Skip to main content
NCBI home page
Social Cognitive and Affective Neuroscience logoLink to Social Cognitive and Affective Neuroscience
. 2025 Apr 26;20(1):nsaf036. doi: 10.1093/scan/nsaf036

Neural correlates of power-related postures and their behavioural consequences: a preliminary electrophysiological investigation

Soren Wainio-Theberge 1,2,3, Jorge L Armony 4,5,6,*
PMCID: PMC12083452  PMID: 40285428

Abstract

Social dominance is conveyed by expansive and contractive body postures, which also have feedback effects on individuals’ own mood and behaviour. These feedback effects are the subject of the ‘power posing’ paradigm, which has grown in popularity in psychology; however, the neural mechanisms of feedback from expansive and contractive postures have never been investigated. We report here for the first time an exploratory neuroimaging study using electroencephalography during a ‘power posing’ design to investigate the neural correlates of this effect. We find that right-lateralized frontal asymmetry in neural activity was increased as a result of taking an expansive posture and that this asymmetry was correlated with the effects the posture exerted on participants’ mood. We interpret this finding in the context of recent theories of frontal alpha asymmetry and motivational conflict.

Keywords: posture, frontal alpha asymmetry, electroencephalography, power posing

Introduction

Social dominance behaviours are a set of verbal and nonverbal attitudes and actions, which guide competition for resources among individuals (Johnson et al. 2012); they allow organisms to avoid direct conflict, either by intimidating other individuals (dominance) or by signalling that they wish to avoid a fight (submission) (Koski et al. 2015). Body posture is a crucial medium of this dominance and submission signalling (Burgoon and Dunbar 2006). Both animals and humans display dominance using open and expansive postures and submission through closed and contractive ones (Burgoon and Dunbar 2006). In humans, these postural signals often have higher-order social consequences: individuals with more open postures are perceived as being of higher social status and being more competent and successful (Weisfeld and Beresford 1982, Hall et al. 2005).

Recent research has suggested that beyond their role in conveying social information to others, the expansive and contractive postures used in dominance signalling have feedback effects on the individuals who adopt them. For example, adopting an expansive posture appears to transiently produce several of the behavioural signatures of social power, including increased risk-taking (Cesario and McDonald 2013) and higher likelihood to commit moral violations (Yap et al. 2013), as well as enhanced positive affect and self-esteem (Körner et al. 2021). Despite earlier controversies (Simmons and Simonsohn 2017), recent meta-analyses suggest that some of the behavioural effects of the ‘power posing’ paradigm (Carney et al. 2010) are robust (Elkjaer et al. 2022, Körner et al. 2022). In particular, influences of posture on individuals’ mood and feelings of power are viewed as particularly reliable, even by those critical of the paradigm in general (Gronau et al. 2017, Jonas et al. 2017). Yet, the neural processes associated with these postural feedback effects remain largely unknown.

Social power has been examined independent of posture using fMRI, implicating regions such as the medial prefrontal cortex, dorsolateral prefrontal cortex, and inferior parietal sulcus (Chiao 2010, Li et al. 2021). However, fMRI is impractical for studying posture manipulations, as participants must lie flat in the scanner, and even small motions create large artefacts which are problematic for analysis. In contrast, electroencephalography (EEG) is well suited for studying the brain correlates of postural feedback effects, despite its lower spatial resolution (Muthukumaraswamy 2013), as it allows participants more freedom of motion. Moreover, EEG allows the assessment of several relevant metrics, including frequency band power (which is linked to numerous cognitive processes; Buzsáki 2006), as well as putative markers of emotional state (Lee et al. 2014, Yang et al. 2020). Indeed, EEG has been used successfully in the past to investigate feedback effects related to approach motivation and postural lean (Harmon-Jones et al. 2011). Nonetheless, to the best of our knowledge, EEG has yet to be applied to study power-related postures.

Thus, the aim of the present project was to use EEG to directly investigate the neural mechanisms of postural feedback using a ‘power posing’ paradigm. As power pose effects on mood have typically been the most robust, we chose to focus on emotional state using questionnaires. Given the lack of previous research on the neural correlates of adopting power-related postures, we undertook an exploratory approach, testing for differences in frequency band power across the EEG power spectrum as these are commonly analysed features of resting-state EEG with known cognitive and behavioural correlates (Buzsáki 2006). Additionally, as the mood questionnaire outcome measures had dimensions for arousal and valence, we also analysed two a priori metrics which have previously been associated with these dimensions: namely, the EEG spectral exponent (Lendner et al. 2020) and frontal alpha asymmetry (Harmon-Jones and Gable 2018), respectively. The former has been associated with arousal in sleep, sedation, and disorders of consciousness (Lendner et al. 2020), while the latter has been associated with clinical depression (Gotlib 1998), positive affect (Coan and Allen 2003), but also more specifically behavioural activation and approach motivation (Gable et al. 2018). Relative to contractive postures, expansive postures have also been shown to increase both emotional arousal and positive-valenced mood (Nair et al. 2015, Körner et al. 2021); we thus expected that they would likewise produce reductions in the EEG spectral exponent and increased left-lateralized frontal cortical asymmetry.

Materials and methods

Participants

A total of 147 participants (mean age = 21.8, SD = 3.6, 109 self-reported as female) were recruited from the McGill Psychology Department Participant Pool and the general public via social media advertisements. The study was approved by the McGill University Faculty of Medicine Institutional Review Board (IRB no. A01-B03-15A) and written informed consent was obtained from all participants. Participants were either compensated with CAD 50$ for their participation or received course credit. The study procedures were carried out in accordance with the Declaration of Helsinki.

