Nice people are easy to smile with: Deliberate imitation of facial expressions depends on personal characteristics

Facial expressions are important in social interactions, where they are often mutually exchanged in face-to-face situations but little is known about the factors that influence such deliberate expressions. Here, we investigated the effect of personality characteristics on voluntary facial expressions. A short statement describing a positive, negative, or neutral characteristic of a target person was followed by her smiling or frowning portrait. Participants were to imitate this expression. Reaction times and accuracy of the facial responses were derived from electromyography of M. corrugator and M. zygomaticus major ; in addition, EEG-derived event-related potentials (ERP) were obtained. Responses were faster and more accurate when facial expressions were congruent with the personal characteristics. Congruency effects in ERPs, were observed in the P200 component, the early posterior negativity (EPN), and the P300. Hence, personal characteristics can modulate the deliberate imitation of facial expressions, based on modulations of reflexive attention and proceeding through several cognitive processing levels. This is the first demonstration that deliberate expressions of emotions are influenced by affective knowledge about the communication partner. 1 Introduction Facial expressions are important venues of communication between people (Huang et al., 2024) – correctly interpreting and responding to others’ facial expressions is conducive to social adaptation and interaction. In real life we often have to regulate our emotional expressions, for example, we sometimes have to smile at or smile with people we do not like because of their previous actions or their expressed opinions. In such a situation, conflicts may arise between our attitude towards the communication partner and the facial expression required in the situation. The present study instructed participants to imitate the facial expression of a (fictitious) communication partner who was associated with positive, negative or neutral personal information via a short verbal characteristic. This setup reflects many real-life situations where an exchange of emotional expressions takes place with a person about whom there is affectively loaded biographical knowledge. This exchange requires both the perception and production of facial expressions, and the access to the biographical information about the partner. The neurocognitive system of expression recognition includes several key components. In their model, Haxby et al. (2000) argued that face processing is hierarchical and comprises and an extended system. The core system includes the fusiform gyrus for perceiving invariant aspects of faces (such as face identity), and the superior temporal sulcus for changeable aspects (e.g., facial expressions and other facial and bodily movements). The extended system integrates additional neural regions, e.g. in the temporal pole to extract social meaning. The cognitive system stores affective and biographical person knowledge, which may influence an individual’s recognition and feedback of facial expressions, for example, returning a smile. Affective person knowledge is the result of learning about the emotionally characteristics of persons that can be formed even based on sparse behavioral information (Abdel Rahman, 2011), such as knowing that someone “beats his wife” or “is a friendly man”. Personal characteristics are relatively stable psychological traits, behavioral patterns, and social functions of an individual, which together constitute an individual’s identity and influence their behavior in different situations (Bateman & Crant, 1993). According to the “Big Two” theory (Abele et al., 2013; Fiske et al., 2002), social cognition consists of two basic dimensions, warmth and competence. These dimensions serve as key criteria in social cognition research, such as in studies exploring the relationship between job applications and hiring decisions (Susanne et al., 2021). They are also used to examine how preferences for attractiveness, social status, and interpersonal warmth align with the “big two” personality dimensions, helping to explain the similarity-attractiveness hypothesis in mate selection (Gebauer et al., 2012). Consequently, human traits, characteristics and personality descriptors are frequently used as independent variables or stimuli in empirical studies on person perception as well as social cognition. These traits serve various purposes, such as facilitating impression formation (Park et al., 2016), activating stereotypes (Wang et al., 2019), or supporting trait-based reasoning (Shi et al., 2019). More specifically, personal characteristics may facilitate the perception and recall of facial expressions (Chang et al., 2024), when they are congruent with the valence of personality information (Luo et al., 2016). Shen et al. (2022) found that positive personal characteristics drive unconscious mimicry more than negative personal characteristics. Although effects of personal traits on face imitation have been found — for instance, impressions of others’ traits (e.g., perceiving them as friendly or hateful) can systematically modulate unconscious facial mimicry (Likowski et al., 2008) — these effects remain at the unconscious level. However, to the best of our knowledge, conscious and deliberate imitation of facial expressions and it’s modulation by affective knowledge about the communication partner has not yet been explored. Among the neurocognitive components of face processing, the storage and extraction of affective biographical knowledge about communication partners plays a particularly important role (Abdel Rahman, 2011; Suess et al., 2015) because they help to understand and interpret current facial expressions based on previous experience (Schoeneman Patel et al., 2022). Biographical knowledge may affect the interpretation of emotional expressions of unfamiliar faces (Abdel Rahman, 2011). For example, Suess and colleagues (2015) found that facial expressions were judged as more negative after learning negative emotional biographical information about the person. Conversely, after learning positive biographical knowledge, these faces were rated as more pleasant than faces associated with neutral biographical knowledge and elicited emotion-typical ERP responses (Abdel Rahman, 2011). Apparently, upon seeing a face, affective person knowledge may rapidly influence perception and later processes, including emotional responses and social evaluations (Todorov et al., 2007). In face-to-face communication it is common to repeat or imitate the partner’s facial expression. Embodied simulation theory (Gallese, 2014) suggests that we recognize emotions and empathize with others by unconsciously and automatically mimicking their facial expressions. In a seminal study Dimberg (1997) presented participants with pictures of emotional faces and recorded facial EMG; the emotional face pictures covertly and subtly activated the participants’ facial muscles, corresponding to the presented facial expressions even when imitation was not the task (Kraft-Feil et al., 2023). However, it is difficult to link the idea of automatic mimicry with the influence of affective knowledge of emotion processing, as previous studies often lacked facial expression ratings or semantic tasks (Abdel Rahman, 2011; Suess et al., 2014). As a result, most research on cognitive systems influencing facial expressions (Ding & Zhang, 2022 ; Dollinger, et al., 2021; Moret-Tatay et al., 2022; Istvan, Uddin, 2023; Abdel Rahman, 2011; Suess et al., 2014) did not examine facial expression imitation. Nevertheless, there