies used the same stimuli (3, 4, 6, & 7) shown under similar conditions, a similar PAS rating scale, and the same discrimination task (below or above 5). A possible explanation for the discrepancy is that results of the previous study (Wiens et al., 2023) were confounded because detection VAN and identification VAN were assessed in separate tasks rather than concurrent tasks, as discussed earlier. However, when we compared the present data with that of our previous study with a Gabor ring (Wiens et al., 2023), exploratory analyses (see Figure 4a) showed that identification VAN was greater (more negative) in the present digit task than the former ring task (Wiens et al., 2023). These findings suggest that the finding of a greater (more negative) detection VAN than identification VAN may not necessarily be a false positive (Koivisto et al., 2017). It is unclear though why the pattern of results apparently differed between the present study and the previous study (Koivisto et al., 2017), given that they were very similar in task and stimuli. In sum, despite different tasks and stimuli, robust detection VAN and identification VAN were observed, and the size of VAN was similar. These findings suggest that VAN is sensitive to different tasks and stimuli, but it is not specific to low-level (detection) experiences. Instead, VAN is sensitive to both low-level (detection) experiences and high-level (identification) experiences. Whereas previous studies provided only ambiguous evidence for identification VAN (Koivisto et al., 2017; Wiens et al., 2023), the present results from different tasks and stimuli suggest that identification VAN is a general finding. With regard to LP, the present results matched those of previous studies (Koivisto et al., 2017; Wiens et al., 2023). Results showed that LP was less sensitive to identification awareness than detection awareness, and the identification LP was small (Figures 3b and 8b). Behavioral results showed that at the identification threshold, subjects could discriminate the stimuli more accurately when they were aware rather than unaware. At the detection threshold, subjects performed close to chance regardless of whether they were aware or unaware. These findings support the validity of the awareness ratings in capturing detection awareness and identification awareness. Experiment 2 was a follow-up and was conducted for two main reasons. First, because previous studies provided only ambiguous evidence for identification VAN (Koivisto et al., 2017; Wiens et al., 2023), we performed Experiment 2 to examine whether identification VAN is a robust finding. Second, exploratory analyses suggested that identification VAN was larger (more negative) for the digit task in Experiment 1 than for the ring task in our previous study (Figure 4a). The reason for this difference is unclear. An obvious difference was that stimuli were smaller in the digit task than the ring task. Although a previous study did not find that detection VAN differed with stimulus sizes of Gabor patches (Eklund & Wiens, 2018), we manipulated stimulus size in Experiment 2 to rule out any confound from stimulus size. In Experiment 2, subjects performed the same digit task as in Experiment 1. However, opacity of the digits was adjusted only to the identification threshold. On separate trials, digits were shown in small or big size. Small digits had the same size as the digits in Experiment 1, and big digits had the same size as the Gabor ring in Wiens et al. (2023). In two separate but intertwined staircases, opacity for each digit size was continuously adjusted to the identification threshold. We hypothesized that the identification VAN would be observed for both small and big digits, consistent with results of Experiment 1. Furthermore, we hypothesized that the identification VAN would be smaller (less negative) for big than small digits, consistent with the idea that differences in stimulus size may account for differences in identification VAN between the digit task and the ring task. For LP, we predicted that identification LP would be unaffected by stimulus size. Experiment 2 - Method Preregistration and data sharing Experiment 2 was preregistered at OSF. All material, data, and scripts are available at a public university repository (Wiens, 2025). Participants The research was approved by the Swedish Ethical Review Authority (2022-06409-01). Recruitment was similar to that in Experiment 1. Twenty-nine subjects were recruited. After initial data screening, one participant was excluded because of an unreasonable response pattern (many nothing responses at maximum opacity). Additionally, eight subjects were excluded in one condition (big digit) because they had too few valid trials (< 25 trials); however, their data were retained in the other condition (small digit). The final sample comprised 28 subjects (10 males; 27 right-handed, age: M = 26, SD = 5). Stimuli, procedure, and data analysis The design of Experiment 2 was similar to that of Experiment 1 with some exceptions. The stimuli were digits (3, 4, 6, and 7) shown in either small size (5 mm x 7 mm) or big size (20 mm x 30 mm). For small, stimulus parameters were identical to those in the digit task in Experiment 1. For big, digit height was identical to that of the Gabor ring in a previous study (Wiens et al., 2023). As in Experiment 1, opacity was continuously adjusted over trials in two separate but interleaved staircases. Here, however, both staircases adjusted opacity to the identification threshold: one for small digits and the other for big digits. In each trial, a small or big digit was shown. Subjects rated their visual experience of the digit (nothing, something, identify) and whether the digit was smaller or larger than 5. The probability of a catch trial (i.e., transparent stimulus) was 0.10 on each trial. On average, subjects completed 412 trials ( SD = 55) including catch trials. After testing six subjects, we noticed that for big digits, some subjects responded mainly identify [PAS3] or nothing [PAS1] (skipping something [PAS2]). Informal observations suggested that some subjects may have responded identify if they experienced only parts of the digits. For subsequent subjects, we clarified instructions by emphasizing that they should only respond identify if they could clearly see the entire digit. Analyses were similar to those in Experiment 1. The difference of identify [PAS3] minus something [PAS2] captured the identification threshold. In the main analyses, some subjects ( n = 8) were excluded from the big condition because they mainly responded nothing and identify (four of these subjects were tested after instructions were clarified, see above). Thus, these subjects lacked trials (< 25) to measure the difference between identify and something (identification threshold). Because opacity could be manipulated only in 33 levels (and not in 256 levels), opacity was recoded to vary from 0 (transparent) to 32 (opaque; for details, see Wiens, 2025). For the big digits, mean opacity was 1.02 for unaware and 1.64 for aware. For the small digits, mean opacity was 3.62 for unaware and 3.92 for aware. Statistical models were similar to those in Experiment 1. An exploratory analysis of identification VAN and LP compared the trial data from Experiment 1 and 2 with those of our previous study (Wiens et al., 2023). Data