Study procedures

The present study reports data from the same experiment as Wainio-Theberge et al. (2024), although the final sample was not the same due to differences in data exclusion criteria; that study reported the results of a mediation analysis focusing on photogrammetric measures of body posture and electromyography, while the present study focuses on EEG. Briefly, the experiment involved a series of resting-state EEG recordings with accompanying mood questionnaires, interspersed with blocks of a perceptual task. Participants were pseudo-randomly assigned to an expansive or contractive posture group, with postures defined following the original ‘power posing’ study (Carney et al. 2010). In the section of the experiment reported here, participants completed a mood questionnaire, followed by a 3-min EEG resting state. They then held an expansive or contractive posture for 3 min and subsequently completed another mood questionnaire and resting state. The mood questionnaire included the Affect Valuation Index (AVI; Tsai et al. 2006), the Authentic Pride (AP) Scale (Tracy and Robins 2007), and the Toronto Hospital Alertness Test (Shapiro et al. 2006). The latter scale was used as an attention and compliance control, as well as to support the cover story: before starting the experiment, participants were told that the study was about the effects of posture and heart-rate variability and cognition and that the questionnaires were presented as control measures for the perceptual task. Following the post-posture resting state, participants answered several questions about their experience of the preceding posture, including how uncomfortable they found the posture, how much difficulty they had maintaining the posture, and how natural the posture felt. EEG was recorded with a 96-channel BioSemi ActiveTwo system with active Ag/AgCl electrodes at a sampling rate of 2048 Hz. The montage was based on the standard BioSemi 64-channel headcap, with 32 additional electrodes added following the BrainVision setup. Channel DC offsets were maintained <50 mV during recording.

Behavioural analysis

The AVI is an instrument based on the affective circumplex (Posner et al. 2005), a model which assumes that emotions map onto a two-dimensional space whose axes represent arousal (physiological activation in response to a situation) and valence (the positive or negative quality ascribed to a stimulus). The AVI contains nine subscales reflecting different directions (e.g. high arousal positive emotion, high arousal neutral, and high arousal negative) in this circumplex space; here, we combined them into two factors reflecting the simple directions of arousal and valence, as in Wainio-Theberge et al. (2024). Specifically, arousal was calculated by subtracting the low arousal subscales (low arousal, low arousal negative, and low arousal positive) from the high arousal subscales (high arousal, high arousal positive, and high arousal negative), such that higher scores indicate higher emotional arousal and lower scores indicate lower emotional arousal, regardless of valence. Valence was calculated by subtracting the negative valence subscales (negative, high arousal negative, and low arousal negative) from the positive valence subscales (positive, high arousal positive, and low arousal positive), such that higher scores indicate more positive emotional valence and lower scores indicate more negative valence, independent of arousal levels. The AP Scale was analysed as its own dimension. Differently from Wainio-Theberge et al. (2024) and to ensure consistency with the neural analyses, we assessed the posture effects on arousal, valence, and pride by entering the post-posture values into a linear model with posture group as an independent variable and pre-posture arousal, valence, or pride as a covariate; this has been shown to yield higher statistical power than difference-based approaches, while having the same interpretation (Senn 2006, McKenzie 2012).

In separate control analyses, we included quantitative measures of participants’ posture (namely participants’ neck angle) and self-report measures of how difficult they found the posture, as we previously observed they mediated the effects of posture on mood changes (Wainio-Theberge et al. 2024). The difficulty score was created as a composite variable from three self-report questions, which were answered on a 10-point Likert scale (difficulty, discomfort, and naturalness, coded negatively); participants’ responses to these questions were z-scored and the difficulty score was created as the sum of these z-scored responses. The neck angle was computed from video recordings taken during the posture. Videos were downsampled to 1 frame per second, and the neck angle was calculated on each frame using OpenPOSE [Cao et al. 2019; see Wainio-Theberge and Armony (2024) for details on the computation]. Frames where the neck angle was less than −20° were automatically excluded, and the neck angle variable was computed for each participant as the median neck angle across frames during the posture. Higher values of neck angle correspond to a more forward or slumped head posture, while lower or more negative values correspond to a more upright head posture. Note that a number of participants were missing neck angle data (N = 22), due to either issues with the onboard memory storage of the camera (N = 10) or failures of OpenPOSE to correctly estimate posture from the recordings (N = 12).

EEG preprocessing and calculation of measures

Following recording, EEG data were downsampled to 256 Hz and preprocessed according to the HAPPE pipeline (Gabard-Durnam et al. 2018). Data were first filtered with a 1-Hz Finite Impulse Response (FIR) high-pass filter in EEGLAB (Delorme and Makeig 2004) and then notch-filtered at 60 and 120 Hz to remove line noise (this step replaced the ‘Cleanline’ procedure in the HAPPE, as it failed to suppress nonstationary line noise in our data). Bad channels were rejected using HAPPE’s normalized log-power heuristic, and wavelet-threshold Independent Component Analysis (ICA) was performed to clean data for ICA decomposition. ICA was performed using the Infomax algorithm in EEGLAB, and artefactual ICA components were rejected using MARA (Winkler et al. 2011). Data were then segmented into 2-s consecutive nonoverlapping epochs, and bad channels and segments were repaired using FASTER (Nolan et al. 2010). No segments were rejected outright at this step to avoid introducing discontinuities, which could affect frequency analysis. Finally, data were re-referenced to a common average reference.

To further clean the data, we conducted a k-means clustering analysis using three of the HAPPE pipeline’s preprocessing metrics (median residual artefact probability, number of channels initially rejected, and number of ICA components rejected). A cluster of recordings with high values of all of these metrics was identified; any subject containing at least one of these bad recordings (in either the pre-posture, during-posture, or post-posture resting states) was excluded from further analysis (N = 31 participants).