is evidence for the interplay of expression imitation and the cognitive system. Specifically, in a study by Matsuda et al. (2022) unconscious mimicry of facial expressions was inhibited by requiring participants to bite on chopsticks, which led to a decline in accuracy in recognizing facial expressions. Conversely, conscious mimicry of facial expressions improved recognition accuracy in participants with deficits in recognizing facial expressions (Kyranides et al., 2022). Furthermore, Künecke et al. (2014) showed that individual differences in the ability to recognize facially expressed emotions are related to facial mimicry. These findings suggest that (automatic) facial mimicry may be an important element for (conscious) facial expression recognition. In the present study we aim to investigate whether affective person knowledge influences the imitation of a facial expression using a Stroop-like paradigm. Faces of unfamiliar persons were paired with information about emotionally positive, neutral or negative characteristics. The stimulus faces displayed happy or angry expressions, which participants were asked to imitate. Hence, the emotional expression perceived and imitated could be either congruent with the associated characteristic (positive trait/smile; negative trait/frown), incongruent (positive trait/frown; negative trait/smile), or neutral (neutral trait/smile; neutral trait/frown). Congruency effects on imitation performance (error rates or reaction times) would demonstrate that person knowledge influences facial expression perception or imitation. The participants’ imitation behavior was measured by electromyography (EMG). With reference to previous studies (Kawamura et al., 2024; Xu et al., 2023), two muscles, used in frowning and smiling (i.e., M. corrugator and M. zygomaticus major, respectively) were targeted. If congruency effects can indeed be found in the EMG responses, the question arises, where in the neurocognitive system these effects are functionally localized. Due their excellent time resolution and based on knowledge from other paradigms, EEG-derived ERPs can provide functional localization of experimental effects. The occipital P1 component is a positive deflection with a peak latency of 100-130 ms (Mück et al., 2020) and reflect peristriate early visual processes. The N170 is a negative-going deflection over temporo-occipital sites and is taken to indicate structural face processing (Itier et al., 2020). Crucially, N170 amplitude was found to increase to faces with negative biographical information (Krasowski et al., 2021) and negative personality traits (Luo et al., 2016) and when scene and face valence were congruent (Hietanen & Astikainen, 2013). The P200 is a fronto-central positivity around 200ms, which appears reflect early stimulus discrimination and response selection (Olofsson et al., 2007). Kaufmann et al. (2015) and Zinchenko et al. (2017) found it to increase in in response to faces that are context congruent rather than incongruent situations, supporting a role in conflict processing. Furthermore, in the studies of Kong et al. (2012) and Li et al. (2024) P200 was larger to words describing negative rather than positive personality traits. The early posterior negativity (EPN) is a relative negativity around 150-300 ms at occipito-temporal sites, associated with the reflexive visual attention, for example to emotional stimuli (Schupp et al. 2003). EPN amplitudes have been reported to be for faces paired with negative as compared to neutral biographical information (Junghöfer et al., 2017; Krasowski et al., 2021). Importantly, congruency effects have been reported by Aldunate et al. (2024), Diéguez-Risco et al. (2015) and Hietanen and Astikainen (2013) with larger EPN amplitudes when emotional stimuli were congruent rather than incongruent with the context. The P300 is a positive centro-parietal waveform peaking approximately 300–600 ms after stimulus onset. It primarily reflects the allocation of attentional resources, stimulus evaluation, and working memory updating. Zhunussova et al. (2024) found that in congruent conditions where faces and voices matched, the P300 was more positive than in incongruent conditions. Proverbio et al. (2023) reported that negative facial expressions elicited larger P300 amplitudes than positive emotional expressions. Finally, the late posterior positivity (LPP), starting around 400 ms, is a sustained posterior positivity, reflecting underlying emotional reactivity and emotion regulation functions (Kong et al., 2021), which has been linked with motivated attention towards the eliciting event (Schupp et al., 2003). Emotionally stimuli elicit larger LPPs than neutral stimuli (Kong et al., 2012; Ku et al., 2020) and so do neutral faces paired with positive personality description (Luo et al., 2016). However, congruency effects in the LPP are mixed; it was found to be larger in incongruent conditions (Pell et al., 2022) but also larger when the emotional face was congruent with a background scene (Xu et al., 2017). We asked whether the imitation of emotional expressions will be affected by the congruency with affective person characteristics and whether this effect might depend on the valence of the imitated expression. Depending on the functional locus of the interplay between affective person knowledge and the expression to be mimicked we expected congruency effects in the corresponding ERP components. Specifically, effects on visual processes and face perception should bear out on congruency effect on the P1 and N170 amplitude, respectively, whereas higher cognitive loci should be progressively reflected in the subsequent components. 2 Methods 2.1 Participants We performed an a priori power analysis using GPower software. Based on a repeated measures ANOVA with two within-subjects’ factors, assuming a medium effect size, statistical efficacy 1-β = 0.80, indicating 80% efficacy, and significance level α= 0.05, a minimum of 19 participants was required. A total of 33 native German-speaking adults were recruited from the Internet. Referring to previous studies (Künecke et al., 2014; Varcin et al., 2019; Xu et al., 2023), five subjects whose facial imitation was less than 70% correct were excluded; hence, data of 28 participants (10 males, M = 30.10, SD = 6.57 years, 18 females, M = 26.17, SD = 5.90 years) were included in the analysis. According to self-report, all participants had normal or corrected-to-normal visual acuity, were right-handed and had no history of psychological or neurological disorders. Participants gave written informed consent to the study, which had been approved by the Ethics Committee of the Department of Psychology, Humboldt University Berlin, where it was also conducted. 