EEG synchronisation reveals the impact of group identity and membership duration on social cognitive bias

Abstract

Group identity induces social cognitive biases, and membership duration may amplify these effects. This study aimed to examine such bias by analysing similarities in neural processing among individuals in competitive scenarios. The fans of two Japanese baseball teams, the Hanshin Tigers and Orix Buffaloes, watched baseball matches between the teams, and EEG synchronisation was analysed for in-group (same team) and out-group (different team) pairs, considering fan history as a factor representing membership duration. The results revealed that in-group pairs showed stronger centroparietal alpha-phase synchronisation than out-group pairs, suggesting that top-down spatial attention modulated early visual processing in a similar way among in-group members. Furthermore, in-group pairs with longer fan histories exhibited higher parietal alpha power synchronisation, probably reflecting shared engagement and emotional responses, whereas this effect was absent in out-group pairs. Interestingly, longer fan histories were associated with reduced parietal delta and theta phase synchronisations, possibly due to diverse late-stage attentional processes among experienced fans. Additionally, frontal alpha-phase synchronisation increased with fan history, indicating enhanced auditory attention in long-term fans. These findings highlight how group identity and membership duration shape neural processing, and EEG synchronisation analysis provides a robust method for examining biases.

Introduction

Social division is a critical issue in modern society. One source of such intergroup conflicts is the psychological tendency arising from the relationship between self and others¹; people tend to favour members of their own group (i.e. in-group), while being less accepting of those from other groups (i.e. out-group). The asymmetry in attitudes toward in-groups and out-groups has been a long-standing subject of study in social psychology. Social identity theory² posits that individuals derive a sense of identity from their group memberships, which influences their cognition and emotions. As people categorise themselves as part of a group, they are motivated to view their group positively to maintain or enhance self-esteem, leading to in-group favouritism and, at times, out-group derogation. This attitude is reflected in cognitive bias, whereby the same event is perceived in opposite ways by in-group and out-group members. For example, in team sports, a team’s score is viewed positively by its fans but negatively by rival team supporters.

Several neuroscientific studies have demonstrated that social identity shapes neural responses associated with value-evaluation systems. For instance, in gambling tasks, feedback-related negativity (FRN) is elicited by the losses of cooperative others and gains of adversarial others³. FRN is an event-related brain potential (ERP) that reflects evaluation of an event⁴. Furthermore, when devoted baseball fans watched an animation featuring key moments in a game between their favourite and rival teams, functional magnetic resonance imaging (fMRI) scans revealed distinct patterns: the rival team’s success activated the anterior cingulate cortex (ACC) and insula, while their failure triggered activity in the ventral striatum⁵. The ACC and insula are commonly activated when experiencing negative emotions or pain^6,7whereas the ventral striatum is associated with reward processing⁸.

While such research highlights the impact of social identity on neural processing, traditional methods have limitations in studying cognitive biases in social situations. In conventional frameworks of cognitive neuroscience, including the studies mentioned above, a model-based approach is typically employed to interpret the observed brain activity⁹. This approach focuses on specific neural processes and examines how experimental manipulation influences these processes. To reliably capture brain responses, specific experimental paradigms (e.g. gambling tasks or simplified animations) are used, consisting of repetitive short trials. However, social cognitive biases are often most pronounced in dynamic, competitive scenarios such as the progression of a sports match or political debate. These situations involve complex interactions that unfold over time, and the context shapes the significance of each event. For instance, a positive event has a different meaning depending on whether one’s group is in a disadvantaged or superior position. Such complexities cannot be adequately captured using traditional experimental paradigms. Addressing this limitation requires the use of more natural stimuli. Moreover, the neural processes influenced by group identity extend beyond value-evaluation systems and affect perception and emotion. For example, the recognition of whether a person belongs to the in-group or out-group influences N170, which is an ERP reflecting face perception¹⁰. The amygdala, a brain region tightly associated with emotion processing, shows differential activity when viewing faces of racial in-group versus out-group members¹¹. Investigating such individual cognitive processes typically requires tasks specifically designed for the purpose (e.g. facial presentations) and analyses targeting particular brain regions or responses. Consequently, model-based approaches struggle to capture the cognitive biases occurring across various levels of brain processing in the same social setting. To overcome these challenges, this study used naturalistic stimuli and analysed the inter-subject correlation (ISC) of the EEG during the stimuli.

ISC analysis identifies stimulus-driven neural dynamics that are shared across people⁹. This involves presenting the same stimulus to each individual, measuring their brain activity during this period, and quantifying the similarities in their neural time series. The logic of ISC is as follows⁹: the measured brain activity over time can be divided into three components—(1) neural activity reflecting stimulus processing shared across participants, (2) neural activity reflecting stimulus processing unique to each individual, and (3) spontaneous neural firing. Among these, shared processing (1) is reflected in the similarity of brain activity between individuals, and involves not only low-level sensory perception but also higher-level processes such as emotional and semantic interpretation^12,13. ISC analysis is a data-driven approach for identifying brain activity that reflects shared processing; thus, it does not rely on predefined assumptions regarding brain responses to stimuli⁹. This model-free nature of ISC makes it possible to use naturalistic, extended-duration stimuli, such as videos and narratives¹⁴and allows us to investigate the shared neural processing in such real-world contexts. In practice, several studies have revealed brain activity synchronisation during naturalistic stimuli by ISC using various imaging modalities, such as fMRI, magnetoencephalography (MEG), and EEG^{12,13,15, 16, 17, 18, 19, 20, 21, 22, 23–24}. For example, during an fMRI scan while watching a movie clip, the subjects’ brain activity showed correlations not only in the visual and auditory areas but also across broad regions, including association areas¹¹. The EEG of the test subjects became more synchronised during scenes of a popular TV drama that received many reactions on social media²⁰.

In recent years, ISC analysis has been applied not only to assess similarities but also to explore dissimilarities in brain activities associated with individual traits²⁵. In subject-unique neural processing (2), which is one of the three components mentioned above, parts that vary systematically with the participant’s characteristics can be identified by analysing the relationship between neural synchronisation and those characteristics²⁵. For example, the ISC of MEG power during speech perception differs between typically developing individuals and those with dyslexia²⁶. Furthermore, the degree of similarity in fMRI signal patterns while watching movie clips predicts proximity within social networks²⁷.