Frequency band power in EEG was calculated as follows. Power spectra were calculated for each channel using Welch’s (1967) method. Frequency bands were defined as 1–4 Hz delta, 30–80 Hz low gamma, and 80–120 Hz high gamma (Buzsáki 2006). Theta, alpha, and beta bands were determined using participants’ individualized alpha peak frequencies. Individual alpha frequencies and bandwidths were calculated following the method from Corcoran et al. (2018); the theta band was then defined as 4 Hz to the lower bound of the alpha peak, while beta power was defined as the upper bound of the alpha peak to 30 Hz. Power in each band was calculated by integrating the power spectral density within each frequency band using Simpson’s rule. Frequency-band power was then log-transformed to render them more normally distributed (Smulders et al. 2018). At this step, six additional subjects were removed for being outliers (>3 median absolute deviations) in any measure of power in at least one recording.

Our two a priori measures were calculated as follows. The EEG spectral exponent was calculated as in Kolvoort et al. (2020); Welch’s periodogram estimate was used to estimate the power spectrum for each electrode, and a linear fit was performed between 30 and 50 Hz following Lendner et al. (2020). Frequencies were interpolated to be evenly spaced on a logarithmic scale in order to equalize the impact of high and low frequencies in the linear fit (Wainio-Theberge et al. 2022). Spectral exponent values were then averaged over all sensors prior to analysis. To calculate frontal alpha asymmetry, we averaged the log-transformed individualized alpha power within five frontal electrodes (AF4, AF8, F4, F6, and F8) and took the difference with their left-hemisphere analogs (Smith et al. 2017); as such, higher asymmetry scores indicate greater right-lateralized alpha activity, which means greater left-lateralized frontal cortical activity as alpha is inversely associated with neural activity (Laufs et al. 2003). Alpha asymmetry scores were calculated using the individualized alpha power values described earlier.

Statistical analysis

Statistical analyses of EEG bandpower measures were carried out using the Fieldtrip package (Oostenveld et al. 2011) in MATLAB 2020a. We assessed neural effects of the expansive and contractive posture conditions by taking the normalized power in each band (i.e. the log of the ratio between power in the post-posture resting state and power in the pre-posture resting state) and comparing this between the expansive and contractive conditions using a cluster-based permutation test (Maris and Oostenveld 2007). Cluster tests are a nonparametric method of statistical testing which take advantage of correlations between neighbouring channels and brain regions in electrophysiological data. The procedure involves applying a univariate statistical test at each channel and then summing the test statistics over adjacent significant channels; data are then shuffled to generate a permutation distribution of this summed cluster statistic, and a P-value is generated from this distribution. In our study, the procedure was carried out using an unpaired t-test at each channel between the expansive and contractive conditions, and shuffling 2000 times.

The a priori selected measures (scaling exponent and frontal asymmetry) were analysed using linear models. The value of each measure during the post-posture resting state was used as the dependent variable, while posture group and pre-posture resting state values of each measure were used as categorical and parametric independent variables, respectively (contractive posture was treated as the reference category for posture group). These models (with only posture and pre-posture resting state as independent variables) are referred to throughout the text as ‘Model 1’.

Additional variables were added to the model if there was a significant or trend-level effect of posture in ‘Model 1’. ‘Model 2’ included mood change, defined as post-posture minus pre-posture values of ‘arousal’, ‘valence’, and ‘pride’, as an additional independent variable. If significant or trend-level effects of mood were observed, ‘Model 3’ and ‘Model 4’ were computed, with perceived difficulty and neck angle included as independent variables, respectively; each of these has been shown to be a relevant mediator of posture effects in Wainio-Theberge et al. (2024). Models included all two- and three-way interactions among factors.

Finally, to aid in interpretability, subgroup analyses treating each posture condition separately were conducted whenever a significant or trend-level interaction with posture was observed. These consisted of the same ‘Models 1–4’, but without posture as a factor of interest, and referred to as ‘Model Xc’ for contractive and ‘Model Xe’ for expansive (X = 1–4).

Results

Expansive and contractive postures have expected effects on mood

Following exclusion of participants as indicated in the Materials and methods, the final sample size was 104 (mean age = 21.7, SD = 3.3, 74 women). As previously reported (Wainio-Theberge et al. 2024), we observed significant effects of the posture on mood variables; these are reported again here due to slight differences in the analysis strategy (use of linear models vs. change scores—see Materials and methods for details) and the final sample included (original N = 100, Noverlap = 78). Mood questionnaires showed acceptable-to-high internal consistency [αarousal = 0.77, confidence interval (CI) = 0.69–0.83; αvalence = 0.80, CI = 0.74–0.85; αpride = 0.93, CI = 0.91–0.95]. Controlling for pre-posture mood, we observed significant posture effects on arousal (t(101) = 3.04, P = .003), valence (t(101) = 3.11, P = .002), and marginally on pride [t(101) = 2.03, P = .05; in the analysis in Wainio-Theberge et al. (2024), all these effects were significant]. Moreover, in the subset of participants in this sample who had neck angle data (N = 82), neck angle was again correlated with changes in mood for arousal (r = −0.25, P = .03), valence (r = −0.37, P < .001), and pride (r = −0.32, P = .004). We refer to these analyses as ‘Model 0’ as they are expected results already in the previous publication (Wainio-Theberge et al. 2024); they are reported again here for completeness due to the slight differences in the final sample composition produced by the EEG preprocessing and data exclusion.

Posture manipulations have minimal effects on EEG frequency bands

We first explored whether expansive and contractive postures would differ in their effects on classic EEG frequency bands. To do this, we conducted independent samples t-tests on all sensors, comparing the normalized post-posture resting state power between each posture group. We found differences between the expansive and contractive postures in a frontal cluster of sensors in the theta and beta bands; expansive postures showed higher power in both bands (Fig. 1). However, this effect did not survive the cluster-based permutation test procedure, which takes into account multiple comparisons (theta: cluster P = .27; beta: cluster P = .40).

Figure 1.

Figure 1.