2.2 Stimuli Stimuli were simple declarative sentences containing statements about personality characteristics, in the form of “She/He is a x person.”, where x stands for a personality characteristic. To select the sentence materials, an open-ended questionnaire was used to collect positive, neutral, and negative words describing personality characteristics and after eliminating duplicates, there were 52 neutral, 28 positive, and 27 negative words. . Thirty-six native German-speaking university students (16 males, M = 25.20, SD = 4.27 years, 18 females, M = 24.37, SD = 4.10 years) evaluated the valence and familiarity of the collected words on seven-point Likert scales on the online platform SoSci Survey. Based on these ratings, we selected a total of 36 sentences containing 12 sentences each of positive, negative and neutral characteristics, for example, “He is a friendly person”, “She is an evil person”, or “He is a busy person”, respectively; for a complete list of the stimulus sentences, please see appendix A. Independent samples t- tests were conducted on the valence and familiarity of the three types of words. In terms of valence, positive personality characteristics ( M = 6.47, SD = 0.13) differed significantly from negative personality characteristics ( M = 1.48, SD = 0.20), t (22) = 71.98, p < 0.001, and neutral personality characteristics ( M = 4.03, SD = 0.13 ) t (22) = 46.72, p < 0.001, the difference between negative and neutral personality characteristics was also significant ( t (22) = -36.31, p < 0.001). The mean familiarity of positive personality characteristics ( M = 6.59, SD = 0.26) was significantly different from negative ( M = 6.11, SD = 0.33) t (22) = 4.01, p = 0.001, and neutral personality characteristics ( M = 6.21, SD = 0.47) t (22) = 2.49, p = 0.021, but the familiarity of negative and neutral personality characteristics did not differ significantly ( t (22)= -0.61, p =0.550). Ten stimulus faces each with happy or angry emotional expressions were selected from the Radboud Faces Database (Langner et al., 2010), (five female faces for each expression). Hair, ears, and other features were removed from all images and faces were placed within an oval frame (see Fig. 1 for an example). Figure 1. Examples of male and female stimulus faces, displaying happy and angry expressions. 2.3 Experimental design The experiment used a within-subjects design with factors facial expression (happy, angry) and trait valence (positive, negative, neutral), yielding six experimental condition combinations with 60 trials each, totaling 360 trials. The specific experimental design is shown in Table 1. Table 1. Types of Prime -Target (conditions) as used in the experiment. | positive | happy | Congruent | | angry | Incongruent | | | negative | happy | Incongruent | | angry | Congruent | | | neutral | happy | Neutral | | angry | Neutral | 2.4 Procedure Participants were seated in a quiet room facing an LCD screen. The background screen was gray (RGB = 227/227/227). All sentences were written in German with black Arial font, size 15, all placed at the centerline of the screen. The images were presented at a viewing angle of 4.01º× 5.35º，and at a distance of 70 cm. All visual stimuli were centered at the screen. The personal pronoun she/he corresponded to the gender of the face shown. At the beginning of each trial (see Fig. 2), a sentence containing personal characteristics appeared for 1 s. Then a 500-ms fixation cross was shown, followed by a 2-s presentation of an emotional face. Participants were required to imitate the facial expression while the face was visible. The face was replaced by a stop signal, shown for 900-1000 ms, prompting participants to relax their facial muscles and return to a neutral expression. The interval between the end of the stop-signal and the beginning of the next trial varied between 900 to 1000 ms. All trials were presented randomly. The experiment included a practice phase consisting of 12 trials; the experiment proper consisted in 360 trials, divided into 6 blocks. Participants could take a break of self-determined duration in-between blocks. Figure 2. Schematic diagram of an example trial. Participants were required to imitate the emotional expression of the face as soon as it appeared and to return to a neutral expression after the appearance of the Stop signal. 2.5 Data recording and analysis Facial electromyographic signals (EMG) were acquired through two pairs of Ag/AgCl electrodes, filled with conductive gel and fixed with adhesive pads to the right side of the face above the corrugator and zygomaticus major muscles, according to the configuration described in Fridlund and Cacioppo (1986). The ground electrode was placed at the midline of the forehead, approximately 3-4 cm above the upper border of the inner eyebrow. Impedances were kept below 10 kΩ. Impedances were kept below 10 kΩ. The EMG signal was amplified, rectified, and integrated (TC = 10 ms) using a Coulbourn V76–04a Hi amplifier. Low- and high-pass filters were set at 10 kHz and 8 Hz, respectively. These signals were recorded using BrainVision Recorder software (Brain Products GmbH, Munich, Germany) at sampling rates of 1000 Hz. Continuous EEG was recorded using electrode caps with 32 electrodes mounted according to the extended International 10-20 system. Horizontal electrooculograms (HEOG) and vertical electrooculograms (VEOG) were recorded at the outer canthi of the eyes and from below the left eye. The left mastoid was used as the common reference electrode during data collection. The electrode impedance was kept below 5 kΩ throughout the experiment. Signals were amplified with Brain amp hardware (Brain Products GmbH), at a band pass of 0.05–100 Hz and a sampling rate of 1000 Hz. Brain vision Analyzer 2.1 software was used to preprocess the facial EMG and EEG data. The EMG was preprocessed with 50-Hz notch filtering, the continuous signals were segmented from -200 ms to 2000 ms relative to face presentation onset. The amplitudes and latencies of the facial muscle activities were used to calculate the accuracy rate and response times of the participant’s facial expression imitation. A response was registered whenever the activity of an EMG channel exceeded 20% of the maximum value within the first second after the face had appeared; the time to reach 20% of the maximum amplitude was taken as reaction time (RT). A correct response in a frown trial (requiring the participant to frown to an angry face stimulus) was defined as above-threshold zygomaticus activation in combination with below threshold corrugator activity, while a correct response in a smile trial (requiring a smile to a happy face) required above-threshold zygomaticus and below-threshold corrugator activation. Incorrect trials and trials without any response were excluded as were trials with reaction times < 150 ms; any participants with mean accuracy < 70% were also excluded from the analyses (Kandice et al., 2019; Künecke et al., 2014; Xu et al., 2023). Offline, EEG data were re-calculated to common average reference, filtered from 0.1-30 Hz and subjected to independent component analysis (ICA) to remove artifacts. The Infomax ICA algorithm was used to decompose the continuous data, yielding 32 separate components. Component identification was a semi-automated process: first an automatic pre-classification was performed by the software’s built-in ICA component classifier, followed by manual verification and final determination of each component in combination with topographic, activity time-series, and spectral features and individually recorded transverse and longitudinal electrooculographic waveforms. Components that were jointly determined to be artifacts (mainly ocular and muscle artifacts) were eliminated. After completing the ICA processing, the channel designators were reduced to electrode names by Inverse ICA processing. Then, EEG data were segmented for -200 to 2000 ms. Artifact-free and artefact-corrected trials were averaged separately for conditions, electrodes and participants, regardless of whether the facial response was correct or not. A repeated-measures ANOVA for factor facial expression (happy, angry) and congruency (congruent, incongruent, neutral) was conducted with SPSS 25.0 software on the average number of remaining trials (see Table 2), revealing a significant main effect of facial expressions ( F (1,32) = 5.00, p = 0.032, η p 2 = 0.14); more trials were left for happy expressions ( M = 56.27, SE = 0.92) than for angry expressions ( M = 53.54, SE = 1.32, p = 0.032). No significant main effects for congruency or interaction between the factors was observed ( p s > 0.05). | MIN | MAX | M | SD | || | happy face | congruent | 33 | 60 | 56.15 | 6.51 | | incongruent | 38 | 60 | 56.30 | 4.81 | | | neutral | 36 | 60 | 56.36 | 5.57 | | | Angry face | congruent | 30 | 60 | 53.94 | 7.16 | | incongruent | 31 | 60 | 53.67 | 7.92 | | | neutral | 31 | 60 | 53.00 | 8.11 | Relevant intervals and electrodes were selected with reference to previous studies and slightly adjusted according to the observed waveforms and topographic maps. The P1 peak amplitude was measured at electrodes O1 and O2 within a 90-120 ms time window (Schmuck et al., 2024). The N170 peak amplitude was measured at electrodes P9 and P10 within a 140-180 ms time window (Li et al., 2023; Maximilian et al., 2021). The average P200 amplitude was calculated within a 200-260 ms time window over a posterior ROI comprising POz, Pz, P5, P6, O1, and O2 (Baus et al., 2017). The average EPN amplitude was obtained within a 160-220 ms time window over a centro-parietal ROI comprising P9, P10, PO9, PO10, PO7, PO8, O1 and O2 (Aguado et al., 2019). The average P300 amplitude was calculated within a 270-330 ms time window over a centro-parietal ROI comprising CP3, CP4, CPz, P1, P2 and Pz (Xu et al., 2023). Finally, the average LPP amplitude was derived within a 350-600 ms time window at a centro-parietal ROI comprising CP3, CP4, CPz, P1, P2, Pz (Aguado et al., 2019; Schmuck et al., 2024). All dependent variables were analyzed with the same factors as the number of trials above. For ERP amplitudes a factor electrode was added. For the P1/N170/EPN the factor Electrode was divided into left and right according to the hemispheric location of the electrodes. The electrodes for the P200/P300/LPP were divided into left, right, and midline locations and the electrodes within these sections were averaged. In this case the factor electrode had three levels. Post-hoc tests were conducted with correction for multiple comparisons with LSD (Least Significant Difference). The test level α was set at 0.05 unless stated otherwise. 3 Results 3.1 Performance A main effect of congruency was found in reaction times ( F (2, 54) = 11.25, p < 0.001, η p 2 = 0.29); post-hoc comparisons revealed faster responses in congruent ( M = 456.81, SE = 18.48) than both incongruent ( M = 476.81, SE = 19.89, p = 0.001) and neutral conditions ( M = 474.39, SE = 19.26, p < 0.001), while the difference between incongruent and neutral conditions was not significant ( p = 0.533). In addition, the interaction between facial expression and congruency was significant ( F (2,54) = 3.62, p = 0.033, η p 2 = 0.12). Responses in the congruent condition were faster than in the incongruent and neutral conditions for both happy and angry faces; however, only for angry faces responses were significantly slower in the neutral than incongruent condition (see Fig. 3B). The ANOVA on mean accuracies found a significant main effect of facial expression ( F (1, 27) = 4.11, p = 0.053, η p 2 = 0.13); as compared to angry faces, the accuracy for happy faces was higher (95% vs. 88.3%). The main effect of congruency was significant ( F (2,54) = 8.15, p = 0.001, η p 2 = 0.23), and post-hoc comparisons revealed that the accuracy in congruent conditions ( M =0.93, SE =0.02) was significantly higher than in the incongruent conditions ( M = 0.91, SE = 0.02), p < 0.001; accuracy in the congruent condition was also significantly higher than in the neutral conditions ( M = 0.92, SE = 0.01), p = 0.048; incongruent conditions did not differ significantly from neutral conditions ( p = 0.073). Figure 3. Mean M. zygomaticus major (Zyg) and M. corrugator supercilii (Corr) responses to happy and angry faces (left vs. right panel) in congruent, incongruent and neutral conditions. (A). Mean response times and accuracy rates (including standard errors) per condition (B). 3.2 ERPs 3.2.1 P1 (90-120 ms) Figure 4 shows the P1 component at the occipital electrodes O1 and O2 and its typical scalp topography. The ANOVA of P1 amplitude revealed a significant main effect of facial expression ( F (1, 32) = 4.26, p = 0.047, η p 2 = 0.12), the amplitude was significantly more positive for angry faces ( M = 5.46 µV, SE = 0.66) than happy faces ( M = 5.15µV, SE = 0.64, p = 0.047). The interaction between electrode and facial expression was significant ( F (1, 32) = 6.70, p = 0.014, η p 2 = 0.18): At left electrodes (O1) the amplitude was significantly more positive for angry faces ( M = 5.44µV, SE = 0.69) than happy faces ( M = 4.85µV, SE = 0.63, p = 0.008). Figure 4 . Grand-average ERPs at the O1 and O2 electrodes for responses to happy and angry faces (A) and topography of P1 in the 90 to 120 ms interval (B). Electrode locations are marked in the topography. 3.2.2 N170 (140-180 ms) Figure 5 shows the N170 component at P9 and P10 electrodes and the typical scalp distribution of this component. The ANOVA of the N170 amplitude revealed a significant interaction of electrode and facial expression ( F (1, 32) = 5.20, p = 0.029, η p 2 = 0.14): At both hemispheres, the difference between facial expressions did not reach significance. Figure 5. Grand-average ERPs at the P9 and P10 electrodes for responses to happy and angry faces (A) and topographies of N170 happy and angry faces conditions in the 140 to 180 ms interval (B). Electrode locations are marked in the topography. 3.2.3 EPN (160-260 ms) Figure 5 shows the ERPs at the left and right ROIs of the EPN component and the scalp distribution of the difference waves between congruent minus incongruent and congruent minus neutral. The difference topographies show posterior negativities as would be expected from a component that indicates reflexive visual attention. ANOVA revealed a significant interaction of electrode and facial expression ( F (1, 32) = 6.90, p = 0.013, η p 2 = 0.18); however, there were no significant simple main effects of facial expression at any individual electrode site. Importantly, the three-way interaction of hemisphere, facial expression, and congruency was significant ( F (2, 64) = 3.27, p = 0.044, η p 2 = 0.09): For angry faces, in the left ROI the amplitude was significantly more negative in congruent ( M = -0.39 µV, SE = 0.48) than incongruent condition ( M = -0.08 µV, SE = 0.49, p = 0.026), and neutral condition ( M = -0.13 µV, SE = 0.47, p = 0.021). There was no effect for happy faces. No other significant main effects and interactions were found ( p s > 0.05). Figure 6. Grand-average ERPs at the EPN left ROI (P9, PO9, PO7, O1) and right ROI (P10, PO10, PO8, O2) for responses to happy and angry faces in congruent, incongruent and neutral conditions (A) and the difference topographies of angry faces (congruent minus incongruent and congruent minus neutral) in the 160 to 220 ms interval (B) and Bar graph of the difference in wave amplitude between left and right electrode sites under happy and angry faces (C). Electrode locations are marked in the topographies. 3.2.4 P200 (200-260 ms) Figure 7 shows the P200 component at the left, right, and midline P200 ROIs and the scalp distribution of the difference congruent minus incongruent in the 200-260 ms time window. The difference topography resembles the ones found for the EPN (see Fig. 5) but do not show a notable activity at the typical electrodes for the P200 component. Nevertheless, for the P200 amplitude in the predetermined ROIs and time window there was a significant interaction of factors ROI and congruency ( F (4, 128) = 3.41, p = 0.011, η p 2 = 0.10): At the midline ROI (POz/Pz) the amplitude was significantly more positive in congruent ( M = 2.56µV, SE = 0.39) than neutral condition ( M = 2.37µV, SE = 0.38, p = 0.014). No other significant main effects and interactions were found ( ps > 0.05). We believe the congruency effect at the midline is a spillover from the EPN, which occurs during the same time window and some shows positive activity in the parietal midline (see Fig. 6). Therefore, we will not discuss the congruency effect in the midline P200 as an independent effect. Figure 7. Grand-average ERPs at the left (P5, O1), middle (POz, Pz), and right P200 ROI (P6, O2) P200 for responses to congruent, incongruent and neutral conditions (A) and the difference topography congruent minus incongruent in the 200 to 260 ms interval (B). Electrode locations of the predetermined ROI are marked in the topography. 