were analyzed with Bayesian robust mixed models (see Wiens, 2025). Experiment 2 - Results Behavior Bayesian logistic mixed models showed that subjects performed better when they were aware than unaware (at the identification threshold). For big digits, the estimated proportion correct was 0.97 for aware trials and 0.66 for unaware trials (aware minus unaware: BF10 > 100). For small digits, the estimated proportion correct was 0.95 for aware and 0.71 for unaware (aware minus unaware: BF10 > 100). There was ambiguous evidence that big and small did not differ in this pattern (1 < BF01 < 3). On average, subjects received 40.8 catch trials ( SD = 8.4). It was unlikely that subjects made a false alarm by responding something or identify ( M = 8.4%, SD = 10.4). ERP Figure 5 shows grand mean ERPs (Figure 5a) and topographies (Figure 5b) for the VAN-relevant amplitudes for big and small digits at the identification threshold. Figure 7a summarizes the estimated awareness effects. As shown by di ff in Figure 5a, the difference of aware minus unaware yielded a negativity for VAN-relevant amplitudes (i.e., identification VAN). Results provided very strong evidence for identification VAN for both big and small digits ( BF10 s > 30) and moderate evidence ( BF01 > 3) that identification VAN did not differ with stimulus size. Figures 6 and 7b show the results for LP-relevant amplitudes. Results provided moderate to strong evidence for identification LP for both big and small digits ( BF10 s > 3) and moderate evidence ( BF01 > 3) that identification LP did not differ with stimulus size. An exploratory analysis combined the identification data from Experiment 1, 2, and our previous study with a Gabor ring (Wiens et al., 2023). Figure 8 summarizes the estimated task differences among the tasks for VAN (Figure 8a) and LP (Figure 8b). In each comparison, the ring task served as the baseline. Letter refers to the letter task in Experiment 1 (minus the ring task). Digit refers to the digit task in Experiment 1, digit_small refers to the small-digit task in Experiment 2 (note that stimulus parameters were identical to the digit task in Exp. 1), digit_big refers to the big-digit task in Experiment 2, and all_digits refers to the average of digit (Exp. 1), digit_small (Exp. 2), and digit_big (Exp. 2). For VAN, results showed that identification VAN was larger (more negative) for digit (Exp. 1) than Gabor ring ( BF10 > 100, see also Figure 4). Similarly, identification VAN was larger for big digits and across all digit conditions ( BF10s > 3). For LP and for task performance, all comparisons provided ambiguous to moderate evidence against any task effects. Figure 5 Experiment 2 grand mean ERPs (a) and topographies (b) for VAN-relevant amplitudes at the identification threshold. (a) Grand mean ERPs for big and small digits. (b) Topographies for big and small digits. Figure 6 Experiment 2 grand mean ERPs (a) and topographies (b) for LP-relevant amplitudes at the identification threshold. (a) Grand mean ERPs for big and small digits. (b) Topographies for big and small digits. Experiment 2 - Discussion For both small and big digits, there was very strong to extreme support for identification VAN (Figure 7a). In contrast to previous studies that found only ambiguous evidence for identification VAN (Koivisto et al., 2017; Wiens et al., 2023), the present finding suggests that identification VAN is a robust phenomenon. Identification VAN did not differ with stimulus size (Figure 7a). Thus, results complemented those of similar detection VAN irrespective of size of Gabor patches (Eklund & Wiens, 2018). An exploratory analysis showed that identification VAN was larger (more negative) that in our previous study with a Gabor ring (Figure 8a)). This finding is inconsistent with our hypothesis that identification VAN would be smaller (less negative) to big digits than small digits. Thus, differences in identification VAN between Figure 7 Experiment 2: Estimated e ff ects of identification awareness for VAN-relevant amplitudes (a) and LPrelevant amplitudes (b). BFlvl refers to the evidence category of the Bayes Factor. (a) Estimated awareness e ff ects for big and small digits. (b) Estimated awareness e ff ects for big and small digits. Figure 8 Estimated task di ff erences of Exp. 1 and 2 compared with ring study for VAN (a) and LP (b). In each comparison, the ring task served as the baseline. Letter and digit refer to the letter and digit tasks (Exp. 1), digit_small and digit_big refer to the small and big digit tasks (Exp. 2), and all_digits refers to the average of digit (Exp. 1), digit_small (Exp. 2), and digit_big (Exp. 2). BFlvl refers to the evidence category of the Bayes Factor. (a) Estimated task e ff ects for VAN (ring as baseline). (b) Estimated task e ff ects for LP (ring as baseline). the digit task and the Gabor ring (Wiens et al., 2023) cannot be attributed to differences in stimulus size. Results showed that LP is sensitive to identification awareness and that identification LP does not differ with stimulus size (Figures 7b). Although identification LP was relatively small (about 0.6 µV), identification LP to big and small digits was similar to that in our previous study with a Gabor ring (Figure 8b). General discussion The main goal of the present research was to examine whether VAN is sensitive to high-level experiences (identification) as well as low-level experiences (detection). The main findings of Experiment 1 and 2 can be summarized as follows: With regard to VAN, Experiment 1 showed that VAN was similarly sensitive to identification awareness and detection awareness, for a digit task and for a letter task in two separate subject samples. Experiment 2 used the same digit task as in Experiment 1 with two stimulus sizes and found identification VAN, and this identification was unaffected by stimulus size. With regard to LP, Experiment 1 showed that LP was smaller (less positive) for identification awareness than detection awareness. Experiment 2 confirmed a small identification LP (about 0.5 µV) that was unaffected by stimulus size. The present results showed that VAN was sensitive to high-level experiences (identification) as well as low-level experiences (detection). Identification VAN as well as detection VAN were observed in three subject samples with different tasks (digit and letter) and stimulus sizes (Figures 3a and 7a). Most previous studies did not clearly distinguish awareness in terms of identification and detection (Derda et al., 2019; Jimenez et al., 2021; Koivisto & Revonsuo, 2008; Koivisto et al., 2005, 2009; Tagliabue et al., 2016; Wilenius & Revonsuo, 2007). In the two studies that separated identification from detection awareness (Koivisto et al., 2017; Wiens et al., 2023), results provided only ambiguous evidence for identification VAN. In contrast, the present findings provided convincing evidence that VAN is sensitive to high-level experiences as well as low-level experiences. Effect of Task Previous studies (Koivisto et al., 2017; Wiens et al., 2023) found that VAN was smaller (less negative) to identification awareness than detection awareness. In contrast, the present study found no differences in VAN between identification and detection awareness. This difference was unexpected, given that the present study design closely resembled that of a previous study (Koivisto et al., 2017). Specifically, Experiment 1 in the present study used the same stimuli (3, 4, 6, & 7) shown under similar conditions, a similar PAS rating scale, and subjects were required to discriminate whether the digit was below or above 5. However, a possible explanation is that the studies differed in design: In the present study, subjects performed a combination of detection and identification tasks whereas in the previous study, detection and identification were assessed in separate tasks. This difference in execution may have affected results. For example, because a supplementary analysis (Koivisto et al., 2017) found that detection LP differed for the detection task and identification task, VAN may have been affected by separate tasks. In support, research has shown that different brain areas activate depending on task, even to the same stimuli (Andersen et al., 2022; Straube & Fahle, 2011). For example, in an fMRI study (Straube & Fahle, 2011), participants responded whether geometric figures were left or right of fixation (detection) or whether the figures were symmetric (identification). Participants performed these tasks separately, and performance was matched for both tasks. Although both tasks activated object-selective areas, relatively high-level areas (e.g., lateral occipital complex) were more active during identification than detection. If these areas contribute differently to VAN, then these differences in neural generators could explain differences in VAN depending on the task. Effect of Stimulus However, task differences cannot explain that our previous study (Wiens et al., 2023) found a smaller (less negative) identification VAN than detection VAN, as that study also used a combination of detection and identification tasks but with a Gabor ring. Exploratory analyses in this study showed that whereas detection VAN did not differ between studies, identification VAN did: Compared to the small identification VAN in the ring task, identification VAN was large (more negative) in the digit task in Experiment 1 (Figure 4a) and in the digit task with big digits in Experiment 2 (Figure 8a). As shown in Figure 8a, strong evidence supported that identification VAN was greater (more negative) across all three digit conditions, than in the ring task. It is unclear why there were differences in identification VAN between studies: First, stimulus size is an unlikely reason because the stimuli in the big-digits condition had the same size as the ring. Second, differences in PAS ratings are an unlikely reason. In both studies, behavioral results were similar for the detection threshold and the identification threshold. At the detection threshold, subjects discriminated the stimuli close to chance (about 50% correct) irrespective of their rating (nothing or something). At the identification threshold, subjects discriminated the stimuli more accurately when they rated identify (about 95% correct) rather than something (about 70% correct). Third, given that there are systematic differences between levels of experience for different stimuli, these may be difficult to detect because many studies did not have enough power to evaluate consecutive changes between PAS ratings. So, there may be systematic differences, but these are difficult to find across studies unless data are shared openly (Lakens et al., 2016; Munafò et al., 2017). In this context, our exploratory analysis between studies is interesting because it suggests that VAN to a high-level experience of identification differs between Gabor ring and digits, even if the reason is currently unknown. Effect of Awareness Levels One of the main features of the study was to focus on effects on VAN of a particular low-level experience (detection) and a particular high-level experience (identification) within the same task. Other studies examined effects of gradual changes in experience on VAN-relevant amplitudes, using various methods of manipulating awareness (Kim & Blake, 2005; Koivisto & Revonsuo, 2010). In general, these studies can be grouped in terms of suppression of sensory signals ( perceptual blindness ) or attentional failures to access sensory signals ( attentional blindness , Kanai et al., 2010). Examples of perceptual blindness are contrast manipulation (as in the present study), and backward masking (Kim & Blake, 2005). Such studies suggest that VAN varies with PAS ratings (Derda et al., 2019; Jimenez et al., 2018, 2021; Koivisto et al., 2017; Tagliabue et al., 2016). Examples of attentional blindness are inattentional blindness, change blindness, and attentional blink (Koivisto & Revonsuo, 2010). Studies of inattentional blindness showed that unaware stimuli per se elicit VAN, but these studies did not manipulate level of experience (for a review see: Hutchinson, 2019). In change blindness, subjects sometimes seem to sense a change without being able to identify the change. These sense trials may be considered as low-level experiences when compared to see trials. Although visual areas activate more to sense trials, than to trials in which a change is not detected (Scrivener et al., 2021), research is mixed as to whether VAN-relevant amplitudes on sense trials differs from VAN-relevant amplitudes on see trials (Scrivener et al., 2019, 2021). In contrast, research on the attentional blink found that VAN-relevant amplitudes to faces changed linearly across PAS ratings; thus, VAN tracked increases in level of experience (Eiserbeck et al., 2021; Roth-Paysen et al., 2022). Notably, one of these studies (Roth-Paysen et al., 2022) compared the attentional blink condition with another condition in which visibility was manipulated through contrast manipulation (as in the present study); for each subject, the stimulus contrast was adjusted to match performance in the attentional blink condition. Results for both conditions showed similar ERPs (N170, VAN, and P3) that varied linearly across PAS ratings. These findings suggest that effects of attentional blink may resemble those of contrast manipulation, maybe because the underlying processes are more similar than previously believed (Dehaene et al., 2006; Kanai et al., 2010). Taken together, research that used different approaches to manipulate awareness has shown that VAN-relevant amplitude is sensitive to increases in level of experience. Neural Generators of VAN It is unclear whether the present findings can be generalized to the alternative conceptualizations of low-level and high-level experiences. If neural generators of VAN can be assumed to be similar or in close vicinity, results in EEG should be similar to the present results because EEG may not have sufficient spatial resolution to resolve these differences (Biasiucci et al., 2019; Luck, 2014). To illustrate, although studies have found amplitude differences in VAN, the topography and timing of VAN appears to be relatively insensitive to different experimental contexts. In support, visual inspection of the grand mean ERPs and topographies (Figures 1 and 5) suggests that the temporal and spatial patterns of VAN were similar for low-level and high-level experiences, for digits and letters, and for two stimulus sizes. Further, VAN matched that in our previous study with a Gabor ring in terms of electrodes and interval (Wiens et al., 2023). Another reason for the consistency of VAN is that similar