Open in a new tab

Results of the exploratory analysis of frequency band power as a function of posture condition. Top row of each set represents the topography of power in each posture condition, normalized to the baseline pre-posture resting state. Bottom row represents the difference in power between conditions. Black dots represent electrodes which formed part of a cluster where significant channel-level condition differences were found, which were subjected to the cluster-based permutation test; no clusters were found significant using the cluster test.

Expansive and contractive postures affect frontal asymmetry

We next examined posture effects on our two a priori EEG markers of arousal and valence, namely the EEG spectral exponent and frontal asymmetry. We analysed these metrics with linear models, examining posture effects on post-posture values of the metrics while controlling for pre-posture values (conceptually equivalent to analysing the pre-post difference; see ‘Model 1’, Materials and methods). We found no significant posture differences in the EEG spectral exponent (t(101) = 1.01, P = .32). In contrast, a significant posture effect was observed for frontal asymmetry (t(101) = −2.63, P = .01); however, post hoc t-tests in each group separately (‘Models 1c and 1e’) revealed that the effect was in the opposite direction as hypothesized (see Discussion). Namely, we observed a significant right-lateralized frontal alpha asymmetry (FAA) in the contractive group (t(45) = 2.17, P = .04) and a left-lateralized FAA in the expansive one which failed to reach significance (t(55) = −1.63, P = .11).

We then conducted follow-up analyses to determine if the frontal asymmetry group differences were related to the behavioural changes observed, described earlier. We thus entered mood changes (i.e. pre-post arousal, valence, and pride differences) in the previous model, including an interaction term with posture (‘Model 2’, Materials and methods). We found that the frontal asymmetry effect was correlated with valence changes (t(99) = −2.37, P = .02), but not with arousal (t(99) = −0.46, P = .65) or pride (t(99) = −0.97, P = .33) changes; increases in emotional valence were associated with increased left-lateralized FAA. We also observed a trend for a group interaction effect for valence (t(99) = −1.92, P = .06), which prompted us to investigate the expansive and contractive condition separately (a trend was also observed for the interaction with pride: t(99) = −1.7, P = .09). We found that correlations of FAA with valence were only observed in the expansive condition (expansive: t(54) = −3.18, P = .002; contractive: t(44) = −0.16, P = .87; Fig. 2).

Figure 2.

Figure 2.

Open in a new tab

Results of the analysis of frontal asymmetry in the power posing paradigm. (a) Topography of alpha power in expansive and contractive conditions—residuals are plotted after regressing out pre-posture alpha power from each electrode. Black dots indicate the electrodes used for the computation of frontal asymmetry. (b) Correlations of frontal asymmetry with valence in each posture condition. Each dot represents one participant. Lines of best fit are plotted in solid black, with their 95% CI plotted in the dashed lines. Partial correlation coefficients (controlling for pre-posture asymmetry) are indicated.

Finally, we examined two additional variables as possible confounders or interacting variables with the effects observed here. First, we added the level of difficulty participants experienced in adopting the posture as a covariate (‘Model 3’). We found a significant main effect of difficulty (t(95) = −2.94, P = .004), such that participants who found the posture more difficult experienced less left-lateralized FAA. We also observed a trend for a two-way interaction between mood and posture similar to the one in ‘Model 2’ (t(95) = −1.85, P = .07). Finally, there was a three-way significant interaction between difficulty, mood, and posture (t(95) = 3.46, P < .001). Examining each condition separately (‘Model 3c’ and ‘Model 3e’), we found an interaction effect between difficulty and mood change within the expansive condition (t(52) = 2.55, P = .01), and a trend for an interaction in the opposite direction in the contractive condition (t(42) = −1.99, P = .05). Follow-up analyses performing a median split on difficulty indicated that correlations between ‘valence’ and asymmetry in the expansive condition were highest for participants who found the posture less difficult (rpartial = −0.55, P = .003), compared to those who found it more difficult (rpartial = −0.16, P = .44), while correlations between mood and asymmetry in the contractive condition were highest for those who found the posture more difficult (more difficult: rpartial = −0.42, P = .06; rpartial = 0.25, P = .27). We also note that there was a group difference in difficulty, as participants found the contractive condition more difficult overall (t(102) = −2.3, P = .02); importantly, the main effect of posture on asymmetry also remained significant when controlling for difficulty (t(100) = −2.9, P = .005).

Finally, following our previous observation in an overlapping dataset, and replicated here, that quantitative measures of neck angle mediated posture effects, we investigated whether neck angle would likewise be a better predictor of frontal asymmetry than the experimental posture condition alone (‘Model 4’). Entering neck angle as a covariate into the posture-asymmetry model, we found a trend for a main effect of neck angle (t(73) = 1.95, P = .05), with participants with more upright neck angles exhibiting more left-lateralized FAA; this was driven by the expansive condition (Figure 3). There was also a significant interaction effect between neck angle and ‘valence’ (t(73) = 2.54, P = .01), such that participants with more upright neck angles exhibited larger correlations of asymmetry and valence; this mirrors the interaction with the categorical posture variable observed in ‘Model 2’.

Figure 3.

Figure 3.

Open in a new tab

Correlations of frontal asymmetry with participant neck angle in each posture condition. As Fig. 2b, each dot represents one participant. Lines of best fit are plotted in solid black, with their 95% CIs plotted in the dashed lines. Partial correlation coefficients (controlling for pre-posture asymmetry) are indicated.

Additional analyses for robustness

Results from ‘Models 0–4 and 1–4c,e’, as well as the complete model syntax are reported in full in Supplementary Table S1. As this was an exploratory study that was not preregistered, we conducted several analyses to ensure the robustness of our findings to different preprocessing steps, analysis approaches, and exclusion criteria. These are reported in Supplementary Table S2; results generally confirm the robustness of the findings presented here.