3.2.5 P300 (270-330 ms) and LPP (350-600 ms) Figure 8 shows the ERPs at left, right, and midline of the P300 ROIs and the difference topography for the congruent minus congruent conditions in the 270-330 ms interval. ANOVA revealed a significant main effect of ROI ( F (2, 64) = 6.69, p = 0.002, η p 2 = 0.17); the amplitude was significantly more positive in the right-hemisphere ROI ( M = 2.91µV, SE = 0.40) than in the left ( M = 2.19 µV, SE = 0.32, p = 0.005), and also more positive than in the midline ROI ( M = 2.52µV, SE = 0.39, p = 0.042), and (as a strong trend) it was more positive in the midline than above the left hemisphere ROI ( p = 0.052). Importantly, the interaction between ROI and congruency was significant ( F (4, 128) = 2.73, p = 0.032, η p 2 = 0.08): At the right ROI (CP4, P2) the P3 amplitude was significantly more positive in congruent ( M = 3.02 µV, SE = 0.40) than incongruent conditions ( M = 2.81 µV, SE = 0.41, p = 0.038). Otherwise, there were no significant main effects or interactions ( ps > 0.05). Figure 8 also shows the ERPs and difference topography for the LPP. ANOVA did not reveal any significant effects for this component (all ps > 0.05). Figure 8. Grand-average ERPs at the left (CP3, P1), midline (CPz, Pz) and right (CP4, P2) P300 and LPP ROIs for congruent, incongruent and neutral conditions with gray shadings for the 270 to 330 ms interval (P300) pink shading for the 350 to 600 ms interval (LPP) (A). Difference topographies congruent minus incongruent in the P300 and LPP intervals (B). Electrode locations of the ROIs are marked in the topographies. 4 Discussion This study investigated the influence of alleged personality characteristics on the imitation performance for facial expressions using a Stroop-like paradigm. EMG results revealed that when personality traits aligned with facial expressions, imitation speed and accuracy were higher compared to incongruent conditions. We used event-related potentials to functionally localize congruency effects between affective information (personality characteristics) and the facial expression imitation performance. We assumed that if the congruency effect is related to processes at that are reflected in specific ERP components, we should observe congruency effects in these components. We analyzed the perception-related components P1 and N170, the EPN, assumed to reflect involuntary attention, the P2 assumed to reflect early stimulus discrimination and response selection, the P3, related to memory activation and the salience of stimuli and the LPP, considered to indicate motivated attention. Except for the P2, which seems not to be present in our data, the observed component topographies conformed reasonably well with the literature. ERP findings demonstrated congruency effects in both the EPN and P300 components. Specifically, the EPN component revealed a more negative amplitude in the left hemisphere for congruent conditions than incongruent conditions during angry face imitation. Furthermore, the P300 components showed more positive amplitudes during congruent than incongruent conditions. 4.1 Performance Our behavioral findings confirm that knowledge about person characteristics influences the deliberate imitation of emotional facial expressions. Specifically, when the valence of the person characteristic is congruent with the valence of the facial expression, participants can imitate facial expressions more quickly and accurately. This aligns with prior research findings on facial expression recognition (Hudson et al., 2019). Knowledge about persona characteristics of the social partner may induce expectations concerning the emotional situation to be encountered (Calbi et al., 2019). This in turn may bias the readiness to produce specific facial expressions or it may even bias imitation. In the present study, positive person characteristics facilitated the imitation of smiles, while negative characteristics facilitated the imitation of frowning responses. Under these conditions, when the required facial expressions align with personal characteristics, participants may more quickly grasp the facial feature (smile or frown) and may more easily chose the appropriate facial expression or more readily produce the expression. The functional locus of the facilitation of facial expressions is not readily derived from performance data; however, as discussed next, ERPs may be helpful at this point. 4.2 ERP Findings Congruency effects were not observed in perceptual processing stages as reflected in the P1 and N170 components. Since one would expect congruency effects to require a confluence of the semantic system processing the information of personal characteristics, the present findings are in line with a relatively late locus of this confluence, but do not add to the evidence of an early locus. Note, however, that although we did not observe it in the current setup, other researchers have shown semantic (knowledge) effects on early face-elicited ERPs (e.g., Abdel Rahman, 2011) and on objects (e.g. Abdel Rahman & Sommer, 2008). It is conceivable that an early locus requires more consolidated or more extensive semantic/affective knowledge than a mere isolated statement as was presented in the current study. In contrast to perceptual ERP components, congruency effects were observed in the EPN and in the P3. The EPN, an increased temporo-occipital negativity, was larger to angry faces that were congruent with the preceding personality information. The EPN has been observed to (both negative and positive) emotional stimuli, whether these are scenes, words or faces (Schacht & Sommer, 2009; Schupp et al., 2003) and an EPN has been observed to large as compared to small non-emotional facial movements (Recio et al. 2017). These findings are consistent with the interpretation of the EPN as indicating the involuntary (reflexive) attention elicited by different kinds of visual stimuli. Accordingly, it seems that the facilitation of the imitation of congruent angry expressions is related to more visual attention elicited by angry expressions that may have been anticipated based on the preceding semantic information about personal characteristics. It remains to be seen whether the absence of a similar effect for positive expressions is systematic or due to insufficient experimental power. More importantly, there was an expression-independent congruency effect in the P3 amplitude, which was larger in congruent than in incongruent conditions. There are many suggestions for the functional interpretation of the P3, for example as reflecting increased attention allocation and stimulus salience (Hajcak et al., 2010) and the updating of memory (Polich, 2007); for a recent review, please see Verleger (2020). Emotional as compared to neutral stimuli have been frequently reported to increase the P3 components and the subsequent LPP, which is commonly interpreted as increased motivated attention and evaluation to emotional stimuli. Interestingly, a recent study found that emotionally congruent face and word stimuli that were presented concurrently, elicited larger P3/LPP amplitudes between 500 and 700 ms than emotionally incongruent stimuli (Huerta-Chavez & Ramos-Loyo, 