neural generators have been reported across different contexts. These generators are occipito-temporal areas including lateral occipital complex (LOC) and infero-temporal cortex (Colombari & Railo, 2024; Colombari et al., 2024; Förster et al., 2020; Y. Liu et al., 2012). For example, research suggests that LOC is activated to different levels of experience: LOC was activated when subjects reported whether they were aware of a Gabor patch (i.e., detection, Y. Liu et al., 2012) or whether they were aware of the tilt of a clearly visible Gabor patch (i.e., identification, Colombari et al., 2024). Further, when neural activity to backward masked visual stimuli (rectangle or rotated rectangle) was recorded with magnetoencephalography (MEG), machine learning algorithms trained on activity in occipital areas between 132 and 320 ms could decode the three consecutive transitions between the four PAS ratings (Andersen et al., 2016). These findings suggest similar neural generators for VAN across contexts. Further, consideration of the visual system explains why similar generators are involved despite different levels of experience. When presented with a visual stimulus, the visual system extracts relevant features from the whole visual field and integrates them into a coherent experience (V. A. F. Lamme, 2018; V. Lamme, 2015). Thus, a low-level experience is actually a high-level experience because it actively excludes high-level experiences of the stimulus identity. For example, an experience of something may be low level in terms of content, but it also excludes the experiences of a digit or a letter. Accordingly, when low-level experiences and high-level experiences occur in the same task (as in the present study), different levels of experiences involve similar neural generators. In support, a recent fMRI meta-analysis (MacLean et al., 2023) found that different levels of experiences have similar effects on sensory cortices. The meta-analysis focused on two comparisons: baseline with unaware conditions, and unaware with aware conditions. Results showed that unaware (vs baseline) activated visual areas (superior lateral occipital cortex), whereas aware (vs unaware) activated only higher-order, non-visual areas. To exemplify with a study (Heinzel et al., 2008), single letters (A or B) activated extrastriate cortex irrespective of whether subjects were unaware of the letters (discrimination performance at chance) or aware of the letters. These findings suggest that in general, sensory cortex may be similarly activated during different levels of experience. Furthermore, attention research has shown that expectation of location, feature, and object category activates stimulus-relevant brain areas even in the absence of a stimulus (Lange et al., 2018). For example, subjects categorized noisy images that slowly transformed into a face or house; for each trial, subjects knew whether to expect a face or a house (Esterman & Yantis, 2010). Analyses of fMRI data showed anticipatory activity in category-specific brain areas (e.g., fusiform gyrus for faces) even before the images contained enough category-specific information. This suggests that expectations of highlevel stimuli evoke anticipatory activity in relevant neural generators irrespective of the final level of experience. Whereas EEG has low spatial resolution, it has excellent temporal resolution (Biasiucci et al., 2019; Luck, 2014). In general, simple stimulus features are processed faster than complex stimulus features (Bullier, 2001). For example, in a MEG study, subjects identified individual digits and letters that were shown left or right of fixation (Gwilliams & King, 2020). Stimulus position was decoded maximally by 120 ms whereas stimulus identity was decoded by 225 ms. These findings suggest timing difference between detection and identification, but such a long timing difference is atypical. Because the main stimulus was flanked by similar stimuli (Gwilliams & King, 2020), processing identity was more demanding than processing position, and this processing difference may have contributed to the delay. In fact, when stimuli are shown without flankers at fixation, feedforward processes are fast and reach frontal areas within 100 ms (Bullier, 2001). Similarly, intracranial recordings show that object category (e.g., animals vs faces) can be decoded after 100 ms in inferior temporal cortex, and this response is insensitive to variations in scale and viewpoint (H. Liu et al., 2009). These findings show that even though visual processing is hierarchical, any inherent delays are relatively small. Because stimuli in the present study were shown at fixation without flankers, any delay between detection and identification was probably too small to affect VAN. Also, even if feedforward processes may be faster for simple than complex stimuli, VAN is believed to reflect mainly recurrent processes that occur later (Knight et al., 2024; V. A. F. Lamme, 2018; V. A. F. Lamme et al., 2000). PAN as an NCC Although the concept of an awareness negativity (i.e., PAN) is useful as an NCC, it has its limits. The main issue is that it is not sensitive to all aspects of a perceptual experience. In vision, when subjects were shown tilted Gabor patches (left or right), machine learning algorithms trained on EEG data could decode awareness in terms of detection but not identification (Rodríguez-San Esteban et al., 2025). Neither voltage EEG nor time-frequency representations of EEG contained information about experiences of the direction of the Gabor patch. Notably, this result is not inconsistent with the present findings. In the previous study (Rodríguez-San Esteban et al., 2025), identification referred to the experience of either left or right tilt, whereas in the present study, identification referred to the experience of the digit or letter versus the experience of something. Similarly, the auditory component of PAN may be insensitive to lateralized experiences under some conditions. In vision and touch, lateral stimulation produces mainly a contralateral PAN, consistent with anatomical pathways (PAN, Dembski et al., 2021). In hearing, an overall PAN has been found in detecting a sound or identifying a recurrent tone hidden in background noise (Dykstra et al., 2017; Eklund & Wiens, 2019). Additionally, lateralized clicks presented over headphones produced a contralateral awareness negativity to clicks (Eklund et al., 2021). However, when the clicks were presented via loudspeakers (i.e., in auditory free-field conditions), results did not clearly support a lateralized awareness negativity (Gerdfeldter et al., 2024). Also, when tones were presented monaurally, results did not provide clear evidence for a lateralized awareness negativity (Gerdfeldter et al., 2025). These findings suggest that PAN may not be sensitive to some experiences. It is possible though that other NCC will be discovered that map onto these experiences. For example, in somatosensory perception, the mere detection of a pulse is associated with a contralateral negativity (i.e., PAN, Dembski et al., 2021). However, in a temporal discrimination task with two stimulation pulses that are well above threshold, the experience of two pulses rather than one pulse is associated with a frontocentral positivity after about 150 ms (Förster et al., 2025). This research emphasizes that the search for NCCs needs to be expanded beyond detection in terms of different experiences (Bachmann & Aru, 2023) and in terms of their development over time (Aru & Bachmann, 2017; Hense et al., 2024). Future research is needed to establish the limits of PAN in indexing different levels of experiences, and to explore whether other NCCs may be found instead. Another issue is that even if PAN is an NCC, it is debated whether it is affected by processes that precede, run in parallel, or follow the actual experience (Aru et al., 2012; de Graaf et al., 2012). Although PAN appears to be independent from post-perceptual processes (Dembski et al., 2021; Förster et al., 2020), PAN may be partly affected by attention (Bola & Doradzinska,´ 2021; Koivisto & Revonsuo, 2010). Furthermore, the concept of PAN is consistent with the idea that it reflects recurrent processing, and that this recurrent processing is an NCC. However, recurrent processing occurs even for unconscious processes, suggesting that recurrent processing may be necessary but not sufficient for conscious experiences (V. A. F. Lamme, 2018). Also, in computation modeling, recurrent processes are not inherently different from feedforward processes, as any recurrent process can be unfolded into feedforward models, albeit feedforward models are more computational inefficient and less flexible (Kreiman & Serre, 2020). Finally, because research has exclusively focused on establishing PAN at the group level, it is unresolved whether it can be reliably measured at the single-subject level (Fischer et al., 2024). This would be necessary for it to be useful to measure awareness in patients with disorders of consciousness. LP as an NCC As seen in Figures 3b and 7b, results showed a weak identification LP that was smaller in amplitude than the detection LP. As seen in Figures 4b and 8b, the present results were similar to our previous study with a Gabor ring (Wiens et al., 2023). Even though LP is correlated with phenomenal experience, it can be separated from it; thus, LP-relevant amplitudes reflect processes related to task relevance, evidence accumulation, and response preparation (Cohen et al., 2020, 2024; Dellert et al., 2021, 2022; Dembski et al., 2021; Förster et al., 2020; Koivisto & Revonsuo, 2010; Kronemer et al., 2022; O’Connell & Kelly, 2021; Pitts et al., 2014; Schlossmacher et al., 2020; Schröder et al., 2021). With regard to evidence accumulation, LP is closely related to a central-parietal positivity (CPP, O’Connell & Kelly, 2021). The CPP provides a convincing theoretical framework for the classic P3b or P300 in terms of an internal mechanism of evidence accumulation (O’Connell & Kelly, 2021; Twomey et al., 2015). For example, when subjects rated their experiences of weak circular patches, CPP increased with PAS ratings independent from stimulus intensity (Tagliabue et al., 2019). Interestingly, the pattern of results across PAS ratings (their Figure 4) closely matched the pattern of LP-relevant amplitudes in the present study: As shown in Figure 2a, LP-relevant amplitudes increased gradually across unaware-detect, aware-detect, unaware-identify, and aware-identify. However, the increase from unaware-detect to aware-detect was larger than the increase from unaware-identify to aware-identify, but this was not tested by Tagliabue et al. (2019). Consequently, the present results are consistent with LP as as a measure of evidence accumulation (as CPP) or as an NCC (Tagliabue et al., 2016, 2019). Critically, several studies using various tasks such as inattentional blindness (e.g., Pitts et al., 2014), attentional blink (e.g., Dellert et al., 2021), and no-report tasks (Cohen et al., 2020; Kronemer et al., 2022) have shown that LP can be reduced, if not eliminated, if task relevance or response preparation are minimized. Taken together, previous research has confirmed that LP can be separated from phenomenal experience. Conclusion The present results showed that when a specific low-level experience (detection) is compared with a specific high-level experience (identification), a similar VAN was observed for both experiences. These results confirm the sensitivity of VAN to both low-level and high-level experiences, consistent with recurrent processing theory. However, results emphasize the limited specificity of VAN in separating between low-level and high-level experiences, suggested by the similarity of VAN in both conditions. Declarations This work was supported by a grant to Stefan Wiens from Marianne and Marcus Wallenberg Foundation (MMW 2019-0102). We thank Tone Lindholm and Laura Lilliehöök for their assistance in data collection for Experiment 2. References 1. Alday, P. M., & van Paridon, J. (2021, June). Away from arbitrary thresholds: Using robust statistics to improve artifact rejection in ERP (preprint). PsyArXiv. https://doi.org/10.31234/osf.io/wqrb5 Andersen, L. M., Pedersen, M. N., Sandberg, K., & Overgaard, M. (2016). Occipital MEG activity in the early time range (<300 ms) predicts graded changes in perceptual consciousness. Cerebral Cortex , 26 (6), 2677– 2688. https://doi.org/10.1093/cercor/bhv108 Andersen, L. M., Vinding, M. C., Sandberg, K., & Overgaard, M. (2022). Task requirements affect the neural correlates of consciousness. European Journal of Neuroscience , ejn.15820. https://doi.org/10.1111/ejn.15820 Aru, J., & Bachmann, T. (2017). In and Out of Consciousness: How Does Conscious Processing (D)evolve Over Time? Frontiers in Psychology , 8 . https://doi.org/10.3389/fpsyg.2017.00128 Aru, J., Bachmann, T., Singer, W., & Melloni, L. (2012). Distilling the neural correlates of consciousness. Neuroscience & Biobehavioral Reviews , 36 (2), 737–746. https://doi.org/10.1016/j.neubiorev.2011.12.003 Auksztulewicz, R., & Blankenburg, F. (2013). Subjective Rating of Weak Tactile Stimuli Is Parametrically Encoded in Event-Related Potentials [Publisher: Society for Neuroscience Section: Articles]. Journal of Neuroscience , 33 (29), 11878–11887. https://doi.org/10.1523/JNEUROSCI.4243-12.2013 Bachmann, T., & Aru, J. (2023). Conscious interpretation: A distinct aspect for the neural markers of the contents of consciousness. Consciousness and Cognition , 108 , 103471. https://doi.org/10.1016/j.concog.2023. 103471 Biasiucci, A., Franceschiello, B., & Murray, M. M. (2019). Electroencephalography. Current Biology , 29 (3), R80– R85. https://doi.org/10.1016/j.cub.2018.11.052 Block, N. (2005). Two Neural Correlates of Consciousness. Trends in Cognitive Sciences , 9 (2), 46–52. https: //doi.org/10.1016/j.tics.2004.12.006 Bola, M., & Doradzinska, Ł. (2021). Perceptual Awareness Negativity—Does It Reflect Awareness or Attention?´ Frontiers in Human Neuroscience , 15 , 601. https://doi.org/10.3389/fnhum.2021.742513 Brown, V. A. (2021). An Introduction to Linear Mixed-Effects Modeling in R. Advances in Methods and Practices in Psychological Science , 4 (1), 251524592096035. https://doi.org/10.1177/2515245920960351 Bullier, J. (2001). Integrated model of visual processing. Brain Research Reviews , 36 (2), 96–107. https://doi.org/ 10.1016/S0165-0173(01)00085-6 Bürkner, P.-C. (2017). Brms : An R Package for Bayesian Multilevel Models Using Stan. Journal of Statistical Software , 80 (1), 1–28. https://doi.org/10.18637/jss.v080.i01 Bürkner, P.-C. (2018). Advanced Bayesian Multilevel Modeling with the R Package brms. The R Journal , 10 (1), 395. https://doi.org/10.32614/RJ-2018-017 Cohen, M. A., Dembski, C., Ortego, K., Steinhibler, C., & Pitts, M. (2024). Neural signatures of visual awareness independent of postperceptual processing. Cerebral Cortex , 34 (11), bhae415. https://doi.org/10.1093/ cercor/bhae415 Cohen, M. A., Ortego, K., Kyroudis, A., & Pitts, M. (2020). Distinguishing the Neural Correlates of Perceptual Awareness and Postperceptual Processing. The Journal of Neuroscience , 40 (25), 4925–4935. https: //doi.org/10.1523/JNEUROSCI.0120-20.2020 Colombari, E., Parisi, G., Tafuro, A., Mele, S., Mazzi, C., & Savazzi, S. (2024). Beyond primary visual cortex: The leading role of lateral occipital complex in early conscious visual processing. NeuroImage , 298 , 120805. https://doi.org/10.1016/j.neuroimage.2024.120805 Colombari, E., & Railo, H. (2024). Multiple independent components contribute to event-related potential correlates of conscious vision. Consciousness and Cognition , 126 , 103785. https://doi.org/10.1016/j.concog. 2024.103785 de Graaf, T. A., Hsieh, P.-J., & Sack, A. T. (2012). The ‘Correlates’ in Neural Correlates of Consciousness. Neuroscience & Biobehavioral Reviews , 36 (1), 191–197. https://doi.org/10.1016/j.neubiorev.2011.05.012 Dehaene, S., Changeux, J.-P., Naccache, L., Sackur, J., & Sergent, C. (2006). Conscious, preconscious, and subliminal processing: A testable taxonomy. Trends in Cognitive Sciences , 10 (5), 204–211. https://doi.org/ 10.1016/j.tics.2006.03.007 Dellert, T., Krebs, S., Bruchmann, M., Schindler, S., Peters, A., & Straube, T. (2022). Neural correlates of consciousness in an attentional blink paradigm with uncertain target relevance. NeuroImage , 264 , 119679. https://doi.org/10.1016/j.neuroimage.2022.119679 Dellert, T., Müller-Bardorff, M., Schlossmacher, I., Pitts, M., Hofmann, D., Bruchmann, M., & Straube, T. (2021). Dissociating the Neural Correlates of Consciousness and Task Relevance in Face Perception Using Simultaneous EEG-fMRI. The Journal of Neuroscience , 41 (37), 7864–7875. https://doi.org/10.1523/ JNEUROSCI.2799-20.2021 Dembski, C., Koch, C., & Pitts, M. (2021). Perceptual awareness negativity: A physiological correlate of sensory consciousness. Trends in Cognitive Sciences , 25 (8), 660–670. https://doi.org/10.1016/j.tics.2021.05.009 Derda, M., Koculak, M., Windey, B., Gociewicz, K., Wierzchon, M., Cleeremans, A., & Binder, M. (2019). The´ role of levels of processing in disentangling the ERP signatures of conscious visual processing. Consciousness and Cognition , 73 , 102767. https://doi.org/10.1016/j.concog.2019.102767 Dienes, Z. (2008). Understanding psychology as a science: An introduction to scientific and statistical inference . Palgrave Macmillan. Dienes, Z. (2016). How Bayes factors change scientific practice. Journal of Mathematical Psychology , 72 , 78–89. https://doi.org/10.1016/j.jmp.2015.10.003 Dykstra, A. R., Cariani, P. A., & Gutschalk, A. (2017). A Roadmap for the Study of Conscious Audition and Its Neural Basis. Phil. Trans. R. Soc. B , 372 (1714), 20160103. https://doi.org/10.1098/rstb.2016.0103 Eiserbeck, A., Enge, A., Rabovsky, M., & Abdel Rahman, R. (2021). Electrophysiological Chronometry of Graded Consciousness during the Attentional Blink. Cerebral Cortex , bhab289. https://doi.org/10.1093/cercor/ bhab289 Eklund, R., Gerdfeldter, B., & Wiens, S. (2021). The early but not the late neural correlate of auditory awareness reflects lateralized experiences. Neuropsychologia , 158 , 107910. https : / / doi . org / 10 . 1016 / j . neuropsychologia.2021.107910 Eklund, R., & Wiens, S. (2018). Visual awareness negativity is an early neural correlate of awareness: A preregistered study with two gabor sizes. Cognitive, A ff ective, & Behavioral Neuroscience , 18 (1), 176–188. https://doi.org/10.3758/s13415-018-0562-z Eklund, R., & Wiens, S. (2019). Auditory awareness negativity is an electrophysiological correlate of awareness in an auditory threshold task. Consciousness and Cognition , 71 , 70–78. https://doi.org/10.1016/j.concog. 2019.03.008 Esterman, M., & Yantis, S. (2010). Perceptual Expectation Evokes Category-Selective Cortical Activity. Cerebral Cortex , 20 (5), 1245–1253. https://doi.org/10.1093/cercor/bhp188 Fahrenfort, J. J., Johnson, P. A., Kloosterman, N. A., Stein, T., & Van Gaal, S. (2024, February). Criterion placement threatens the construct validity of neural measures of consciousness. https://doi.org/10.1101/2024. 02.22.581517 Fischer, H., Nilsson, M. E., & Ebner, N. C. (2024). Why the Single-N Design Should Be the Default in Affective Neuroscience. A ff ective Science , 5 (1), 62–66. https://doi.org/10.1007/s42761-023-00182-5 Förster, J., Koivisto, M., & Revonsuo, A. (2020). ERP and MEG correlates of visual consciousness: The second decade. Consciousness and Cognition , 80 , 102917. https://doi.org/10.1016/j.concog.2020.102917 Förster, J., Nierhaus, T., Schröder, P., & Blankenburg, F. (2025). Perceptual experience in somatosensory temporal discrimination is indexed by a mid-latency fronto-central ERP difference [Publisher: Nature Publishing Group]. Scientific Reports , 15 (1), 7674. https://doi.org/10.1038/s41598-025-91580-1 Franke, M., & Roettger, T. B. (2019, July). Bayesian regression modeling (for factorial designs): A tutorial (tech. rep.) (type: article). PsyArXiv. https://doi.org/10.31234/osf.io/cdxv3 Frith, C., Perry, R., & Lumer, E. (1999). The neural correlates of conscious experience: An experimental framework. Trends in Cognitive Sciences , 3 (3), 105–114. https://doi.org/10.1016/S1364-6613(99)01281-4 Gerdfeldter, B., Andersson, A., & Wiens, S. (2024). Examining the lateralization of electrophysiological correlates of auditory awareness. Psychophysiology , 61 (11), e14656. https://doi.org/10.1111/psyp.14656 Gerdfeldter, B., Greiciˇ ut¯ e, M., & Wiens, S. (2025). Examining experienced lateralization of sounds over head-˙ phones with electroencephalography. Neuropsychologia . https://doi.org/https://doi.org/10.1016/j. neuropsychologia.2024.109064 Gramfort, A., Luessi, M., Larson, E., Engemann, D. A., Strohmeier, D., Brodbeck, C., Goj, R., Jas, M., Brooks, T., Parkkonen, L., & Hämäläinen, M. (2013). MEG and EEG data analysis with MNE-python. Frontiers in Neuroscience , 7 , 1–13. https://doi.org/10.3389/fnins.2013.00267 Gramfort, A., Luessi, M., Larson, E., Engemann, D. A., Strohmeier, D., Brodbeck, C., Parkkonen, L., & Hämäläinen, M. S. (2014). MNE software for processing MEG and EEG data. NeuroImage , 86 , 446–460. https: Crossref Google Scholar 2. Gwilliams, L., & King, J.-R. (2020). Recurrent processes support a cascade of hierarchical decisions (T. Serre & M. J. Frank, Eds.) [Publisher: eLife Sciences Publications, Ltd]. eLife , 9 , e56603. https://doi.org/10.7554/ eLife.56603 Heinzel, A., Hautzel, H., Poeppel, T., Boers, F., Beu, M., & Mueller, H.-W. (2008). Neural correlates of subliminal and supraliminal letter processing—An event-related fMRI study. Consciousness and Cognition , 17 (3), 699–713. https://doi.org/10.1016/j.concog.2008.01.008 Hense, A., Peters, A., Bruchmann, M., Dellert, T., & Straube, T. (2024). Electrophysiological correlates of sustained conscious perception [Publisher: Nature Publishing Group]. Scientific Reports , 14 (1), 10593. https: //doi.org/10.1038/s41598-024-61281-2 Hutchinson, B. T. (2019). Toward a theory of consciousness: A review of the neural correlates of inattentional blindness. Neuroscience & Biobehavioral Reviews , 104 , 87–99. https://doi.org/10.1016/j.neubiorev.2019. 