Discussion

Here, we report the first analysis, to our knowledge, of the neural correlates of postural feedback in a ‘power posing’ paradigm. Participants adopted either an expansive or contractive posture with mood questionnaires and resting-state EEG obtained before and after the posture. Qualitatively, the involvement of prefrontal regions was consistent across frequency bands and concurs with fMRI studies of social power mentioned in the Introduction (Chiao 2010, Li et al. 2021): although we found no significant differences between the posture conditions when correcting for multiple comparisons, at the channel level we did observe higher theta and beta power over frontal sensors in the expansive posture. We also analysed two a priori putative neural markers of arousal and valence. For the former, we used an EEG spectral exponent, which revealed no significant effects of pose. In contrast, we did find a significant association between posture and frontal asymmetry, traditionally viewed as an index of valence (Coan and Allen 2003). Intriguingly, this effect was in the opposite direction as hypothesized: relative to contractive postures, expansive postures were associated with increases in left-lateralized frontal alpha asymmetry. Given alpha’s negative correlation with underlying cortical activity, this suggests a ‘right’-lateralized asymmetry in neural activity, which has typically been observed in depression and negative affect in both EEG and fMRI studies (Allen and Reznik 2015); we will thus refer to the underlying brain activity throughout the Discussion, to put our findings in context with other imaging modalities.

Our finding relating right frontal cortical activity to the behavioural effects of expansive postures, namely a significant positive correlation between right frontal asymmetry and increases in valence, appears counterintuitive, given that right frontal cortical activity has typically been associated with negative affect and depression (Allen and Reznik 2015). However, some researchers have noted that while left-lateralized frontal cortical activity appears to consistently relate to approach motivation, right-lateralized frontal activity does not consistently relate to avoidance (Harmon-Jones and Gable 2018). In particular, it has been proposed (Gable et al. 2018) that right frontal asymmetry is related to a Behavioural Inhibition System (r-BIS) (Gray and McNaughton 2000), which, rather than promoting avoidance or withdrawal behaviours, serves to regulate motivational conflicts between approach and avoidance systems and is implicated in cognitive control. Although still limited, there is growing evidence supporting the proposal that right frontal cortical activity reflects activity of this r-BIS system, rather than avoidance. For example, Lacey et al. (2020) found that participants who needed to suppress their emotional responses (including avoidance responses) for a reward showed increased right-lateralized frontal activity, and Nash et al. (2012) found associations between right frontal cortical activity and the amplitude of the error-related negativity, an ERP component associated with conflict monitoring. Moreover, it has been suggested that the r-BIS (and right frontal asymmetry) is particularly engaged in individuals who are in an unstable power state, associated with uncontrollability and potential threats to its maintenance (Deng et al. 2018), which is arguably the case here: individuals achieved a position of perceived power through artificial means, in the absence of any information about possible outcomes arising from that position. In this context, the r-BIS is activated to prepare the individual to resolve any conflict that may threaten the status quo and help them maintain their powerful status. This interpretation is consistent with the finding that the right frontal asymmetry, and index of the r-BIS system, correlated with post-posture positive valence, particularly in those who found the posture natural and comfortable. Nonetheless, additional experiments are needed to replicate our results and further clarify the relationship between powerful postures and behavioural activation and inhibition. Examining the interactions of expansive and contractive postures with actual power roles (which have been shown to increase left frontal asymmetry; Boksem et al. 2012) may be fruitful; expansive postures in a low-power role or contractive postures in a high-power role may increase motivational conflict, while expansive postures in a high-power role should increase behavioural activation and contractive postures in a low-power role should increase inhibition.

Another potential reason for the unexpected effects of posture on asymmetry is that while power is generally thought to activate approach motivation and the behavioural activation system, this may not be the case for manipulations of postural expansion and contraction. Notably, the contractive posture, in slumping the torso and putting the head down, encourages the participant to lean forward, while the expansive posture encourages them to lean backwards; this forward and backward lean has itself been previously found to activate approach and avoidance tendencies (Price and Harmon-Jones 2016), such that the forward lean in the contractive posture is associated with approach motivation and the backward lean in expansion is associated with avoidance. This could create paradoxical effects, whereby the expansive posture increases neural signatures of avoidance while simultaneously being associated with approach-related behavioural outcomes such as positive affect. Future studies could address this conflict by separating different components of postural expansion and contraction: the horizontal dimension of postural expansion (i.e. opening the chest to make oneself appear larger) is not confounded with forward or backward leans, and comparing these components of expansive and contractive postures has been recommended for different reasons by Körner and Schütz (2020). We note that both of these hypotheses are speculative and do not represent the full breadth of possible interpretations for our findings; they are provided to encourage future researchers to replicate our effects and investigate the contexts that might modulate them.

Limitations and future directions

The present work did have a number of limitations that need to be taken into account when interpreting the findings reported. First, our sample size, while similar to, or even larger than, most typical EEG studies, could be considered relatively small in the context of behavioural power pose experiments. Furthermore, this sample was further reduced due to a high rate of data loss; while this was a deliberate strategy to minimize noise in the EEG data, it resulted in several near- or marginally significant effects that need to be confirmed, or not, in future studies. As mentioned in the Introduction, our analysis was largely exploratory, given the lack of previous findings, or a strong theoretical framework, on the neural correlates of power-related induced postures. Nonetheless, we believe our initial findings should be useful to generate specific hypotheses in future, preregistered studies. Finally, as is common in psychology studies, our sample was predominantly (75%) women; as sex differences are of particular relevance for research on social dominance (Cale and Lilienfeld 2002), future studies should attempt to recruit more men, or make use of individual difference measures which measure ‘masculine’ traits in the population as a whole (Mahalik et al. 2003).