2024). At first sight, these findings seem to be consistent with the present one. However, the congruency effect reported by Huerta-Chavez and Ramos-Loyo (2024) was found in a relatively late interval (500-700ms) and was not attributed specifically to the P3 or the LPP. In contrast, in the present study the posterior congruency effects were considerably earlier and confined to the typical P3 latency range (270-330 ms), whereas, in the more typical LPC window (350-600 ms) we did not observe a congruency effect. Therefore, we must conclude that the present effect of congruency is related to an increased P3 rather than to any changes in the LPP. These two components can be distinguished and seem to reflect at least partially different processes (Nowparast-Rostami et al. 2015). How can we interpret the congruency effects on the P3 amplitude? First, performance tells us that the congruent condition is easier than the incongruent condition. Discussing the large number of theoretical constructs that have been linked to the P3 or P3b (e.g. Verleger, 2020) as possible explanations for the present findings is outside the present scope future research may try to pin down more specific explanations. In our opinion a plausible option might be semantic priming exerted by the personality information. For example, Schweinberger (1996) used cross-modality associative priming where faces or names of famous persons could be primed by the name or face of a related person, for example the face of Gorbatchov could be preceded by the name of Yeltsin. Consistently, the items that were preceded (primed) by related items elicited larger late positivity than unprimed items. In the present experiment, we seem to be dealing with facilitation coming from the semantic system; due to the absence of perceptual ERP effects of congruency, the semantic priming does not appear to target emotion expression recognition but rather, later, higher level cognitive processes reflected in the P3, such as memory updating, attention allocation, stimulus categorization, etc. This interpretation – vague as it may be – is consistent with the absence of inhibition due to incongruent relative to the unrelated (neutral) conditions. In performance, the incongruent conditions were either indistinguishable from the neutral condition or – in angry faces – responses were even somewhat faster (cf. Fig. 2). Hence, the present effects do not seem to reflect any conflicts between personality-based and expression-based valence, which should have led to performance decrements in incompatible as compared to neutral conditions. Instead, congruency between personality-based and expression-based valence seems to cause a benefit-only situation with no costs when there is incongruence rather than no relationship. The absence of incongruency costs also speaks against a mimicry-based source of the observed effects. If the valence of affective information about personality would bias or prime the mimicry of a specific facial expression incongruent affective information would prime the mimicry of a false expression, increasing error rates and reaction time relative to a non-primed neutral condition. This, however, was not found; therefore, we may conclude that the congruency effect is likely not substantially located in motor mimicry but more probably at higher cognitive stages. 4.3 Limitations As pointed out above, it is not clear whether the absence of a congruency effect in the left EPN and the right P300 for smiling faces is due to more experimental power needed for this condition. In the present ERP analyses we did not distinguish between correct and incorrect responses; the reason was that response correctness depended at least in part on EMG activations that could be influenced by factors unrelated to brain processes. Future studies with enhanced power might allow to address these questions. 4.4 Conclusions and perspectives The present study shows to the best of our knowledge for the first time that deliberate imitation of facial expressions, which is a common situation in face-to-face interaction, is influenced by the congruency with knowledge about the interaction partner, even when based on very sparse information. Thus, it seems to be easier to smile at persons with alleged positive characteristics and easier to frown at “bad” persons. Altogether, when the valence of affective information of a person and the emotional facial expression to be imitated match, performance is facilitated independent of the type of facial expression. This seems to be a benefit-only pattern because we did not find costs of incongruency relative to a no-relation (neutral) condition. The P3 component of the ERP revealed that the expression-independent benefit of congruent conditions is related to higher order cognitive processes. In addition, for angry expressions there seems to be an additional functional locus (EPN) of the congruency benefit during reflexive visual attention when the angry expression matches the negative person characteristics. Since the present experiment is the first that used an expression imitation task to study emotional social synchronization and its neural concomitants, the approach can be extended into many directions, such as different emotional expressions, different social situations, or the degree of familiarity with the interaction partner. Acknowledgments We thank Dr. Anoushiravan Zahedi for his assistance with the experimental setup. We also thank Ulrike Bunzenthal and Thomas Pinkpank for technical support. Funding This research was funded by the China Scholorship Council (CSC）under Grant No. 201908330157.

Abdel Rahman, R. (2011). Facing good and evil: Early brain signatures of affective biographical knowledge in face recognition. Emotion, 11 (6), 1397–1405. https://doi.org/10.1037/a0024717 Abdel Rahman, R., Sommer, W. (2008). Seeing what we know and understand: How knowledge shapes perception. Psychonomic Bulletin & Review, 15 (6), 1055-1063. https://doi.org/10.3758/pbr.15.6.1055 Abele, A. E., Wojciszke, B. (2013). The Big Two in social judgment and behavior. Social Psychology, 44 (2), 61-62. https://doi.org/10.1027/1864-9335/a000137 Aguado, L., Parkington, K. B., Dieguez-Risco, T., et al. (2019). Joint modulation of facial expression processing by contextual congruency and task demands. Brain Sciences, 9 (5), 116-116. https://doi.org/10.3390/brainsci9050116 Aldunate, N., López, V., Rojas-Thomas, F., et al. (2024). Emotional text messages affect the early processing of emoticons depending on their emotional congruence: evidence from the N170 and EPN event related potentials. Cognitive Processing, 25 (4), 621-634. https://doi.org/10.1007/s10339-024-01223-y Bateman, T. S., Crant, J. M. (1993). The proactive component of organizational behavior: A measure and correlates. Journal of Organizational Behavior, 14 (2), 103–118. https://doi.org/10.1002/job.4030140202 Baus, C., Bas, J., Calabria, M., et al. (2017). Speak my language and I will remember your face better: An ERP study. Frontiers in Psychology, 8, 709. https://doi.org/10.3389/fpsyg.2017.00709 Calbi, M., Siri, F., Heimann, K., et al. (2019). How context influences the interpretation of facial expressions: a source localization high-density EEG study on the “Kuleshov effect”. Scientific Reports, 9 (1), 1-16. https://doi.org/10.1038/s41598-018-37786-y Chang, C. H., Drobotenko, N., Ruocco, A. C., et al. (2024). Perception and memory-based representations of facial emotions: Associations with personality functioning, affective states and recognition abilities. Cognition, 245, 105724. https://doi.org/10.1016/j.cognition.2024.105724 Diéguez-Risco, T., Aguado, L., Albert, J., et al. (2015). Judging emotional congruency: explicit attention