06.003 Jimenez, M., Grassini, S., Montoro, P. R., Luna, D., & Koivisto, M. (2018). Neural correlates of visual awareness at stimulus low vs. high-levels of processing. Neuropsychologia , 121 , 144–152. https://doi.org/10.1016/j. neuropsychologia.2018.11.001 Jimenez, M., Hinojosa, J. A., & Montoro, P. R. (2020). Visual awareness and the levels of processing hypothesis: A critical review. Consciousness and Cognition , 85 , 103022. https://doi.org/10.1016/j.concog.2020.103022 Jimenez, M., Poch, C., Villalba-García, C., Sabater, L., Hinojosa, J. A., Montoro, P. R., & Koivisto, M. (2021). The Level of Processing Modulates Visual Awareness: Evidence from Behavioral and Electrophysiological Measures. Journal of Cognitive Neuroscience , 1–16. https://doi.org/10.1162/jocn_a_01712 Kanai, R., Walsh, V., & Tseng, C.-h. (2010). Subjective discriminability of invisibility: A framework for distinguishing perceptual and attentional failures of awareness. Consciousness and Cognition , 19 (4), 1045– 1057. https://doi.org/10.1016/j.concog.2010.06.003 Kim, C.-Y., & Blake, R. (2005). Psychophysical Magic: Rendering the Visible ‘Invisible’. Trends in Cognitive Sciences , 9 (8), 381–388. https://doi.org/10.1016/j.tics.2005.06.012 Knight, R. S., Chen, T., Center, E. G., Gratton, G., Fabiani, M., Savazzi, S., Mazzi, C., & Beck, D. M. (2024). Bypassing input to V1 in visual awareness: A TMS-EROS investigation. Neuropsychologia , 198 , 108864. https://doi.org/10.1016/j.neuropsychologia.2024.108864 Koch, C., Massimini, M., Boly, M., & Tononi, G. (2016). Neural correlates of consciousness: Progress and problems. Nature Reviews Neuroscience , 17 (5), 307–321. https://doi.org/10.1038/nrn.2016.22 Koivisto, M., Grassini, S., Salminen-Vaparanta, N., & Revonsuo, A. (2017). Different electrophysiological correlates of visual awareness for detection and identification. Journal of Cognitive Neuroscience , 29 (9), 1621–1631. https://doi.org/10.1162/jocn_a_01149 Koivisto, M., Kainulainen, P., & Revonsuo, A. (2009). The relationship between awareness and attention: Evidence from erp responses. Neuropsychologia , 47 (13), 2891–2899. https://doi.org/10.1016/j.neuropsychologia. 2009.06.016 Koivisto, M., & Revonsuo, A. (2008). The role of selective attention in visual awareness of stimulus features: Electrophysiological studies. Cognitive, A ff ective, & Behavioral Neuroscience , 8 (2), 195–210. https://doi. org/10.3758/CABN.8.2.195 Koivisto, M., & Revonsuo, A. (2010). Event-Related Brain Potential Correlates of Visual Awareness. Neuroscience & Biobehavioral Reviews , 34 (6), 922–934. https://doi.org/10.1016/j.neubiorev.2009.12.002 Koivisto, M., Revonsuo, A., & Salminen, N. (2005). Independence of visual awareness from attention at early processing stages. NeuroReport , 16 (8), 817. https://doi.org/10.1097/00001756-200505310-00008 Kreiman, G., & Serre, T. (2020). Beyond the feedforward sweep: Feedback computations in the visual cortex. Annals of the New York Academy of Sciences , 1464 (1), 222–241. https://doi.org/10.1111/nyas.14320 Kretzschmar, F., & Alday, P. M. (2020, April). Principles of statistical analyses: Old and new tools (preprint). PsyArXiv. https://doi.org/10.31234/osf.io/nyj3k Kronemer, S. I., Aksen, M., Ding, J. Z., Ryu, J. H., Xin, Q., Ding, Z., Prince, J. S., Kwon, H., Khalaf, A., Forman, Crossref Google Scholar 3. Lakens, D., Hilgard, J., & Staaks, J. (2016). On the reproducibility of meta-analyses: Six practical recommendations. BMC Psychology , 4 (1), 24. https://doi.org/10.1186/s40359-016-0126-3 Lamme, V. A. F. (2006). Towards a true neural stance on consciousness. Trends in Cognitive Sciences , 10 (11), 494–501. https://doi.org/10.1016/j.tics.2006.09.001 Lamme, V. A. F. (2010). How Neuroscience Will Change Our View on Consciousness. Cognitive Neuroscience , 1 (3), 204–220. https://doi.org/10.1080/17588921003731586 Lamme, V. A. F. (2018). Challenges for Theories of Consciousness: Seeing or Knowing, the Missing Ingredient and How to Deal with Panpsychism. Phil. Trans. R. Soc. B , 373 (1755), 20170344. https://doi.org/10. 1098/rstb.2017.0344 Lamme, V. A. F., Supèr, H., Landman, R., Roelfsema, P. R., & Spekreijse, H. (2000). The Role of Primary Visual Crossref Google Scholar 4. Lamme, V. (2015). The Crack of Dawn: Perceptual Functions and Neural Mechanisms that Mark the Transition from Unconscious Processing to Conscious Vision [Publisher: [object Object]]. Open MIND . https: //doi.org/10.15502/9783958570092 Lange, F. P. d., Heilbron, M., & Kok, P. (2018). How Do Expectations Shape Perception? [Publisher: Elsevier]. Trends in Cognitive Sciences , 22 (9), 764–779. https://doi.org/10.1016/j.tics.2018.06.002 Liu, H., Agam, Y., Madsen, J. R., & Kreiman, G. (2009). Timing, Timing, Timing: Fast Decoding of Object Information from Intracranial Field Potentials in Human Visual Cortex [Publisher: Elsevier]. Neuron , 62 (2), 281–290. https://doi.org/10.1016/j.neuron.2009.02.025 Liu, Y., Paradis, A.-L., Yahia-Cherif, L., & Tallon-Baudry, C. (2012). Activity in the lateral occipital cortex between 200 and 300 ms distinguishes between physically identical seen and unseen stimuli. Frontiers in Human Neuroscience , 6 . https://doi.org/10.3389/fnhum.2012.00211 Luck, S. J. (2014). An Introduction to the Event-Related Potential Technique (Second edition). The MIT Press. MacLean, M. W., Hadid, V., Spreng, R. N., & Lepore, F. (2023). Revealing robust neural correlates of conscious and unconscious visual processing: Activation likelihood estimation meta-analyses. NeuroImage , 273 , 120088. https://doi.org/10.1016/j.neuroimage.2023.120088 Makin, T. R., & Orban de Xivry, J.-J. (2019). Ten common statistical mistakes to watch out for when writing or reviewing a manuscript. eLife , 8 , e48175. https://doi.org/10.7554/eLife.48175 Makowski, D., Ben-Shachar, M., & Lüdecke, D. (2019). bayestestR: Describing Effects and their Uncertainty, Existence and Significance within the Bayesian Framework. Journal of Open Source Software , 4 (40), 1541. https://doi.org/10.21105/joss.01541 Makowski, D., Ben-Shachar, M. S., Chen, S. H. A., & Lüdecke, D. (2019). Indices of Effect Existence and Significance in the Bayesian Framework. Frontiers in Psychology , 10 , 2767. https://doi.org/10.3389/fpsyg. 2019.02767 Matta, T. H., Flournoy, J. C., & Byrne, M. L. (2018). Making an unknown unknown a known unknown: Missing data in longitudinal neuroimaging studies. Developmental Cognitive Neuroscience , 33 , 83–98. https: //doi.org/10.1016/j.dcn.2017.10.001 Meyer, K. (2011). Primary Sensory Cortices, Top-down Projections and Conscious Experience. Progress in Neurobiology , 94 (4), 408–417. https://doi.org/10.1016/j.pneurobio.2011.05.010 Munafò, M. R., Nosek, B. A., Bishop, D. V. M., Button, K. S., Chambers, C. D., Sert, N. P. d., Simonsohn, U., Wagenmakers, E.-J., Ware, J. J., & Ioannidis, J. P. A. (2017). A manifesto for reproducible science [Number: 1 Publisher: Nature Publishing Group]. Nature Human Behaviour , 1 (1), 1–9. https://doi.org/10. 1038/s41562-016-0021 O’Connell, R. G., & Kelly, S. P. (2021). Neurophysiology of Human Perceptual Decision-Making. Annual Review of Neuroscience , 44 (1), 495–516. https://doi.org/10.1146/annurev-neuro-092019-100200 Peirce, J., Gray, J. R., Simpson, S., MacAskill, M., Höchenberger, R., Sogo, H., Kastman, E., & Lindeløv, J. K. (2019). PsychoPy2: Experiments in behavior made easy. Behavior Research Methods , 51 (1), 195–203. https://doi.org/10.3758/s13428-018-01193-y Pitts, M. A., Martínez, A., & Hillyard, S. A. (2012). Visual Processing of Contour Patterns under Conditions of Inattentional Blindness. Journal of Cognitive Neuroscience , 24 (2), 287–303. https://doi.org/10.1162/ jocn_a_00111 Pitts, M. A., Metzler, S., & Hillyard, S. A. (2014). Isolating Neural Correlates of Conscious Perception from Neural Correlates of Reporting One’s Perception. Frontiers in Psychology , 5 . https://doi.org/10.3389/ fpsyg.2014.01078 Quarto: An Open-Source Scientific and Technical Publishing System . (2024). https://quarto.org R core Team. (2016). R: A language and environment for statistical computing. Retrieved August 1, 2019, from https://www.R-project.org/ Ramsøy, T. Z., & Overgaard, M. (2004). Introspection and Subliminal Perception. Phenomenology and the Cognitive Sciences , 3 (1), 1–23. https://doi.org/10.1023/B:PHEN.0000041900.30172.e8 Rodríguez-San Esteban, P., Gonzalez-Lopez, J. A., & Chica, A. B. (2025). Neural representation of consciously seen and unseen information [Publisher: Nature Publishing Group]. Scientific Reports , 15 (1), 7888. https: //doi.org/10.1038/s41598-025-92490-y Roth-Paysen, M.-L., Bröcker, A., Bruchmann, M., & Straube, T. (2022). Early and late electrophysiological correlates of gradual perceptual awareness in- and outside the Attentional Blink window. NeuroImage , 263 , 119652. https://doi.org/10.1016/j.neuroimage.2022.119652 RStudio Team. (2020). Rstudio: Integrated development environment for R. http://www.rstudio.com Sandberg, K., & Overgaard, M. (2015, March). Using the perceptual awareness scale (PAS) [ef]. In M. Overgaard (Ed.), Behavioral Methods in Consciousness Research (pp. 181–196). Oxford University Press. Sassenhagen, J., & Alday, P. M. (2016). A common misapplication of statistical inference: Nuisance control with null-hypothesis significance tests. Brain and Language , 162 , 42–45. https://doi.org/10.1016/j.bandl.2016. 08.001 Schlossmacher, I., Dellert, T., Pitts, M., Bruchmann, M., & Straube, T. (2020). Differential Effects of Awareness and Task Relevance on Early and Late ERPs in a No-Report Visual Oddball Paradigm. The Journal of Neuroscience , 40 (14), 2906–2913. https://doi.org/10.1523/JNEUROSCI.2077-19.2020 Schröder, P., Nierhaus, T., & Blankenburg, F. (2021). Dissociating Perceptual Awareness and Postperceptual Processing: The P300 Is Not a Reliable Marker of Somatosensory Target Detection [Publisher: Society for Neuroscience Section: Research Articles]. Journal of Neuroscience , 41 (21), 4686–4696. https://doi.org/ 10.1523/JNEUROSCI.2950-20.2021 Scrivener, C. L., Malik, A., Lindner, M., & Roesch, E. B. (2021). Sensing and seeing associated with overlapping occipitoparietal activation in simultaneous EEG-fMRI. Neuroscience of Consciousness , 2021 (1), niab008. https://doi.org/10.1093/nc/niab008 Scrivener, C. L., Malik, A., Marsh, J., Lindner, M., & Roesch, E. B. (2019). An EEG study of detection without localisation in change blindness [Company: Springer Distributor: Springer Institution: Springer Label: Springer Number: 10 Publisher: Springer Berlin Heidelberg]. Experimental Brain Research , 237 (10), 2535–2547. https://doi.org/10.1007/s00221-019-05602-2 Snyder, J. S., Yerkes, B. D., & Pitts, M. A. (2015). Testing Domain-General Theories of Perceptual Awareness with Auditory Brain Responses. Trends in Cognitive Sciences , 19 (6), 295–297. https://doi.org/10.1016/j. tics.2015.04.002 Straube, S., & Fahle, M. (2011). Visual detection and identification are not the same: Evidence from psychophysics and fMRI. Brain and Cognition , 75 (1), 29–38. https://doi.org/10.1016/j.bandc.2010.10.004 Tagliabue, C. F., Mazzi, C., Bagattini, C., & Savazzi, S. (2016). Early local activity in temporal areas reflects graded content of visual perception. Frontiers in Psychology , 7 . https://doi.org/10.3389/fpsyg.2016.00572 Tagliabue, C. F., Veniero, D., Benwell, C. S. Y., Cecere, R., Savazzi, S., & Thut, G. (2019). The EEG signature of sensory evidence accumulation during decision formation closely tracks subjective perceptual experience. Scientific Reports , 9 (1), 4949. https://doi.org/10.1038/s41598-019-41024-4 Twomey, D. M., Murphy, P. R., Kelly, S. P., & O’Connell, R. G. (2015). The classic P300 encodes a build-tothreshold decision variable. European Journal of Neuroscience , 42 (1), 1636–1643. https://doi.org/10. 1111/ejn.12936 Wagenmakers, E.-J., Love, J., Marsman, M., Jamil, T., Ly, A., Verhagen, J., Selker, R., Gronau, Q. F., Dropmann, Crossref Google Scholar 5. Wagenmakers, E.-J., Morey, R. D., & Lee, M. D. (2016). Bayesian benefits for the pragmatic researcher. Current Directions in Psychological Science , 25 (3), 169–176. https://doi.org/10.1177/0963721416643289 Wasserstein, R. L., & Lazar, N. A. (2016). The ASA statement on p -values: Context, process, and purpose. The American Statistician , 70 (2), 129–133. https://doi.org/10.1080/00031305.2016.1154108 Wiens, S. (2025). Open data: EEG correlates of detection and identification awareness for digits and letters. https: Crossref Google Scholar 6. Wiens, S., Andersson, A., & Gravenfors, J. (2023). Neural electrophysiological correlates of detection and identification awareness. Cognitive, A ff ective, & Behavioral Neuroscience , 23 (5), 1303–1321. https://doi.org/ 10.3758/s13415-023-01120-5 Wiens, S., & Nilsson, M. E. (2017). Performing contrast analysis in factorial designs: From NHST to confidence intervals and beyond. Educational and Psychological Measurement , 77 (4), 690–715. https://doi.org/10. 1177/0013164416668950 Wilenius, M. E., & Revonsuo, A. T. (2007). Timing of the Earliest ERP Correlate of Visual Awareness. Psychophysiology , 44 (5), 703–710. https://doi.org/10.1111/j.1469-8986.2007.00546.x Windey, B., Vermeiren, A., Atas, A., & Cleeremans, A. (2014). The graded and dichotomous nature of visual awareness. Philosophical Transactions of the Royal Society B: Biological Sciences , 369 (1641), 20130282. Crossref Google Scholar Information & Authors Information Version history V1 Version 1 02 July 2025 Peer review timeline Published Neuropsychologia Version of Record 1 Feb 2026 Published Copyright This work is licensed under a Non Exclusive No Reuse License. Authors Affiliations Stefan Wiens 0000-0003-4531-4313 [email protected] Stockholm universitet Psykologiska institutionen View all articles by this author Mingailė Greičiūtė Stockholm universitet Psykologiska institutionen View all articles by this author Billy Gerdfeldter 0000-0002-3222-8056 Stockholm universitet Psykologiska institutionen View all articles by this author Annika Andersson Stockholm universitet Psykologiska institutionen View all articles by this author Metrics & Citations Metrics Article Usage 251 views 115 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Stefan Wiens, Mingailė Greičiūtė, Billy Gerdfeldter, et al. Electrophysiological correlates of detection and identification awareness for digits and letters. Authorea . 02 July 2025. DOI: https://doi.org/10.22541/au.175143836.64282168/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.175143836.64282168/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'a00cab22bd5558d3',t:'MTc3OTYyOTU2OA=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();