Conclusion

Expansive and contractive body postures are used extensively in social communication to signal dominance and submission; adopting these postures can have feedback effects on the organism adopting them, causing them to exhibit some of the traits and behaviours of dominant or submissive individuals. While this feedback from expansive and contractive body postures has been well studied in the context of the ‘power posing’ paradigm, the neural mechanisms of these effects have never been investigated. The present study reports the first neuroimaging investigation of power pose effects. Contrary to our hypothesis, we found that expansive postures increased right-lateralized frontal alpha activity, traditionally seen as a marker of avoidance motivation or negative affect. While future research is needed to confirm this, we tentatively interpret these results in the context of motivational conflicts between the posture and the participant’s overall social role and context.

Supplementary Material

nsaf036_Supp
nsaf036_supp.zip (30.5KB, zip)

Acknowledgements

We wish to thank Martin Frébourg and Kaitlin Pearson, who generously assisted with the data collection. This research was made possible with infrastructure from the Center for Research on Brain, Language, and Music, a strategic cluster funded by the Fonds de Recherche du Québec.

Contributor Information

Soren Wainio-Theberge, Department of Psychiatry, McGill University, Montreal, QC H3A 1A1, Canada; Douglas Mental Health University Institute, Montreal, QC H4H 1R3, Canada; Centre for Research on Brain, Language, and Music, Montreal, QC H3A 1G1, Canada.

Jorge L Armony, Department of Psychiatry, McGill University, Montreal, QC H3A 1A1, Canada; Douglas Mental Health University Institute, Montreal, QC H4H 1R3, Canada; Centre for Research on Brain, Language, and Music, Montreal, QC H3A 1G1, Canada.

Author contributions

Soren Wainio-Theberge (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Validation, Visualization, Writing—original draft, Writing—review and editing) and Jorge L. Armony (Conceptualization, Project administration, Funding acquisition, Resources, Supervision, Writing—review and editing)

Supplementary data

Supplementary data is available at SCAN online.

Conflict of interest:

None declared.

Funding

This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC, no. 2017‐05832) and the Social Sciences and Humanities Research Council of Canada (no. 435-2019-1211) to J.L.A. S.W.-T. was funded by a Canada Graduate Scholarship–Master’s fellowship from NSERC.

Data availability

Data and code underlying the present article are available at https://osf.io/9x8rn/.