In the present study, we conducted an ISC analysis of EEG data during social competition scenarios to investigate cognitive biases influenced by group identity and assessed the effect of membership duration on bias. Previous research has reported that the amount of time spent as a group member is positively associated with the strength of group identity^28,29suggesting that membership duration may be an important factor which potentially intensifies bias. To the best of our knowledge, this factor has been overlooked. To achieve this goal, we measured fans’ EEGs while watching baseball games between two competing teams. As previously noted, the ISC framework allows the use of naturalistic and complex stimuli. Furthermore, EEG is more easily measured than fMRI or MEG using simple devices. Therefore, a combination of ISC analysis and EEG measurements makes it feasible to study brain activity during experiences that closely resemble real-life situations. ISC can reveal the similarities and differences in brain activity that reflect individual characteristics. When applied to social cognitive biases, this approach suggests that members of the same group share the same cognitive processes. In contrast, members of different groups may display distinct brain activities as they interpret the same situation from different perspectives. Based on this, we hypothesised that the degree of neural synchronisation would be greater among in-group members than among out-group members. Additionally, we propose that, within the in-group, a longer membership duration correlates with a higher degree of synchronisation, whereas this effect is not evident in the out-group. The first aim of this study was to examine whether EEG signals exhibit synchronisation patterns that align with these hypotheses. In the experiment, the stimuli were videos of matches between two Japanese professional baseball teams—the Hanshin Tigers and the Orix Buffaloes. Fans from both teams were recruited; they viewed the videos individually, and their EEGs were recorded. Subsequently, the data from the two participants were paired, and the degree of synchronisation between their EEGs was analysed. Pairs in which both participants supported the same team (e.g. Hanshin Tigers fans paired with Hanshin Tigers fans, or Orix Buffaloes fans paired with Orix Buffaloes fans) were categorised as in-group pairs. The remaining pairs in which participants supported different teams (e.g. a Hanshin Tigers fan paired with an Orix Buffaloes fan) were categorised as out-group pairs. The membership duration for each pair was determined by the shorter fan histories (unit: years) of the two participants. The second aim was to determine which cognitive processes are related to bias, considering the features of brain activity where the effects of group identity and membership duration were observed. The ISC reflects similarities in the brain activity associated with the same social setting, encompassing not only low-level sensory perceptions but also higher-level cognitive processes. This interpretation provides valuable insights for future research on the neural mechanisms underlying social cognitive biases.

Methods

Participants

Thirty-four individuals participated in this study, of which data from 32 participants (17 females, mean age: 32.8, range: 20–48) were analysed (but for ECG analyses, we analysed data from 30 participants. See ECG analyses section for details), after excluding one participant who withdrew from the experiment and one whose data contained irreparable errors. Among these participants, 18 were fans of the Hanshin Tigers and the remaining 14 were fans of the Orix Buffaloes. This study, approved by the Ethics Committee of the National Institute of Information and Communications Technology, was conducted in accordance with the principles of the Declaration of Helsinki. Prior to the experiment, the participants were provided with an explanation of the study. Informed consent was obtained from all participants, using a form approved by the committee.

Stimuli

Figure 1 shows a photograph of the actual laboratory setting. The experimental stimuli were presented on a wall (width: 4 m; height: 2.25 m) using a projector (EH-TW8300, Seiko Epson Corporation, Nagano, Japan) with audio playback delivered through an audio system (LIFESTYLE 650, Bose Corporation, Framingham, USA). The participants sat on a chair approximately 4 m from the wall to watch the videos. The stimuli comprised three videos automatically generated via computer-controlled gameplay of a baseball computer game. This setting was to rule out any games that the participants may have already watched when using videos of real matches. The computer game used was the Professional baseball spirits 2019 (Konami Digital Entertainment Co., Ltd., Tokyo, Japan), which features real-life teams and players and utilises highly detailed 3D graphics to closely replicate the characteristics of actual players and the atmosphere of real baseball matches. The audio conveyed the audience’s cheers and the commentators’ commentary in alignment with the game’s progression.

The selected stimuli were games between the Hanshin Tigers and Orix Buffaloes. From a large pool of automatically generated videos, the experimenters selected three videos with exciting dynamics (e.g. high hit counts) to maximise engagement. These videos depicted a game won by the Hanshin Tigers, a game won by the Orix Buffaloes, and a draw. In baseball, a standard game consists of nine innings, with full-length matches in a video game lasting over an hour. To accommodate time constraints, this study used footage from the top of the 6th inning. To ensure that participants understood the context of the matches, each video was preceded by a 30-second still image summarising the progress of the game up to the end of the 5th inning. The duration of the videos was as follows: 28 min 58 s for the Hanshin Tigers victory, 25 min 48 s for the draw, and 33 min 14 s for the Orix Buffaloes victory.

[See PDF for image]

Fig. 1

The laboratory setup. The participants watched baseball games on a large screen that provided an immersive experience. The screen in this figure is a composite image. Although the lighting is bright in this figure to make the laboratory setting easier to understand, it was dimmed to create a dark environment for video projection during the actual experiment.

Procedure

Each participant was seated in a comfortable chair in front of the screen and informed about the structure of the videos (e.g. they started with the top of the sixth inning) and that they would watch three approximately 30-minute baseball game videos. They were not informed which team would win each match. Physiological measurement devices were attached to the participants, and their recordings were started. The participants were asked to complete questionnaires before, between, and after viewing the videos (see the section below for details of each questionnaire). Between each video, the participants were offered a short break if desired. The order of videos varied for each participant.

Questionnaires and the analysis

In the questionnaire, before watching the videos, participants were asked about their history as fans of their preferred teams. This included the number of years they had been fans and the number of games of their favoured team they had watched in the past year (2023). They were also asked how much they would favour the rival team (e.g. Hanshin Tigers fans and Orix Buffaloes) if they watched a match between the rival team and another team (excluding games between the Hanshin Tigers and Orix Buffaloes). Responses were recorded on a 7-point scale ranging from “strongly cheer” to “neutral” to “strongly cheer the opponent team”. Additionally, participants rated their current emotional state: valence (7-point scale: “very negative” to “neutral” to “very positive”) and arousal level (7-point scale: “very calm” to “neutral” to “very excited”).

Immediately after watching each video, the participants answered the same questions regarding their emotional states. They were also asked how much they enjoyed the game they had just watched, using a 7-point scale ranging from “did not enjoy it at all” to “neutral” to “enjoyed it very much”.

After completing all video viewing, participants were asked about their familiarity with the video game used in the experiment, whether they understood the unique information displayed in the game (e.g. player status parameters), and whether these elements drew their attention during viewing. However, these responses were not included in the analysis. To assess the extent of each participant’s fandom, multiple-choice quizzes were administered: 20 questions about Hanshin Tigers and Orix Buffaloes, each offering four possible answers. One-sample t-tests revealed that the quiz accuracy rates for Hanshin Tigers fans (mean = 46.7%, SE = 2.55%) and Orix Buffaloes fans (mean = 49.3%, SE = 3.74%) were significantly higher than the chance level of 25% (ps < 0.001). No significant difference was noted in the quiz performance between Hanshin Tigers and Orix Buffaloes fans (p = .57). These results indicate that both groups had substantial knowledge of baseball, confirming their enthusiasm for the sport.

For statistical tests of emotional valences, arousal levels, and enjoyment scores, permutation tests (with 100,000 iterations to generate the null hypothesis distribution) were conducted separately for Hanshin Tigers and Orix Buffaloes fans. Specifically, pairwise comparisons were made between the video types (Hanshin victory vs. Orix victory, Hanshin victory vs. draw, and Orix victory vs. draw). To address multiple comparisons across the six tests, p-values were adjusted using False Discovery Rate (FDR) correction. The significance threshold was set at p < .05.

Physiological measurement

The EEG, electrooculogram (EOG), and electrocardiogram (ECG) were recorded using a custom 8-channel wearable measurement system based on the Polymate Mini AP108 (Miyuki Giken Co. Ltd., Tokyo, Japan; see Yokota et al.³⁰ for details). This measurement system allows participants to move their heads freely and helps reduce fatigue, owing to the minimal number of electrodes, use of dry electrodes, and wireless data transmission. In this study, we employed this system to enable participants to watch videos in a relaxed setting that resembled real-life conditions. EEG electrodes were placed at the scalp locations Fp1, Fp2, Fz, Cz, and Pz, whereas EOG electrodes were positioned above the left eyebrow (vertical EOG) and on the lateral side of the left outer canthus (horizontal EOG). An ECG electrode was placed below the left clavicle. A reference electrode was attached to the left mastoid and a ground electrode to the right mastoid. The recordings were performed at a sampling rate of 500 Hz. In addition, participants wore eyeglass-type eye trackers (Pupil Core, Pupil Labs GmbH, Berlin, Germany) to measure their gaze and pupil diameters. Eye-tracking data were recorded with the intention of incorporating them into future analyses alongside other experimental measures. However, given that the present study focuses on neural synchrony, to ensure clarity and avoid unnecessary complexity, we chose to exclude the eye-tracking results in this paper.

ECG analyses

R-peak timings were extracted from the ECG data recorded during each movie using the Pan-Tompkins algorithm as implemented in HEPLAB (an EEGLAB plug-in)³¹and they were checked by visual inspection and corrected if necessary. Data from two participants (both Orix Buffaloes fans) were excluded from the analysis due to unusable recordings during the Hanshin victory video. Thus, data from 30 participants (18 Hanshin Tigers and 12 Orix Buffaloes fans) were used for the rest of ECG analyses. Based on these peaks, we calculated the heart rate (HR: beats per minute) and low-frequency to high-frequency (LF/HF) ratio for each movie. High sympathetic activity increases HR³². The LF/HF ratio represents the balance between sympathetic and parasympathetic nervous system activity, with higher values indicating greater sympathetic dominance³³. To calculate the LF/HF ratio, power spectral density was estimated from interpolated inter-beat interval (IBI) time series. Low-frequency (LF: 0.04–0.15 Hz) and high-frequency (HF: 0.15–0.40 Hz) components were calculated by integrating the power spectral density within each band.

For statistical analysis, a mixed design analysis of variance (ANOVA) was performed with HR and LF/HF ratio as dependent variables. In this ANOVA, fan group (Hanshin Tigers vs. Orix Buffaloes fans) was treated as a between-subjects factor, and video type (Hanshin victory, draw, and Orix victory) as a within-subjects factor.

EEG analyses

For preprocessing, the EEG data from each video were filtered using a finite impulse response (FIR) band-pass filter (1–40 Hz, order: 5000th). The filtered data from all videos were appended, and an independent component analysis (ICA) was performed. In this process, vertical and horizontal EOGs were included in the ICA decomposition along with the EEG data to enhance the separation of eye-related activity. ICA components associated with eye movements were manually identified and removed based on visual inspection, focusing on their correspondence with blink and saccade-related waveforms. Two components were removed for 31 participants, and three were removed for one participant. For the remaining EEG analyses, we concentrated on the midline electrodes Fz, Cz, and Pz, which are frequently utilised in EEG studies as they collectively capture activity from the frontal, central, and parietal regions of the brain, respectively. These electrode sites were chosen to achieve a balance between broad neural representation and minimal interference, enabling participants to view the stimuli in a naturalistic manner. Electrodes Fp1 and Fp2 were excluded from the analysis as they are located off the midline and were recorded for separate purposes. The data were then segmented into 1-second intervals for each electrode, and intervals containing activity exceeding ± 100 µV were marked as artefacts. After artefact detection, the mean rates of the unusable data were Fz: 3.6% (SD = 8.7%), Cz: 3.8% (SD = 8.6%), and Pz: 4.4% (SD = 9.4%). Previous studies investigating the ISC of EEG signals have shown that variations in cognitive states (e.g. attentional state) predominantly influence neural synchronisation within low-frequency bands, especially those below 10 Hz^20,24. Based on these findings, our study focused on lower frequency bands including the alpha band and below. The EEG data were re-segmented by video length, and activity in the delta (1–4 Hz), theta (4–8 Hz), and alpha (8–13 Hz) bands were extracted using FIR band-pass filters (order: 3300th). The filtered data are referred to as band-specific data. In this study, the EEG synchronisation indices of phase and power between the two participants were calculated as follows (Fig. 2):

• Phase synchronisation index: Band-specific data were divided into 10-second segments, excluding artefact intervals. For each segment, the data were subjected to a Hilbert transform, and the phase-locking value (PLV)³⁴ between participants was calculated using the following formula:

where and represent the instantaneous phases of the two participants at time n, and N is the number of data points in the segment. The PLV ranges from 0 to 1, with 0 indicating no synchronisation and 1 indicating perfect synchronisation. Two EEGs were considered synchronised if their phase difference remained constant over time. This approach effectively eliminates the influence of individual differences in neural response speed to stimuli when calculating synchronisation. The PLVs for all the segments across all videos were averaged to derive a representative value for phase synchronisation.

• Power synchronisation index: Envelope time-series data were calculated by applying the Hilbert transform to the band-specific data and taking the absolute values. The envelope was smoothed using an FIR low-pass filter at 0.3 Hz (order of 3300th). These protocols were performed as previously reported²⁶. The smoothed data were divided into 10-second segments, excluding artefact intervals. For each segment, the Pearson correlation coefficient (r) was calculated between the two participants, followed by Fisher’s z-transformation using the formula: . The z-transformed coefficients were averaged across all segments for all videos. Subsequently, the inverse Fisher transformation by as applied to these averaged values to compute a representative value for power synchronisation.

[See PDF for image]

Fig. 2

Process for calculating EEG synchronization indices.

The study included 496 pair combinations (i.e. a combination of selecting two from 32), which were categorised into in-group pairs (244 pairs: Hanshin Tigers fans paired with other Hanshin Tigers fans, or Orix Buffaloes fans paired with other Orix Buffaloes fans) and out-group pairs (252 pairs: one Hanshin Tigers fan paired with one Orix Buffaloes fan). For the fan history of each pair, a smaller value (unit: years) was used for the fan histories of two individuals. This indicates that both individuals in the pair have been fans for at least this number of years.

Because these paired data included overlapping participants, they were not independent of each other. To address this, permutation testing was employed³⁵. Following the procedure of multiple regression on distance matrices³⁶the effects of pair category (in-group/out-group), fan history (unit: years; see the previous paragraph for the definition), and their interaction on each synchronisation index were analysed. In this process, a multiple regression model was first fitted to the original data to estimate the regression coefficient (β) for each term. The model is specified as follows:

We did not standardise the variables, as our primary interest was whether the beta coefficients significantly differed from zero, rather than their absolute magnitudes. Since the pair category is a categorical variable, dummy variables were assigned as in-group = 1 and out-group = -1. To assess the significance, null hypothesis distributions were created by permutation of the obtained data. Specifically, the relationship between the data and their labels was randomly shuffled, and a regression model was applied to generate pseudo-regression coefficients. This process was repeated 100,000 times for the null hypothesis distributions. The original coefficients were ranked within these distributions to calculate p-values. As nine tests were conducted across combinations of electrodes (Fz, Cz, and Pz) and frequency bands (delta, theta, and alpha), FDR correction was applied to adjust the p-values. Using a significance threshold of α = 0.05, the coefficients were evaluated to determine whether they significantly differed from zero.

Although the number of participants in the present study was relatively modest (N = 32), the primary analyses were conducted at the pair level, yielding a total of 496 data points (in-group = 244, out-group = 252). From this perspective, the number of observations can be considered relatively large, enhancing the reliability and robustness of the statistical analyses.

Usage of AI-assisted technologies in the writing process

During the preparation of this work, the authors used ChatGPT (GPT-4, OpenAI Inc., United States) to improve the language of the manuscript . After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Results

Subjective and ECG results

The results of the subjective ratings of the questionnaires are summarised in Fig. 3. Higher scores indicated more positive emotions, whereas lower scores reflected more negative emotions. Arousal levels are represented by larger values, indicating higher arousal. Similarly, higher enjoyment scores denoted greater enjoyment of the game content. Permutation tests revealed that Hanshin Tigers fans reported significantly more positive valence after watching the Hanshin Tigers’ victory video compared to that with the Orix Buffaloes’ victory video (p = .002) or draw video (p = .017). Similarly, Orix Buffaloes fans reported significantly more positive valence after watching the Orix Buffaloes’ victory video compared to that with the Hanshin Tigers’ victory video (p = .017) or draw video (p = .044). In terms of arousal level, Hanshin Tigers fans showed significantly higher levels after watching the Hanshin Tigers’ victory video than after the Orix Buffaloes’ victory video (p = .032) and draw video (p = .044). However, no significant differences in arousal levels were observed among Orix Buffaloes fans across the different video types. The enjoyment scores had a similar trend to the arousal level; Hanshin Tigers fans reported significantly higher enjoyment for the Hanshin Tigers’ victory video compared to the Orix Buffaloes’ victory video (p = .026) and draw video (p = .004). In contrast, no significant differences in enjoyment scores were observed among Orix Buffaloes fans across the video types.

Regarding ECG results, Fig. 4 presents the data for HR and the LF/HF ratio. The ANOVAs with the factors of fan group and video types did not show any significant main effect or interaction (ps > 0.05).

[See PDF for image]

Fig. 3

Subjective ratings after each video. Each box plot illustrates the distribution of subjective ratings, divided by fan groups (x-axis) and the winner of the video (indicated by colours). High scores in the emotional valence, arousal level, and enjoyment score mean “positive”, “excited”, and “enjoyed,” respectively. The score of 4 represents a “neutral” response on each scale. “HT” refers to the Hanshin Tigers, while “OB” represents the Orix Buffaloes, and “DR” indicates a draw. The box plots display the median, the lower (Q1) and upper (Q3) quartiles, and outliers. The box represents the interquartile range (IQR), which contains the middle 50% of the data. The edges of the box indicate Q1 and Q3. The bold line inside the box indicates the median. The whiskers extend from the box to the smallest and largest values within 1.5 times the IQR from the quartiles. Data points outside this range are plotted individually as outliers, indicated by black dots. The enjoyment scores of HT fans after watching the HT victory video and OB fans after watching the OB victory video show little variance, with the Q1, Q3, and median lines overlapping.

[See PDF for image]

Fig. 4

ECG results during each video. The bar plots illustrate the HR (left) and LF/HF ratio (right). They are divided by fan groups (x-axis) and the winner of the video (indicated by colours). “HT” refers to the Hanshin Tigers, while “OB” represents the Orix Buffaloes, and “DR” indicates a draw. The error bars show standard errors.

EEG results

Phase synchronisation

Figure 5 summarises the results of phase synchronisation. The permutation-based multiple regression analysis of PLV found that the coefficient for pair category was significantly greater than zero in the alpha band at Cz ( = 0.00089779, p = .011) and Pz ( = 0.00132192, p = .011), wherein PLVs were higher for in-group pairs compared to out-group pairs (Fig. 5b). Furthermore, in the alpha band at Fz, the coefficient for fan history was significantly greater than zero ( = 0.00010938, p = .039, Fig. 5a), indicating that pairs with longer shared experience showed higher PLV. Conversely, in the delta and theta bands at Pz, the coefficient for fan history was significantly less than zero (delta: = -0.0003370758, p = .037, theta: = -0.00014157, p = .037), suggesting that pairs with longer experience exhibited lower PLV. No other significant effects or interactions were observed for any bands or electrodes.

[See PDF for image]

Fig. 5

Phase synchrony results. (a) Scatter plots are organised by three electrodes (Fz, Cz, and Pz) and three frequency bands (delta, theta, and alpha). The significant effects are indicated by asterisk (* p < .05) (b) The PLV distribution in the alpha band at Cz and Pz, designed to highlight the differences between the pair categories. The red and blue lines show medians of in-group and out-group.