to situational context modulates processing of facial expressions of emotion. Biological Psychology, 112, 27–38. https://doi.org/10.1016/j.biopsycho.2015.09.012 Dimberg, U. (1997). Social fear and expressive reactions to social stimuli. Scandinavian Journal of Psychology, 38 (3), 171–174.https://doi.org/10.1111/1467-9450.00024 Dollinger, L., Laukka, P., Hogman, L. B., et al. (2021). Training emotion recognition accuracy: Results for multimodal expressions and facial micro expressions. Frontiers in Psychology, 12, 708867. https://doi.org/10.3389/fpsyg.2021.708867 Fiske, S. T., Cuddy, A. J. C., Glick, P., et al. (2002). A model of (often mixed) stereotype content: Competence and warmth respectively follow from perceived status and competition. Journal of Personality and Social Psychology,82 (6), 878-902. https://doi.org/10.1037/0022-3514.82.6.878 Gallese, V. (2014). Bodily selves in relation: embodied simulation as second- person perspective on intersubjectivity. Philosophical Transactions of the Royal Society B: Biological Sciences, 369 (1644), 20130177. https://doi.org/10.1098/rstb.2013.0177 Gebauer, J. E., Leary, M. R., Neberich, W. (2012). Big Two personality and Big Three mate preferences: Similarity attracts, but country-level mate preferences crucially matter. Personality and Social Psychology Bulletin, 38 (12), 1579-1593. https://doi.org/10.1177/0146167212456300 Haxby, J. V., Hoffman, E. A., Gobbini, M. I. (2000). The distributed human neural system for face perception. Trends in Cognitive Sciences,4 (6),223-233. https://doi.org/10.1016/s1364-6613(00)01482-0 Hajcak, G., MacNamara, A., Olvet, D. M. (2010). Event-related potentials, emotion, and emotion regulation: an integrative review. Developmental Neuropsychology, 35 (2), 129-155. https://doi.org/10.1080/87565640903526504 Hietanen, J. K., Astikainen, P. (2013). N170 response to facial expressions is modulated by the affective congruency between the emotional expression and preceding affective picture. Biological Psychology, 92, 114–124. https://doi.org/10.1016/j.biopsycho.2012.10.005 Huang, L., Zhu, T., Cai, J., et al. (2024). Can perceivers differentiate intense facial expressions? Eye movement patterns. Behavioral Sciences, 14 (3), 248. https://doi.org/10.3390/bs14030185 Hudson, A., Henderson, H., Itier, R. (2019). The effect of auditory semantic cues on face expression processing: An EEG investigation. Journal of Vision, 19 (10), 183. https://doi.org/10.1167/19.10.183 Huerta-Chavez, V., Ramos-Loyo, J. (2024). Emotional congruency between faces and words benefits emotional judgments in women: An event-related potential study. Neuroscience Letters, 822, 137644. https://doi.org/10.1016/j.neulet.2024.137644 Junghöfer, M., Rehbein, M. A., Maitzen, J., et al. (2017). An evil face? Verbal evaluative multi-CS conditioning enhances face-evoked mid-latency magnetoencephalographic responses. Social Cognitive and Affective Neuroscience, 12 (4), 695-705.https://doi.org/10.1093/scan/nsw179 Kaufmann, J., End, A., Schweinberger, S. (2015). Individual differences in the activation of mental representations of famous faces by lookalikes. Journal of Vision, 15 (12), 167. https://doi.org/10.1167/15.12.167 Kawamura, N., Sato, W., Shimokawa, K., et al. (2024). Machine learning-based interpretable modeling for subjective emotional dynamics sensing using facial EMG. Sensors, 24 (5), 1536. https://doi.org/10.3390/s24051536 Kong, F., Zhang, Y., Chen, H. (2012). ERP differences between processing of physical characteristics and personality attributes. Behavioral and Brain Functions, 8, 1-10. https://doi.org/10.1186/1744-9081-8-49 Kraft-Feil, T. L., Ingram, R. E., Gorey, C., et al. (2023). The association of negative mood with automatic and effortful facial expression mimicry. Frontiers in Psychology, 14, 1056535. http://dx.doi.org/10.3389/FPSYG.2023.1056535 Krasowski, C., Schindler, S., Bruchmann, M., et al. (2021). Electrophysiological responses to negative evaluative person-knowledge: Effects of individual differences. Cognitive, Affective, & Behavioral Neuroscience, 21, 822-836. https://doi.org/10.1007/s12144-022-02749-0 Ku, L. C., Chan, S. H., Lai, V. T. (2020). Personality traits and emotional word recognition: An ERP study. Cognitive, Affective, & Behavioral Neuroscience, 20, 371-386. https://doi.org/10.3758/s13415-020-00774-9 Künecke, J., Hildebrandt, A., Recio, G., et al. (2014). Facial EMG responses to emotional expressions are related to emotion perception ability. PLoS ONE, 9 (1), e84053. https://doi.org/10.1371/journal.pone.0084053 Kyranides, M. N., Petridou, M., Gokani, H. A., et al. (2022). Reading and reacting to faces, the effect of facial mimicry in improving facial emotion recognition in individuals with antisocial behavior and psychopathic traits. Current Psychology 42 (17), 14306-14319. https://doi.org/10.1007/s12144-022-02749-0 Likowski, K. U., Mühlberger, A., Seibt, B., et al. (2008). Modulation of facial mimicry by attitudes. Journal of e+Experimental Social Psychology, 44 (4), 1065-1072. https://doi.org/10.1016/j.jesp.2007.10.007 Luo, Q. L., Wang, H. L., Dzhelyova, M., et al. (2016). Effect of affective personality information on face processing: Evidence from ERPs. Frontiers in Psychology, 7, 810. https://doi.org/10.3389/fpsyg.2016.00810 Matsuda, N.,Abe, M. O. (2022). Imitation inhibition and facial expression recognition: Imitation-inhibition training inhibits the impact of interference with facial mimicry. Cognitive Studies: Bulletin of the Japanese Cognitive Science Society, 29 (4),587-594. https://doi.org/10.31234/osf.io/7nh6r Mück, M., Ohmann, K., Dummel, S., et al. (2020). Face perception and narcissism: Variations of event-related potential components (P1 & N170) with admiration and rivalry. Cognitive, Affective, & Behavioral Neuroscience, 20 (5), 1041-1055. https://doi.org/10.3758/s13415-020-00818-0 Moret-Tatay, C., Mundi-Ricos, P., Irigaray, T. Q. (2022). The relationship between face processing, cognitive and affective empathy. Behavioral Sciences, 13 (1),24. https://doi.org/10.3390/bs13010021 Pell, M. D., Sethi, S., Rigoulot, S., et al. (2022). Emotional voices modulate perception and predictions about an upcoming face. Cortex, 149, 148-164. https://doi.org/10.1016/j.cortex.2021.12.017 Polich, J. (2007). Updating P300: an integrative theory of P3a and P3b. Clinical Neurophysiology, 118 (10), 2128-2148. https://doi.org/10.1016/j.clinph.2007.04.019 Schmuck, J., Voltz, E., Gibbons, H. (2024). You’re beautiful when you smile: Event-related brain potential (ERP) evidence of early opposite-gender bias in happy faces. Brain Sciences,14 (8),739. https://doi.org/10.3390/brainsci14080739 Schupp, H. T., Junghöfer, M., Weike, A. I., et al. (2003). Emotional facilitation of sensory processing in the visual cortex. Psychological Science, 14 (1), 7-13. https://doi.org/10.1111/1467-9280.01411 Shen, J., Han, S., Zhou, Y., et al. (2022). The influence of facial attractiveness and personal characteristics on imitation. Journal of Social Psychology, 163 (1), 94–106. https://doi.org/10.1080/00224545.2022.2052002 Suess, F., Rabovsky, M., Abdel Rahman, R. (2015). Perceiving emotions in neutral faces: expression processing is biased by affective person knowledge. Social Cognitive and Affective Neuroscience, 10 (4), 531–536. https://doi.org/10.1093/scan/nsu088 Todorov, A., Gobbini, M. I., Evans, K. K., et al. (2007). Spontaneous retrieval of affective person knowledge in face perception. Neuropsychologia, 45 (1), 163-173. https://doi.org/10.1016/j.neuropsychologia.2006.04.018 Varcin, K. J., Grainger, S. A., Richmond, J. L., et al. (2019). A role for affectivity in rapid facial mimicry: An electromyographic study. Social Neuroscience, 14 (5), 608–617. https://doi.org/10.1080/17470919.2018.1564694 