References

  1. Allen  JJ, Reznik  SJ. Frontal EEG asymmetry as a promising marker of depression vulnerability: summary and methodological considerations. Curr Opin Psychol  2015;4:93–97. doi: 10.1016/j.copsyc.2014.12.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Boksem  MAS, Smolders  R, Cremer  DD. Social power and approach-related neural activity. Soc Cogn Affect Neurosci  2012;7:516–20. doi: 10.1093/scan/nsp006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Burgoon  JK, Dunbar  NE. Nonverbal expressions of dominance and power in human relationships. Sage Handbook Nonverbal Commun  2006;2:279–97. [Google Scholar]
  4. Buzsáki  G. Rhythms of the Brain. New York (NY): Oxford University Press, 2006, xv, 448 doi: 10.1093/acprof:oso/9780195301069.001.0001 [DOI] [Google Scholar]
  5. Cale  EM, Lilienfeld  SO. Sex differences in psychopathy and antisocial personality disorder: a review and integration. Clinic Psychol Rev  2002;22:1179–207. doi: 10.1016/S0272-7358(01)00125-8 [DOI] [PubMed] [Google Scholar]
  6. Cao  Z, Hidalgo Martinez  G, Simon  T  et al.  OpenPose: realtime multi-person 2D pose estimation using part affinity fields. IEEE Trans Pattern Anal Mach Intell  2019;43:172–86.. doi: 10.1109/TPAMI.2019.2929257 [DOI] [PubMed] [Google Scholar]
  7. Carney  DR, Cuddy  AJC, Yap  AJ. Power posing: brief nonverbal displays affect neuroendocrine levels and risk tolerance. Psychol Sci  2010;21:1363–68. doi: 10.1177/0956797610383437 [DOI] [PubMed] [Google Scholar]
  8. Cesario  J, McDonald  M. Bodies in context: power poses as a computation of action possibility. Soc Cogn  2013;31:260–74. doi: 10.1521/soco.2013.31.2.260 [DOI] [Google Scholar]
  9. Chiao  JY. Neural basis of social status hierarchy across species. Curr Opin Neurobiol  2010;20:803–09. doi: 10.1016/j.conb.2010.08.006 [DOI] [PubMed] [Google Scholar]
  10. Coan  JA, Allen  JJB. The state and trait nature of frontal EEG asymmetry in emotion. In: Hugdahl  K and Davidson  RJ (eds), The Asymmetrical Brain. MIT Press, 2003, 565–615 [Google Scholar]
  11. Corcoran  AW, Alday  PM, Schlesewsky  M  et al.  Toward a reliable, automated method of individual alpha frequency (IAF) quantification. Psychophysiology  2018;55:e13064. doi: 10.1111/psyp.13064 [DOI] [PubMed] [Google Scholar]
  12. Delorme  A, Makeig  S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Methods  2004;134:9–21. doi: 10.1016/j.jneumeth.2003.10.009 [DOI] [PubMed] [Google Scholar]
  13. Deng  M, Zheng  M, Guinote  A. When does power trigger approach motivation? Threats and the role of perceived control in the power domain. Soc Pers Psychol Compass  2018;12:e12390. doi: 10.1111/spc3.12390 [DOI] [Google Scholar]
  14. Elkjaer  E, Mikkelsen  M, Michalak  J  et al. Expansive and contractive postures and movement: a systematic review and meta-analysis of the effect of motor displays on affective and behavioral responses. Perspect Psychol Sci  2022;17:276–304. doi: 10.1177/1745691620919358 [DOI] [PubMed] [Google Scholar]
  15. Gabard-Durnam  LJ, Mendez Leal  AS, Wilkinson  CL  et al.  The Harvard automated processing pipeline for electroencephalography (HAPPE): standardized processing software for developmental and high-artifact data. Front Neurosci  2018;12. doi: 10.3389/fnins.2018.00097 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gable  PA, Neal  LB, Threadgill  AH. Regulatory behavior and frontal activity: considering the role of revised-BIS in relative right frontal asymmetry. Psychophysiology  2018;55:e12910. doi: 10.1111/psyp.12910 [DOI] [PubMed] [Google Scholar]
  17. Gotlib  IH. EEG Alpha asymmetry, depression, and cognitive functioning. Cogn Emot  1998;12:449–78. doi: 10.1080/026999398379673 [DOI] [Google Scholar]
  18. Gray  JA, McNaughton  N. The Neuropsychology of Anxiety: An Enquiry into the Functions of the Septo-hippocampal System. New York (NY): Oxford University Press, 2000. [Google Scholar]
  19. Gronau  QF, Erp  SV, Heck  DW  et al.  A Bayesian model-averaged meta-analysis of the power pose effect with informed and default priors: the case of felt power. Compr Results Soc Psychol  2017;2:123–38. doi: 10.1080/23743603.2017.1326760 [DOI] [Google Scholar]
  20. Hall  JA, Coats  EJ, LeBeau  LS. Nonverbal behavior and the vertical dimension of social relations: a meta-analysis. Psychol Bull  2005;131:898–924. doi: 10.1037/0033-2909.131.6.898 [DOI] [PubMed] [Google Scholar]
  21. Harmon-Jones  E, Gable  PA. On the role of asymmetric frontal cortical activity in approach and withdrawal motivation: an updated review of the evidence. Psychophysiology  2018;55:e12879. doi: 10.1111/psyp.12879 [DOI] [PubMed] [Google Scholar]
  22. Harmon-Jones  E, Gable  PA, Price  TF. Leaning embodies desire: evidence that leaning forward increases relative left frontal cortical activation to appetitive stimuli. Biol Psychol  2011;87:311–13. doi: 10.1016/j.biopsycho.2011.03.009 [DOI] [PubMed] [Google Scholar]
  23. Johnson  SL, Leedom  LJ, Muhtadie  L. The dominance behavioral system and psychopathology: evidence from self-report, observational, and biological studies. Psychol Bull  2012;138:692–743. doi: 10.1037/a0027503 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Jonas  KJ, Cesario  J, Alger  M  et al.  Power poses—where do we stand?  Compr Results Soc Psychol  2017;2:139–41. doi: 10.1080/23743603.2017.1342447 [DOI] [Google Scholar]
  25. Kolvoort  IR, Wainio‐Theberge  S, Wolff  A  et al.  Temporal integration as “common currency” of brain and self-scale-free activity in resting-state EEG correlates with temporal delay effects on self-relatedness. Human Brain Mapp  2020;41:4355–74. doi: 10.1002/hbm.25129 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Körner  R, Petersen  L-E, Schütz  A. Do expansive or contractive body postures affect feelings of self-worth? High power poses impact state self-esteem. Curr Psychol  2021;40:4112–24. doi: 10.1007/s12144-019-00371-1 [DOI] [Google Scholar]
  27. Körner  R, Röseler  L, Schuetz  A  et al.  Dominance and prestige: meta-analytic review of experimentally induced body position effects on behavioral, self-report, and physiological dependent variables. Psychol Bull  2022;148:67–85. doi: 10.1037/bul0000356 [DOI] [Google Scholar]
  28. Körner  R, Schütz  A. Dominance or prestige: a review of the effects of power poses and other body postures. Soc Pers Psychol Compass  2020;14:e12559. doi: 10.1111/spc3.12559 [DOI] [Google Scholar]
  29. Koski  J, Xie  H, Olson  IR. Understanding social hierarchies: the neural and psychological foundations of status perception. Soc Neurosci  2015;10:527–50. doi: 10.1080/17470919.2015.1013223 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lacey  MF, Neal  LB, Gable  PA. Effortful control of motivation, not withdrawal motivation, relates to greater right frontal asymmetry. Int J Psychophysiol  2020;147:18–25. doi: 10.1016/j.ijpsycho.2019.09.013 [DOI] [PubMed] [Google Scholar]
  31. Laufs  H, Kleinschmidt  A, Beyerle  A  et al.  EEG-correlated fMRI of human alpha activity. NeuroImage  2003;19:1463–76. doi: 10.1016/S1053-8119(03)00286-6 [DOI] [PubMed] [Google Scholar]
  32. Lee  -Y-Y, Hsieh  S, Daunizeau  J. Classifying different emotional states by means of EEG-based functional connectivity patterns. PLoS One  2014;9:e95415. doi: 10.1371/journal.pone.0095415 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Lendner  JD, Helfrich  RF, Mander  BA  et al.  An electrophysiological marker of arousal level in humans. eLife  2020;9:e55092. doi: 10.7554/eLife.55092 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Li  S, Krueger  F, Camilleri  JA  et al.  The neural signatures of social hierarchy-related learning and interaction: a coordinate- and connectivity-based meta-analysis. NeuroImage  2021;245:118731. doi: 10.1016/j.neuroimage.2021.118731 [DOI] [PubMed] [Google Scholar]
  35. Mahalik  JR, Locke  BD, Ludlow  LH  et al.  Development of the conformity to masculine norms inventory. Psychol Men Masc  2003;4:3. doi: 10.1037/1524-9220.4.1.3 [DOI] [Google Scholar]
  36. Maris  E, Oostenveld  R. Nonparametric statistical testing of EEG- and MEG-data. J Neurosci Methods  2007;164:177–90. doi: 10.1016/j.jneumeth.2007.03.024 [DOI] [PubMed] [Google Scholar]
  37. McKenzie  D. Beyond baseline and follow-up: the case for more T in experiments. J Dev Agric Econ  2012;99:210–21. doi: 10.1016/j.jdeveco.2012.01.002 [DOI] [Google Scholar]
  38. Muthukumaraswamy  SD. High-frequency brain activity and muscle artifacts in MEG/EEG: a review and recommendations. Front Hum Neurosci  2013;7:138. doi: 10.3389/fnhum.2013.00138 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Nair  S, Sagar  M, Sollers  J  III  et al.  Do slumped and upright postures affect stress responses? A randomized trial. Health Psychol  2015;34:632–41. doi: 10.1037/hea0000146 [DOI] [PubMed] [Google Scholar]
  40. Nash  K, Inzlicht  M, McGregor  I. Approach-related left prefrontal EEG asymmetry predicts muted error-related negativity. Biol Psychol  2012;91:96–102. doi: 10.1016/j.biopsycho.2012.05.005 [DOI] [PubMed] [Google Scholar]
  41. Nolan  H, Whelan  R, Reilly  RB. FASTER: fully automated statistical thresholding for EEG artifact rejection. J Neurosci Methods  2010;192:152–62. doi: 10.1016/j.jneumeth.2010.07.015 [DOI] [PubMed] [Google Scholar]
  42. Oostenveld  R, Fries  P, Maris  E  et al.  FieldTrip: open source software for advanced analysis of MEG, EEG, and invasive electrophysiological data. Comput Intell Neurosci  2011;2011:1–9. doi: 10.1155/2011/156869 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Posner  J, Russell  JA, Peterson  BS. The circumplex model of affect: an integrative approach to affective neuroscience, cognitive development, and psychopathology. Dev Psychopathol  2005;17:715–34. doi: 10.1017/S0954579405050340 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Price  TF, Harmon-Jones  E. Chapter three—embodying approach motivation: a review of recent evidence. In: Elliot  AJ (ed.), Advances in Motivation Science, Vol. 3. Cambridge (MA): Elsevier, 2016, 81–111. doi: 10.1016/bs.adms.2015.12.002 [DOI] [Google Scholar]
  45. Senn  S. Change from baseline and analysis of covariance revisited. Stat Med  2006;25:4334–44. doi: 10.1002/sim.2682 [DOI] [PubMed] [Google Scholar]
  46. Shapiro  CM, Auch  C, Reimer  M  et al.  A new approach to the construct of alertness. J Psychosom Res  2006;60:595–603. doi: 10.1016/j.jpsychores.2006.04.012 [DOI] [PubMed] [Google Scholar]
  47. Simmons  JP, Simonsohn  U. Power posing: P-curving the evidence. Psychol Sci  2017;28:687–93. doi: 10.1177/0956797616658563 [DOI] [PubMed] [Google Scholar]
  48. Smith  EE, Reznik  SJ, Stewart  JL  et al.  Assessing and conceptualizing frontal EEG asymmetry: an updated primer on recording, processing, analyzing, and interpreting frontal alpha asymmetry. Int J Psychophysiol  2017;111:98–114. doi: 10.1016/j.ijpsycho.2016.11.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Smulders  FTY, ten Oever  S, Donkers  FCL  et al.  Single‐trial log transformation is optimal in frequency analysis of resting EEG alpha. Eur J Neurosci  2018;48:2585–98. doi: 10.1111/ejn.13854 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Tracy  J, Robins  R. The psychological structure of pride: a tale of two facets. J Pers Soc Psychol  2007;92:506–25. doi: 10.1037/0022-3514.92.3.506 [DOI] [PubMed] [Google Scholar]
  51. Tsai  J, Knutson  B, Fung  H. Cultural variation in affect valuation. J Pers Soc Psychol  2006;90:288–307. doi: 10.1037/0022-3514.90.2.288 [DOI] [PubMed] [Google Scholar]
  52. Wainio-Theberge  S, Armony  JL. Differences in natural standing posture are associated with antisocial and manipulative personality traits. J Pers Soc Psychol  2024;127:1089–102. doi: 10.1037/pspp0000515 [DOI] [PubMed] [Google Scholar]
  53. Wainio-Theberge  S, Spiousas  I, Armony  JL. Quantitative measures of neck flexion mediate behavioural effects of postural feedback. Emotion, in review  2024. doi: 10.31234/osf.io/kt6xr [DOI]
  54. Wainio-Theberge  S, Wolff  A, Gomez-Pilar  J  et al.  Variability and task-responsiveness of electrophysiological dynamics: scale-free stability and oscillatory flexibility. NeuroImage  2022;256:119245. doi: 10.1016/j.neuroimage.2022.119245 [DOI] [PubMed] [Google Scholar]
  55. Weisfeld  GE, Beresford  JM. Erectness of posture as an indicator of dominance or success in humans. Motiv Emot  1982;6:113–31. doi: 10.1007/BF00992459 [DOI] [Google Scholar]
  56. Welch  P. The use of fast Fourier transform for the estimation of power spectra: a method based on time averaging over short, modified periodograms. IEEE Trans Audio Electroacoust  1967;15:70–73. doi: 10.1109/TAU.1967.1161901 [DOI] [Google Scholar]
  57. Winkler  I, Haufe  S, Tangermann  M. Automatic classification of artifactual ICA-components for artifact removal in EEG signals. Behav Brain Funct  2011;7:30. doi: 10.1186/1744-9081-7-30 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Yang  K, Tong  L, Shu  J  et al.  High gamma band EEG closely related to emotion: evidence from functional network. Front Hum Neurosci  2020;14. doi: 10.3389/fnhum.2020.00089 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Yap  A, Wazlawek  A, Lucas  B  et al.  The ergonomics of dishonesty: the effect of incidental posture on stealing, cheating, and traffic violations. Psychol Sci  2013;24:2281–89. doi: 10.1177/0956797613492425 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

nsaf036_Supp
nsaf036_supp.zip (30.5KB, zip)

Data Availability Statement

Data and code underlying the present article are available at https://osf.io/9x8rn/.

Articles from Social Cognitive and Affective Neuroscience are provided here courtesy of Oxford University Press

RESOURCES