Power synchronisation

Figure 6 summarises the results of power synchronisation. The permutation-based multiple regression analysis on power synchronisation index r identified that the interaction term between pair category and fan history was significantly greater than zero in the alpha band at Pz ( = 0.001356, p = .046). To examine this interaction, the coefficient of fan history was evaluated separately for each level of the pair category (i.e. in-group and out-group), with p-values calculated by ranking its position within the null distribution of pseudo-coefficients generated through 100,000 permutations. FDR correction was applied to account for these two comparisons. In the in-group condition, the fan history coefficient was significantly greater than zero ( = 0.002064339, p < .001), whereas in the out-group condition, the coefficient did not significantly differ from zero ( = -0.0006479794, p = .437). No other significant effects or interactions were observed for any bands or electrodes.

[See PDF for image]

Fig. 6

Power synchrony results. Scatter plots organised by three electrodes (Fz, Cz, and Pz) and three frequency bands (delta, theta, and alpha). The significant effect is indicated by asterisk (* p < .05), pair cat. and fan hist. are abbreviation for pair category and fan history.

Discussion

This study explored social cognitive bias by analysing the degree of shared neural processing, focusing on the influence of group identity and membership length. To achieve this, we analysed EEG synchronisation among fans of two baseball teams while watching matches. The subjective results showed that Hanshin Tigers fans felt positive emotions when their team won, whereas Orix Buffaloes fans became positive when they emerged victorious (Fig. 3, left). These emotions likely reflect the cognitive biases associated with each game event. No significant effects of fan group or video type were observed on either HR or the LF/HF ratio. In each video, the winning team was revealed only at the end, with both teams alternating between winning and losing throughout the game. Arousal levels fluctuated substantially in response to the unfolding events, and since HR and LF/HF ratio were calculated based on data from the entire video duration, it is possible that the effects observed in the subjective measures were not captured in these physiological indices. Regarding EEG, we hypothesised that (1) in-group pairs would demonstrate greater synchronisation than out-group pairs. Additionally, we proposed that (2) longer membership durations within the in-group would further amplify synchronisation, whereas no such effect would be expected for the out-group. The results revealed a combination of expected and unexpected patterns. The following sections interpret each observed effect.

This study employed two complementary indices of EEG synchronisation: the PLV and power-based correlation (r), which primarily reflect evoked and induced brain responses, respectively. Evoked responses are neural signals that are phase-locked to the onset of a stimulus. This means that they occur at a consistent time point and with a consistent oscillatory phase relative to the stimulus. In contrast, induced responses are triggered by a stimulus and occur at a similar latency but may vary in phase alignment³⁷. In the context of ISC, PLV quantifies the consistency of phase alignment across participants in response to each stimulus in the videos (i.e. changes in visual or auditory input). This measure is closely related to inter-trial phase clustering (ITPC) and ERPs in within-subject analyses, and it primarily captures stimulus-locked phase resetting associated with perceptual and cognitive processing. In contrast, the power-based synchronisation index r measures the correlation of power fluctuations across participants. This is conceptually related to event-related synchronisation (ERS) and desynchronisation (ERD) in within-subject analyses, which reflect changes in the magnitude of neural activity that are induced by events but not necessarily phase-locked to them. Fluctuations in internal states, such as arousal, are more likely to be reflected in this index. Together, PLV and r offer a more comprehensive account of inter-individual neural synchrony by capturing both phase-locked responses to external stimuli and shared endogenous dynamics.

Consistent with the first hypothesis, we found that the phase of the centroparietal alpha activity was more synchronised within in-group pairs than within out-group pairs (Fig. 5). This effect was independent of membership duration. This result indicates that the alpha responses elicited by the events in the videos were more frequently aligned across in-group pairs. Previous studies have proposed that the posterior alpha response to stimuli is closely associated with P1 and N1, which are ERPs that reflect early visual processing^{38, 39, 40, 41–42}. Furthermore, top-down spatial attention modulated the P1 and N1 amplitudes^{43, 44, 45, 46–47}. Based on these results, the alpha-phase synchronisation observed in this study may reflect neural fluctuations in early visual processing caused by top-down spatial attention. This result can also be interpreted through the lens of the self-categorisation theory⁴⁸which posits that individuals categorise themselves into social groups based on contextual cues, influencing their attention and perception. In the present study, identifying as a Hanshin Tigers or Orix Buffaloes fan may have activated a categorical, identity-based distinction between in-group and out-group members. Such categorisation can occur rapidly based on group labels and does not necessarily require long-term experience. Once established, group identity may guide early stages of visual processing via top-down attentional mechanisms. These identity-driven biases in attentional allocation could lead members of different groups to focus on and encode different aspects of the same events. Over time, such differences in attentional focus may contribute to divergent interpretations of shared experiences, potentially serving as a foundational process in the development and reinforcement of social biases.

The effect, clearly consistent with the second hypothesis, was that power at the parietal alpha showed stronger synchronisation with a longer fan history (i.e. duration of group membership) exclusively among in-group pairs, whereas this effect was absent in out-group pairs (Fig. 6). The ERS/ERD of alpha activity at posterior sites has been argued to reflect the inhibition or de-inhibition of cortical arousal and can be influenced by endogenous attention^42,49,50 and emotional processing⁵¹. These findings suggest that, in this study, in-group pairs with longer shared affiliation experiences exhibited more synchronised changes of arousal across video scenes. One possible explanation is that they may share similar interests and emotional responses to individual scenes. By contrast, out-group pairs may have experienced differing interests or emotional reactions, leading to less-aligned arousal patterns. These findings underscore the important role of accumulated experience in the synchronisation of internal states. Such experience may foster a shared interpretation of events, thereby promoting aligned emotional responses and engagement. According to the social identity theory²emotional investment in the in-group further reinforces this alignment, as individuals are motivated to perceive and respond to events in ways that affirm group identity and bolster self-esteem. Our findings suggest that accumulated experience is an important factor in shaping such social attitudes.

Interestingly, while categorical in-group recognition may modulate attention, as indicated by the alpha-phase findings, the synchrony of arousal dynamics appears to additionally rely on group-related experiential history. This suggests that identity-based categorisation and experience-based identity formation may influence distinct layers of neural information processing: the former shaping attention via rapid, top-down mechanisms, and the latter modulating affective and arousal responses over time through shared experience.

Unexpectedly, the affiliation period was associated with decreased phase synchronisation of delta and theta activity in the parietal region (Fig. 5). These frequency bands are linked to late-stage attentional processes, possibly related to the P300 component^52,53. This study suggests that a longer fan history, regardless of group identity, leads to more varied triggers of late-stage attention. Experienced baseball fans may interpret game details differently, focusing not only on universally significant events (e.g. home runs), but also on subtler aspects (e.g. player movements or expressions). These individualised perceptions are likely to lead to varied attentional patterns, thereby reducing phase synchronisation.

Another effect of membership duration was observed in frontal alpha activity, where phase synchronisation increased with the length of affiliation regardless of group identity (Fig. 5). This activity is likely linked to the auditory N1 component, which involves a rhythm around alpha and is localised in a more anterior region than the visual P1 and N1. Endogenous attention to auditory stimuli influences the auditory N1 amplitude through modulation or superimposition^{54, 55–56}. The frontal alpha-phase synchronisation results may reflect differences in attention allocation between visual and auditory stimuli. In this study, visual stimuli and auditory commentary were presented simultaneously, with visual stimuli providing intuitive and direct information about the game. Participants with shorter fan histories likely prioritised visual stimuli, whereas those with longer histories may have attended to both, resulting in stronger synchronisation with the commentary. This possibility should be investigated in future studies using different types of stimuli.

In summary, this study identified EEG synchronisation patterns consistent with the hypotheses, suggesting that cognitive bias according to group identity and membership duration emerges as a variation in the degree of shared brain processing. Additionally, this study provides insights into how bias affects cognitive processes. First, the centroparietal alpha-phase findings suggest that in-group and out-group distinctions may influence early visual processing through top-down attention mechanisms. Second, the interaction between group affiliation and membership duration observed in parietal alpha power indicates that a sense of belonging can shape the synchronisation of arousal level, reflecting engagement and emotional resonance. These results suggest that while categorical group recognition may rapidly influence attention, the synchrony of arousal dynamics additionally reflects experience-based identity formation, indicating that different layers of neural processing are shaped by distinct aspects of social identity. The results for parietal delta and theta phase synchronisation highlight that fan experience, regardless of group affiliation, may lead to different interpretations of individual game events. Additionally, the observed frontal alpha-phase synchronisation suggests that longer experience allows greater attention to the auditory commentary. Although these interpretations are tentative, they provide valuable insights for future studies. Targeted experiments using various stimuli and refined hypotheses are required to validate these findings. Finally, this study highlights the utility of EEG for investigating cognitive biases in social information processing. EEG synchronisation can serve as an objective measure of cognitive bias, offering valuable scientific insights to help mitigate social division.

Several limitations of this study should be noted. First, a limited number of EEG electrodes was used, which may have restricted the spatial resolution of the observed neural activity. Future research employing high-density EEG systems will help capture a more comprehensive picture of brain dynamics. Second, although we used fan history as a proxy for accumulated group experience, this measure may not fully reflect their fandom. Future studies should consider alternative indices to better capture fan engagement (e.g. subjective report on how much they like the team). Third, while this study computed averages of neural synchrony data across entire video segments, specific events during the game (e.g. scoring, critical plays) likely evoked transient synchronisation patterns. Investigating time-resolved dynamics around such key events may provide deeper insights into how social cognitive bias evolves in real-time processing and should be addressed in future studies.

Acknowledgements

This study was partially supported by KAKENHI 23K21739.

Author contributions

Both authors contributed to the study design and interpretation of the results. MS collected and analysed the data and drafted the initial version of the manuscript. YN provided critical revisions.

Data availability

The data of this study are available upon reasonable request to the corresponding author.

Declarations

Competing interests

The authors declare no competing interests.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. De Dreu, CKW; Gross, J; Fariña, A; Ma, Y. Group cooperation, Carrying-Capacity stress, and intergroup conflict. Trends Cogn. Sci.; 2020; 24, pp. 760-776. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32620334][DOI: https://dx.doi.org/10.1016/j.tics.2020.06.005]

2. Tajfel, H; Turner, JC. Williams, JA; Worchel, S. An integrative theory of intergroup conflict. The Social Psychology of Intergroup Relations; 1979; Belmont, CA, Wadsworth: pp. 33-47.

3. Itagaki, S; Katayama, J. Self-relevant criteria determine the evaluation of outcomes induced by others. NeuroReport; 2008; 19, 383. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18303586][DOI: https://dx.doi.org/10.1097/WNR.0b013e3282f556e8]

4. Gehring, W. & Willoughby, A. J. R. The Medial Frontal Cortex and the Rapid Processing of Monetary Gains and Losses | Science. https://www.science.org/doi/https://doi.org/10.1126/science.1066893 (2002).

5. Cikara, M; Botvinick, MM; Fiske, ST. Us versus them: social identity shapes neural responses to intergroup competition and harm. Psychol. Sci.; 2011; 22, pp. 306-313. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21270447][DOI: https://dx.doi.org/10.1177/0956797610397667]

6. Segerdahl, A. R., Mezue, M., Okell, T. W., Farrar, J. T. & Tracey, I. The dorsal posterior Insula subserves a fundamental role in human pain. Nat. Neurosci.https://www.nature.com/articles/nn.3969 (2015).

7. Shackman, AJ et al. The integration of negative affect, pain and cognitive control in the cingulate cortex. Nat. Rev. Neurosci.; 2011; 12, pp. 154-167.1:CAS:528:DC%2BC3MXitVOnsL8%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21331082][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3044650][DOI: https://dx.doi.org/10.1038/nrn2994]

8. Delgado, MR. Reward-Related responses in the human striatum. Ann. N Y Acad. Sci.; 2007; 1104, pp. 70-88.2007NYASA1104..70D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17344522][DOI: https://dx.doi.org/10.1196/annals.1390.002]

9. Nastase, A. S., Gazzola, V., Hasson, U. & Keysers, C. Measuring shared responses across subjects using intersubject correlation | social cognitive and affective neuroscience | Oxford academic. https://academic.oup.com/scan/article/14/6/667/5489905 (2019).

10. Ratner, KG; Amodio, DM. Seeing Us vs. them: minimal group effects on the neural encoding of faces. J. Exp. Soc. Psychol.; 2013; 49, pp. 298-301. [DOI: https://dx.doi.org/10.1016/j.jesp.2012.10.017]

11. Hart, A et al. Differential response in the human amygdala to Racial outgroup vs ingroup face stimuli. NeuroReport; 2000; 11, 2351.1:STN:280:DC%2BD3M%2FmvFSjtw%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/10943684][DOI: https://dx.doi.org/10.1097/00001756-200008030-00004]

12. Hasson, U; Nir, Y; Levy, I; Fuhrmann, G; Malach, R. Intersubject synchronization of cortical activity during natural vision. Science; 2004; 303, pp. 1634-1640.2004Sci..303.1634H1:CAS:528:DC%2BD2cXhvFCnsbw%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15016991][DOI: https://dx.doi.org/10.1126/science.1089506]

13. Lerner, Y; Honey, CJ; Silbert, LJ; Hasson, U. Topographic mapping of a hierarchy of Temporal receptive windows using a narrated story. J. Neurosci.; 2011; 31, 2906.1:CAS:528:DC%2BC3MXislGitrs%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21414912][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3089381][DOI: https://dx.doi.org/10.1523/JNEUROSCI.3684-10.2011]

14. Hasson, U; Malach, R; Heeger, DJ. Reliability of cortical activity during natural stimulation. Trends Cogn. Sci.; 2010; 14, pp. 40-48. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20004608][DOI: https://dx.doi.org/10.1016/j.tics.2009.10.011]

15. Hasson, U; Furman, O; Clark, D; Dudai, Y; Davachi, L. Enhanced intersubject correlations during movie viewing correlate with successful episodic encoding. Neuron; 2008; 57, pp. 452-462.1:CAS:528:DC%2BD1cXit1SnsbY%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18255037][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2789242][DOI: https://dx.doi.org/10.1016/j.neuron.2007.12.009]

16. Jääskeläinen, I. P. et al. Inter-Subject Synchronization of Prefrontal Cortex Hemodynamic Activity During Natural Viewing. 2 (2008).

17. Lankinen, K; Saari, J; Hari, R; Koskinen, M. Intersubject consistency of cortical MEG signals during movie viewing. NeuroImage; 2014; 92, pp. 217-224.1:STN:280:DC%2BC2cvltlOhtQ%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24531052][DOI: https://dx.doi.org/10.1016/j.neuroimage.2014.02.004]

18. Lankinen, K et al. Consistency and similarity of MEG- and fMRI-signal time courses during movie viewing. NeuroImage; 2018; 173, pp. 361-369. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29486325][DOI: https://dx.doi.org/10.1016/j.neuroimage.2018.02.045]

19. Chang, WT et al. Combined MEG and EEG show reliable patterns of electromagnetic brain activity during natural viewing. NeuroImage; 2015; 114, pp. 49-56. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25842290][DOI: https://dx.doi.org/10.1016/j.neuroimage.2015.03.066]

20. Dmochowski, JP et al. Audience preferences are predicted by Temporal reliability of neural processing. Nat. Commun.; 2014; 5, 4567.2014NatCo..5.4567D1:CAS:528:DC%2BC2MXksVekurk%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25072833][DOI: https://dx.doi.org/10.1038/ncomms5567]

21. Dmochowski, J. P., Sajda, P., Dias, J. & Parra, L. C. Correlated components of ongoing EEG point to emotionally laden Attention – A possible marker of engagement?? Front. Hum. Neurosci.6 (2012).

22. Poulsen, AT; Kamronn, S; Dmochowski, J; Parra, LC; Hansen, L. K. EEG in the classroom: synchronised neural recordings during video presentation. Sci. Rep.; 2017; 7, 43916.2017NatSR..743916P [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28266588][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5339684][DOI: https://dx.doi.org/10.1038/srep43916]

23. Petroni, A. et al. The Variability of Neural Responses to Naturalistic Videos Change with Age and Sex. eNeuro 5 (2018).

24. Madsen, J. & Parra, C. L. Cognitive processing of a common stimulus synchronizes brains, hearts, and eyes. PNAS Nexus 1 (2022).

25. Finn, ES et al. Idiosynchrony: from shared responses to individual differences during naturalistic neuroimaging. NeuroImage; 2020; 215, 116828. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32276065][DOI: https://dx.doi.org/10.1016/j.neuroimage.2020.116828]

26. Thiede, A; Glerean, E; Kujala, T; Parkkonen, L. Atypical MEG inter-subject correlation during listening to continuous natural speech in dyslexia. NeuroImage; 2020; 216, 116799.1:STN:280:DC%2BB38zls1Cguw%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32294536][DOI: https://dx.doi.org/10.1016/j.neuroimage.2020.116799]

27. Parkinson, C; Kleinbaum, AM; Wheatley, T. Similar neural responses predict friendship. Nat. Commun.; 2018; 9, 332.2018NatCo..9.332P [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29382820][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5790806][DOI: https://dx.doi.org/10.1038/s41467-017-02722-7]

28. Mathieu, JE; Zajac, DM. A review and meta-analysis of the antecedents, correlates, and consequences of organizational commitment. Psychol. Bull.; 1990; 108, pp. 171-194. [DOI: https://dx.doi.org/10.1037/0033-2909.108.2.171]

29. Mael, F; Ashforth, BE. Alumni and their alma mater: A partial test of the reformulated model of organizational identification. J. Organ. Behav.; 1992; 13, pp. 103-123. [DOI: https://dx.doi.org/10.1002/job.4030130202]

30. Yokota, Y. & Naruse, Y. Temporal fluctuation of mood in gaming task modulates feedback negativity: EEG study with virtual reality. Front. Hum. Neurosci.15 (2021).

31. Perakakis, PP. HEPLAB: a matlab graphical interface for the preprocessing of the heartbeat-evoked potential. Zenodo; 2019; [DOI: https://dx.doi.org/10.5281/zenodo.2649943]

32. Gordan, R; Gwathmey, JK; Xie, LH. Autonomic and endocrine control of cardiovascular function. World J. Cardiol.; 2015; 7, 204. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25914789][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4404375][DOI: https://dx.doi.org/10.4330/wjc.v7.i4.204]

33. Shaffer, F. & Ginsberg, J. P. An overview of heart rate variability metrics and norms. Front. Public Health5 (2017).

34. Lachaux, JP; Rodriguez, E; Martinerie, J; Varela, FJ. Measuring phase synchrony in brain signals. Hum. Brain Mapp.; 1999; 8, pp. 194-208.1:STN:280:DC%2BD3c%2FosVCgsA%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/10619414][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6873296][DOI: https://dx.doi.org/10.1002/(SICI)1097-0193(1999)8:4<194::AID-HBM4>3.0.CO;2-C]

35. Kriegeskorte, N., Mur, M. & Bandettini, P. A. Representational similarity analysis - connecting the branches of systems neuroscience. Front. Syst. Neurosci.2 (2008).

36. Lichstein, JW. Multiple regression on distance matrices: a multivariate Spatial analysis tool. Plant. Ecol.; 2007; 188, pp. 117-131. [DOI: https://dx.doi.org/10.1007/s11258-006-9126-3]

37. Tallon-Baudry, C; Bertrand, O. Oscillatory gamma activity in humans and its role in object representation. Trends Cogn. Sci.; 1999; 3, pp. 151-162.1:STN:280:DC%2BC2sbgsVGrtg%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/10322469][DOI: https://dx.doi.org/10.1016/S1364-6613(99)01299-1]

38. Gruber, WR; Klimesch, W; Sauseng, P; Doppelmayr, M. Alpha phase synchronization predicts P1 and N1 latency and amplitude size. Cereb. Cortex; 2005; 15, pp. 371-377. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15749980][DOI: https://dx.doi.org/10.1093/cercor/bhh139]

39. Freunberger, R et al. Functional similarities between the P1 component and alpha oscillations. Eur. J. Neurosci.; 2008; 27, pp. 2330-2340. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18445223][DOI: https://dx.doi.org/10.1111/j.1460-9568.2008.06190.x]

40. Klimesch, W; Hanslmayr, S; Sauseng, P; Gruber, WR; Doppelmayr, M. P1 and traveling alpha waves: evidence for evoked oscillations. J. Neurophysiol.; 2007; 97, pp. 1311-1318. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17167063][DOI: https://dx.doi.org/10.1152/jn.00876.2006]

41. Klimesch, W. Evoked alpha and early access to the knowledge system: the P1 Inhibition timing hypothesis. Brain Res.; 2011; 1408, pp. 52-71.1:CAS:528:DC%2BC3MXhtVWrsL3M [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21774917][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3158852][DOI: https://dx.doi.org/10.1016/j.brainres.2011.06.003]

42. Klimesch, W. Alpha-band oscillations, attention, and controlled access to stored information. Trends Cogn. Sci.; 2012; 16, pp. 606-617. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/23141428][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3507158][DOI: https://dx.doi.org/10.1016/j.tics.2012.10.007]

43. Anllo-Vento, L; Hillyard, SA. Selective attention to the color and direction of moving stimuli: electrophysiological correlates of hierarchical feature selection. Percept. Psychophys; 1996; 58, pp. 191-206.1:STN:280:DyaK28vjsFakug%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/8838164][DOI: https://dx.doi.org/10.3758/BF03211875]

44. Luck, SJ et al. Effects of Spatial Cuing on luminance detectability: psychophysical and electrophysiological evidence for early selection. J. Exp. Psychol. Hum. Percept. Perform.; 1994; 20, pp. 887-904.1:STN:280:DyaK2czmsFGnuw%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/8083642][DOI: https://dx.doi.org/10.1037/0096-1523.20.4.887]

45. Hillyard, SA; Luck, SJ; Mangun, GR. Heinze, HJ; Münte, TF; Mangun, GR. The Cuing of attention to visual field locations: analysis with ERP recordings. Cognitive Electrophysiology; 1994; Boston, MA, Birkhäuser: pp. 1-25. [DOI: https://dx.doi.org/10.1007/978-1-4612-0283-7_1]

46. Mangun, GR; Hillyard, SA. Spatial gradients of visual attention: behavioral and electrophysiological evidence. Electroencephalogr. Clin. Neurophysiol.; 1988; 70, pp. 417-428.1:STN:280:DyaL1M%2FjtVCrsA%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/2460315][DOI: https://dx.doi.org/10.1016/0013-4694(88)90019-3]

47. Mangun, GR; Hillyard, SA. Allocation of visual attention to Spatial locations: tradeoff functions for event-related brain potentials and detection performance. Percept. Psychophys; 1990; 47, pp. 532-550.1:STN:280:DyaK3czhtVKntA%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/2367174][DOI: https://dx.doi.org/10.3758/BF03203106]

48. Turner, JC. Ellemers, N; Spears, R; Dossje, B. Some current issues in research on social identity and self-categorization theories. Social Identity: Context, Commitment, Content; 1999; Oxford, UK, Blackwell: pp. 6-34.

49. Worden, MS; Foxe, JJ; Wang, N; Simpson, GV. Anticipatory biasing of visuospatial attention indexed by retinotopically specific alpha-band electroencephalography increases over occipital cortex. J. Neurosci. Off J. Soc. Neurosci.; 2000; 20, RC63.1:STN:280:DC%2BD3c3ns1ersw%3D%3D [DOI: https://dx.doi.org/10.1523/JNEUROSCI.20-06-j0002.2000]

50. Medendorp, WP et al. Oscillatory activity in human parietal and occipital cortex shows hemispheric lateralization and memory effects in a delayed double-step saccade task. Cereb. Cortex N Y N 1991; 2007; 17, pp. 2364-2374.

51. Schubring, D; Schupp, HT. Affective picture processing: Alpha- and lower beta-band desynchronization reflects emotional arousal. Psychophysiology; 2019; 56, e13386. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31026079][DOI: https://dx.doi.org/10.1111/psyp.13386]

52. Güntekin, B; Başar, E. Review of evoked and event-related delta responses in the human brain. Int. J. Psychophysiol.; 2016; 103, pp. 43-52. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25660301][DOI: https://dx.doi.org/10.1016/j.ijpsycho.2015.02.001]

53. Başar-Eroglu, C; Başar, E; Demiralp, T; Schürmann, M. P300-response: possible Psychophysiological correlates in delta and theta frequency channels. A review. Int. J. Psychophysiol.; 1992; 13, pp. 161-179. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/1399755][DOI: https://dx.doi.org/10.1016/0167-8760(92)90055-G]

54. Hillyard, SA; Hink, RF; Schwent, VL; Picton, TW. Electrical signs of selective attention in the human brain. Science; 1973; 182, pp. 177-180.1973Sci..182.177H1:STN:280:DyaE3s3ltVOisQ%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/4730062][DOI: https://dx.doi.org/10.1126/science.182.4108.177]

55. Näätanen, R; Teder, W; Alho, K; Lavikainen, J. Auditory attention and selective input modulation: A topographical ERP study. NeuroReport; 1992; 3, 493. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/1391755][DOI: https://dx.doi.org/10.1097/00001756-199206000-00009]

56. Teder, W; Alho, K; Reinikainen, K; Näätänen, R. Interstimulus interval and the selective-attention effect on auditory erps. Psychophysiology; 1993; 30, pp. 71-81.1:STN:280:DyaK2c%2Fit1OrtQ%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/8416064][DOI: https://dx.doi.org/10.1111/j.1469-8986.1993.tb03206.x]