Xu, Q., Sommer, W., Recio, G. (2023). Control over emotional facial expressions: Evidence from facial EMG and ERPs in a Stroop-like task. Biological Psychology, 181, 108611. https://doi.org/10.1016/j.biopsycho.2023.108611 Xu, Q., Yang, Y., Tan, Q., et al. (2017). Facial expressions in context: Electrophysiological correlates of the emotional congruency of facial expressions and background scenes. Frontiers in Psychology, 8, 2175. https://doi.org/10.3389/fpsyg.2017.02175 Zinchenko, A., Obermeier, C., Kanske, P., et al. (2017). The influence of negative emotion on cognitive and emotional control remains intact in aging. Frontiers in Aging Neuroscience,9,349. https://doi.org/10.3389/fnagi.2017.00349 Zhunussova, A., Pellicciari, M. C., Carreiras, M., et al. (2024). Electrophysiological correlates of cognitive reserve in healthy older adults at different cognitive states. bioRxiv, 2024- 12 . https://doi.org/10.1101/2024.12.30.627948 Appendix A. Personality characteristics and their rating scores for valence and familiarity ( Mean and SD ), and the corresponding priming sentences used in the study. | Mean | SD | Mean | SD | ||||| | Positive | ehrlich | honest | 6.63 | 0.67 | 6.87 | 0.43 | Exp | Dies ist eine ehrliche Frau. / Dies ist ein ehrlicher Mann. | | humorvoll | humorous | 6.60 | 0.77 | 6.50 | 1.04 | Exp | Dies ist eine humorvolle Frau. / Dies ist ein humorvoller Mann. | | | freundlich | kind | 6.57 | 0.63 | 6.80 | 0.48 | Exp | Dies ist eine freundliche Frau. / Dies ist ein freundlicher Mann. | | | aufrichtig | honest | 6.57 | 0.68 | 6.30 | 1.09 | Exp | Dies ist eine aufrichtige Frau. / Dies ist ein aufrichtiger Mann. | | | zuverlässig | trustworthy | 6.53 | 0.68 | 6.83 | 0.38 | Exp | Dies ist eine zuverlässige Frau. / Dies ist ein zuverlässiger Mann. | | | warmherzig | warm | 6.53 | 0.82 | 6.30 | 1.21 | Exp | Dies ist eine warmherzige Frau. / Dies ist ein warmherziger Mann. | | | hilfsbereit | helpful | 6.50 | 0.73 | 6.83 | 0.46 | Exp | Dies ist eine hilfsbereite Frau. / Dies ist ein hilfsbereiter Mann. | | | großherzig | magnanimous | 6.40 | 0.86 | 6.17 | 1.09 | Exp | Dies ist eine großherzige Frau. / Dies ist ein großherziger Mann. | | | witzig | witty | 6.37 | 0.61 | 6.93 | 0.25 | Exp | Dies ist eine witzige Frau. / Dies ist ein witziger Mann. | | | liebenswert | amiable | 6.30 | 0.99 | 6.43 | 0.90 | Exp | Dies ist eine liebenswerte Frau. / Dies ist ein liebenswerter Mann. | | | gerecht | just | 6.30 | 0.92 | 6.67 | 0.71 | Exp | Dies ist eine gerechte Frau. / Dies ist ein gerechter Mann. | | | großzügig | generous | 6.30 | 0.79 | 6.50 | 0.78 | Exp | Dies ist eine großzügige Frau. / Dies ist ein großzügiger Mann. | | | loyal | faithful | 6.27 | 0.58 | 6.57 | 0.86 | Practice | Dies ist eine loyale Frau. / Dies ist ein loyaler Mann. | | | rücksichtsvoll | considerate | 6.40 | 0.77 | 6.40 | 1.04 | Practice | Dies ist eine rücksichtsvolle Frau. / Dies ist ein rücksichtsvoller Mann. | | | optimistisch | optimistic | 6.27 | 0.87 | 6.43 | 0.94 | Practice | Dies ist eine optimistische Frau. / Dies ist ein optimistischer Mann. | | | klug | smart | 6.23 | 0.82 | 6.67 | 0.66 | Practice | Dies ist eine kluge Frau. / Dies ist ein kluger Mann. | | | negative | gewalttätig | cruel | 1.17 | 0.38 | 6.30 | 1.34 | Exp | Dies ist eine gewalttätige Frau. / Dies ist ein gewalttätiger Mann. | | brutal | brutal | 1.20 | 0.48 | 6.33 | 1.21 | Exp | Dies ist eine brutale Frau. / Dies ist ein brutaler Mann. | | | grausam | merciless | 1.33 | 0.61 | 6.17 | 1.53 | Exp | Dies ist eine grausame Frau. / Dies ist ein grausamer Mann. | | | bösartig | vicious | 1.33 | 0.61 | 6.20 | 1.42 | Exp | Dies ist eine bösartige Frau. / Dies ist ein bösartiger Mann. | | | abscheulich | despicable | 1.37 | 0.72 | 5.73 | 1.74 | Exp | Dies ist eine abscheuliche Frau. / Dies ist ein abscheulicher Mann. | | | widerwärtig | disgusting | 1.50 | 0.68 | 5.70 | 1.62 | Exp | Dies ist eine widerwärtige Frau. / Dies ist ein widerwärtiger Mann. | | | hinterlistig | cunning | 1.53 | 0.63 | 6.03 | 1.45 | Exp | Dies ist eine hinterlistige Frau. / Dies ist ein hinterlistiger Mann. | | | heimtückisch | insidious | 1.53 | 0.73 | 5.63 | 1.47 | Exp | Dies ist eine heimtückische Frau. / Dies ist ein heimtückischer Mann. | | | böse | evil | 1.60 | 0.93 | 6.60 | 1.22 | Exp | Dies ist eine böse Frau. / Dies ist ein böser Mann. | | | verlogen | dishonest | 1.63 | 0.85 | 5.77 | 1.55 | Exp | Dies ist eine verlogene Frau. / Dies ist ein verlogener Mann. | | | ekelhaft | nasty | 1.67 | 1.18 | 6.50 | 0.97 | Exp | Dies ist eine ekelhafte Frau. / Dies ist ein ekelhafter Mann. | | | arrogant | arrogant | 1.87 | 0.73 | 6.33 | 1.30 | Exp | Dies ist eine arrogante Frau. / Dies ist ein arroganter Mann. | | | überheblich | presumptuous | 1.87 | 0.78 | 5.97 | 1.54 | Practice | Dies ist eine überhebliche Frau. / Dies ist ein überheblicher Mann. | | | habsüchtig | greedy | 1.93 | 1.01 | 4.93 | 2.03 | Practice | Dies ist eine habsüchtige Frau. / Dies ist ein habsüchtiger Mann. | | | geizig | mean | 1.93 | 0.74 | 6.40 | 1.19 | Practice | Dies ist eine geizige Frau. / Dies ist ein geiziger Mann. | | | unverschämt | impertinent | 2.00 | 1.08 | 6.10 | 1.18 | Practice | Dies ist eine unverschämte Frau. / Dies ist ein unverschämter Mann. | | | Neutral | geschäftig | busy | 4.00 | 0.87 | 5.23 | 1.45 | Exp | Dies ist eine geschäftige Frau. / Dies ist ein geschäftiger Mann. | | still | quiet | 4.00 | 1.17 | 6.63 | 0.81 | Exp | Dies ist eine stille Frau. / Dies ist ein stiller Mann. | | | beschäftigt | busy | 3.97 | 0.56 | 6.37 | 1.03 | Exp | Dies ist eine beschäftigte Frau. / Dies ist ein beschäftigter Mann. | | | ledig | single | 4.03 | 0.32 | 6.10 | 1.30 | Exp | Dies ist eine ledige Frau. / Dies ist ein lediger Mann. | | | anspruchslos | frugal | 4.07 | 1.36 | 6.03 | 1.19 | Exp | Dies ist eine anspruchslose Frau. / Dies ist ein anspruchsloser Mann. | | | adlig | noble | 4.10 | 1.09 | 5.90 | 1.54 | Exp | Dies ist eine adlige Frau. / Dies ist ein adliger Mann. | | | wunschlos | content | 4.10 | 1.24 | 6.10 | 1.35 | Exp | Dies ist eine wunschlose Frau. / Dies ist ein wunschloser Mann. | | | klein | small | 4.10 | 0.71 | 6.77 | 0.68 | Exp | Dies ist eine kleine Frau. / Dies ist ein kleiner Mann. | | | bürgerlich | civil | 3.87 | 0.63 | 5.60 | 1.63 | Exp | Dies ist eine bürgerliche Frau. / Dies ist ein bürgerlicher Mann. | | | normal | normal | 3.87 | 0.78 | 6.60 | 0.97 | Exp | Dies ist eine normale Frau. / Dies ist ein normaler Mann. | | | verheiratet | married | 4.33 | 0.84 | 6.57 | 0.94 | Exp | Dies ist eine verheiratet Frau. / Dies ist ein verheiratet Mann. | | | ernst | serious | 3.87 | 0.94 | 6.60 | 0.67 | Exp | Dies ist eine ernste Frau. / Dies ist ein ernster Mann. | | | dünn | thin | 4.17 | 0.38 | 6.50 | 0.90 | Practice | Dies ist eine dünne Frau. / Dies ist ein dünner Mann. | | | mittelalter | middle aged | 3.83 | 0.46 | 4.90 | 1.63 | Practice | Dies ist eine mittelalte Frau. / Dies ist ein mittelalter Mann. | | | groß | big | 4.20 | 0.66 | 6.63 | 0.85 | Practice | Dies ist eine große Frau. / Dies ist ein großer Mann. | | | volljährig | of legal age | 4.23 | 0.63 | 6.57 | 0.86 | Practice | Dies ist eine volljährige Frau. / Dies ist ein volljähriger Mann. | Information & Authors Information Version history Copyright This work is licensed under a Non Exclusive No Reuse License. Authors Metrics & Citations Metrics Article Usage 405views 116downloads Citations Download citation Hongwei Xing, Xiaoli Ma, Yaping Yang, et al. Nice people are easy to smile with: Deliberate imitation of facial expressions depends on personal characteristics. Authorea. 29 December 2025. DOI: https://doi.org/10.22541/au.176700616.67379175/v1 DOI: https://doi.org/10.22541/au.176700616.67379175/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu.