https://orcid.org/0000-0002-3106-3261
Abstract
Expressing appreciation is essential for establishing interpersonal closeness, but virtual interactions are increasingly common and create social distance. Little is known about the neural and inter-brain correlates of expressing appreciation and the potential effects of virtual videoconferencing on this kind of interaction. Here, we assess inter-brain coherence with functional near-infrared spectroscopy while dyads expressed appreciation to one another. We scanned 36 dyads (72 participants) who interacted in either an in-person or virtual (Zoom®) condition. Participants reported on their subjective experience of interpersonal closeness. As predicted, expressing appreciation increased interpersonal closeness between dyad partners. Relative to 3 other cooperation tasks (i.e. problem-solving task, creative-innovation task, socio-emotional task), we observed increased inter-brain coherence in socio-cognitive areas of the cortex (anterior frontopolar area, inferior frontal gyrus, premotor cortex, middle temporal gyrus, supramarginal gyrus, and visual association cortex) during the appreciation task. Increased inter-brain coherence in socio-cognitive areas during the appreciation task was associated with increased interpersonal closeness. These findings support the perspective that expressing appreciation, both in-person and virtually, increases subjective and neural metrics of interpersonal closeness.
Introduction
Mutual appreciation (we follow Adler and Fagley (2005)’s definition of appreciation as “acknowledging the value and meaning of something—an event, a person, a behavior, an object—and feeling a positive emotional connection to it” (p. 81).) is vital for developing and maintaining healthy social interactions and relationships. Studies have shown that shared appreciation is associated with increased levels of interpersonal closeness and satisfaction in romantic relationships (Algoe et al. 2010; Gordon et al. 2012) and friendships (Badri et al. 2022), and increased well-being and performance at the workplace (Pryce 2011; Fagley and Adler 2012). Relatedly, expressing appreciation increases interpersonal closeness between romantic partners (Schramm et al. 2005; Lambert et al. 2010; Algoe et al. 2013), colleagues (Rusk et al. 2015), and strangers (Williams and Bartlett 2015; Kong and Belkin 2019). Social scientists have developed several theories regarding how expressing appreciation can contribute to new relationships and orientation to, maintenance of, and investment in existing relationships (e.g. Find-remind-and-bind Theory (Algoe 2012); Reciprocity Rules (Cropanzano and Mitchell 2005); and, Self-determination Theory (Ryan and Deci 2000)). A central idea shared across these theories is that feeling appreciation and gratitude prompts behaviors that are desirable in a future relationship partner (e.g. increased prosociality, decreased aggression, and increased relationship investment; Bartlett et al. 2012; Bartlett and DeSteno 2006; Dewall et al. 2012; Williams and Bartlett 2015). Prior studies have focused on appreciation exchanged in-person in lab settings (Williams and Bartlett 2015; Kong and Belkin 2019), but many interactions today are moving to virtual spaces. Thus, the role of appreciation in facilitating interpersonal closeness in an era of relative social distance associated with virtual environments is unclear. Further, little is known about the single-brain and inter-brain processes that are related to sharing of appreciation. Here, we used functional near-infrared spectroscopy (fNIRS) to investigate the neural correlates of expressing appreciation and experiencing interpersonal closeness. In addition, we assessed whether virtual versus in-person interactions affect neural processes and interpersonal closeness.
Functional NIRS is a noninvasive brain imaging technique that measures changes in cortical oxygenation associated with neural activation (Strangman et al. 2002; Cui et al. 2010). Some of the advantages of fNIRS over other neuroimaging technologies are its relative tolerance to movement and increased portability, enabling the assessment of neural activation and behavior under naturalistic conditions (Quaresima and Ferrari 2019). Researchers are increasingly using fNIRS to extend neuroimaging from assessing single-brain functioning to multi-brain “hyperscanning” experiments (Funane et al. 2011; Cui et al. 2012). Researchers have focused on studying inter-brain coherence (IBC; i.e. correlation of cortical activation between brains) and how IBC relates to behavioral measures of interaction (Babiloni and Astolfi 2014; Balters et al. 2020; Czeszumski et al. 2020). Studies have found increased IBC when pairs completed a task together in contrast to completing the identical task (simultaneously) alone (Feng et al. 2020; Fishburn et al. 2018; Hu et al. 2017; Liu et al. 2016; M. Zhang et al. 2021; Zhou et al. 2022). In addition to joint attention, research shows that joint goal-directed intention is related to IBC (Kruse et al. 2021), and dyads demonstrate increased IBC when cooperating in contrast to when competing with one another (Cui et al. 2012; Lu et al. 2019a). Other studies have found increased IBC, particularly in prefrontal regions, to be positively associated with prosocial behavior during collaborative activities such as games (Liu et al. 2016; Pan et al. 2017; Miller et al. 2019), group creative brainstorming (Xue et al. 2018; Lu et al. 2019b), and psychological counseling (Zhang et al. 2018). Functional NIRS hyperscanning research by Balconi and Fronda (2020) found that gift exchange (as a sign of appreciation) increases IBC in the dorsolateral prefrontal cortex (dlPFC), a region associated with social interaction processes, such as emotional attunement (Balconi et al. 2017), theory of mind, and suppression of selfish behaviors (Kalbe et al. 2010; Balconi and Vanutelli 2017). Recently, we demonstrated differences in dyadic IBC between in-person and virtual (Zoom®) interactions (Balters et al. 2023). Compared with in-person interaction, virtual interaction was characterized by increased IBC during problem-solving and creative-innovation tasks but decreased IBC during socio-emotional collaboration. These findings suggest task-dependent effects of virtual interaction on IBC.
Expressing appreciation is widely regarded as an important behavior for facilitating interpersonal closeness (Rusk et al. 2015; Williams and Bartlett 2015; Kong and Belkin 2019). To date, however, social neuroscience research has not considered the neural processes involved in expressing appreciation, which might support or relate to the experience of increased interpersonal closeness in this context. Further, with the rapid growth and increasingly widespread use of video conferencing technology, it is important to explore the potential effects of this technology on neural processes and interpersonal closeness during appreciation. This study was designed to address these gaps. Specifically, we utilized fNIRS hyperscanning to understand single-brain and inter-brain signatures of expressing appreciation. We also assessed whether these signatures differ between in-person and virtual interactions. Using a portable fNIRS neuroimaging system, we measured cortical activation from dyads who interacted in either an in-person or virtual condition (Fig. 1). We also measured participants’ subjective experiences of interpersonal closeness during the experiment. Dyads first performed 3 cooperative paradigms, including a problem-solving task, a creative-innovation task, and a socio-emotional task. After completing these 3 tasks, participants expressed appreciation for one another for 2 min. We hypothesized that relative to the 3 other tasks, expressing appreciation would be characterized by increased interpersonal closeness between dyad partners accompanied by specific patterns of individual neural activity and IBC. Based on prior findings of fNIRS hyperscanning studies focused on prosocial interaction, we further predicted that increased IBC would be observed in the dlPFC during the appreciation task compared with the control tasks. In testing these hypotheses, data analyses focused on: (1) exploring single-brain and inter-brain correlates of expressing appreciation and (2) assessing whether appreciation-related neural activation and IBC differed between in-person and virtual interactions. Taken together, this study has potential implications for improving our understanding of: (a) the neural bases of expressing appreciation, (b) the neural processes that support increased interpersonal closeness during appreciation, and (c) whether modern video conferencing technology impacts neural and subjective metrics of the quality of social interaction during appreciation.
Conditions of the between-subject study. Participants interacted either in an in-person or virtual interaction condition throughout the experiment.
Materials and methods
The study methodology was approved by the Stanford University Institutional Review Board (IRB #18160) and followed COVID-19 regulation for human experimentation as defined by the Stanford University School of Medicine. Written consent was obtained from all participants.
Participants
As described in Balters et al. (2023), participants were recruited through email lists and social media. A total of 72 adults participated in the study (n = 36 female, n = 36 male, mean age: 27.11 years, SD = 7.57 years). The racial and ethnic composition of the sample was 3% African American/Black, 23% Asian/Pacific Islander, 6% Biracial/Multiracial, 15% Hispanic/Latinx, 6% Middle Eastern, 22% South Asians, and 25% Caucasian/White. All participants were healthy, right-handed, and had normal or corrected to normal hearing and vision. The study followed a between-subject design, and participants interacted with their dyad partner either in-person or virtually throughout the experiment. The previously unacquainted dyad partners were randomly assigned to either interaction condition. Condition groups were matched based on age, sex, and race/ethnicity. Both interaction conditions contained 6 female–female, 6 female–male, and 6 male–male dyads. The experimental procedure lasted approximately 3 h, and participants were compensated with an Amazon gift card ($25 USD per hour).
Experimental procedure
As described in Balters et al. (2023), dyads in the in-person condition sat face-to-face at a table (Fig. 1). Dyad partners sat 9 ft apart and wore facemasks in accordance with COVID-19 guidelines. To decrease obstruction of faces, participants wore transparent, anti-fog facemasks (ClearMask™). Participants of the virtual condition also wore a facemask to prevent bias between the conditions. Dyads in the virtual condition sat at desks in 2 separate rooms and interacted over Zoom video conferencing. Given previous research suggesting that the self-view window can heighten self-focused attention during video conferencing (Bailenson, 2021), we maximized the Zoom window in the virtual condition, and no self-view window was displayed. Two identical laptops were used for video conferencing (Lenovo Yoga 730–730-15IKB, 15.6″). Laptops were placed to approximate the facial proportions in the in-person condition.
Functional NIRS caps were attached while participants were in 2 separate rooms. Functional NIRS calibration was executed while participants were in the same room or after the Zoom® connection was already established. Before the experiment, participants performed a silent baseline task for 30 s with closed eyes. Participants were instructed not to talk to one another during that time. Subsequently, participants had 3 min to introduce themselves to one another (Fig. 2). During the experiment, participants were alone in the room(s), and instructions were given through audio prompts. After the experiment, participants were guided to 2 separate rooms and filled out post-experimental questionnaires.
Experimental procedure. Dyads experienced the 3 familiarization and appreciation tasks in either an in-person or virtual interaction condition throughout the experiment. The sample consisted of 36 dyads, 6 dyads per randomized task order across both conditions.
Experimental tasks
Before executing the appreciation task, dyads collaborated on 3 cooperative tasks: a problem-solving scenario, a creative ideation scenario, and an emotion-sharing scenario. These tasks served to familiarize each participant with their dyad partner prior to the appreciation task. Specifically, these 3 tasks provided social experiences that could inform participants’ expression of appreciation toward their partner. Considering diverse collaborative activities allowed participants to draw from a variety of social interaction contexts to express appreciation. In addition, considering these activities allowed us to assess whether there were general or unique effects of expressing appreciation on behavioral and brain outcomes in in-person and virtual interactions. All 3 tasks were 8 min long, and a clock was provided for the participants to keep track of time. As described in Balters et al. (2023), in the problem-solving task, participants were instructed to collaborate and identify 4 traffic rules that have the most significant impact on safety on US highways and to provide reasons for why the traffic rule is essential during the discussion. In the creative-innovation task, dyads were instructed to collaborate and design a solution (a product, service, process, campaign, etc.) to increase water conservation in CA households. In the emotion-sharing task, participants were instructed to collaborate and engage in a modified version of a Nonviolent Communication (NVC) exercise that is used to increase interpersonal closeness between individuals (Rosenberg and Chopra 2015). The order of the 3 familiarization tasks was randomized across dyads, and a 2-min soothing beach video was displayed between the tasks to reduce carry-over effects. After these 3 tasks, participants engaged in the appreciation task. Participants were instructed to share with their partners what they appreciated about each other for a duration of 2 min (“For the next 2 min, please share with your partner what you appreciate about them”). Given the prior familiarization procedure, participants could express appreciation based on a variety of different collaborative contexts. The 2-min duration was selected to facilitate natural expressions of appreciation, which are often of short duration (Williams and Bartlett 2015; Kong and Belkin 2019). During the familiarization tasks, participants did not receive explicit instruction to provide or avoid the expression of appreciation. Dyads collaborated on each task without interruptions. After each of the familiarization tasks and the appreciation task, participants filled out a questionnaire that captured their subjective experience of interpersonal closeness. For each outcome of interest, the average of the 3 familiarization tasks was used as a control against the appreciation task (e.g. IBC during the appreciation task versus average IBC during the problem-solving, creative-innovation, and social–emotional tasks).
Interpersonal closeness metric
After each task (i.e. problem-solving, creative-innovation, socio-emotional, appreciation), participants rated their subjective closeness toward their dyad partner on a 7-point connectedness scale (i.e. “How connected did you feel with your partner during the task?”; [Wiltermuth and Heath 2009]) ranging from “not at all” to “extremely.” This scale has been widely used to assess interpersonal closeness between dyad partners (e.g. Cirelli et al. 2014; Páez et al. 2015).
Post experimental assessments
Personality traits
Participants completed a battery of personality measures, including the NEO Five-Factor Inventory-3 (NEO-FFI-3) (McCrae and Costa Jr 2007), the Adult Attachment Scale Survey (Collins and Read 1990), and Wong and Law’s Emotional Intelligence Survey (Wong and Law 2017). For the NEO-FFI-3, T-scores were calculated for the Neuroticism, Extraversion, Openness, Agreeableness, and Conscientiousness subscales. For the Adult Attachment Scale, Avoidant and Anxious subscales were calculated following the procedures described in Collins (1996). For the Emotional Intelligence score, we calculated the average scores on the self-emotional appraisal, other’s emotion appraisal, use of emotion, and regulation of emotion subscales.
Familiarity with Zoom® video conferencing
Participants rated their prior experience and proficiency with Zoom® video conferencing on a 5-point Likert scale ranging from “not at all” to “extremely.”
Functional NIRS data acquisition and preprocessing
Cortical hemodynamic activation was captured via a continuous wave fNIRS system (NIRSport2 System, NIRX, Germany, 64 x 64 configuration). System parameters were set to record with 2 wavelengths (760 and 850 mm) at a sampling frequency of 10.2 Hz. A total of 128 optodes (32 sources x 32 detectors per person) were divided between the 2 dyad partners to form 100 measurement channels per participant. The channels were placed over the entire cortex according to the international 10–20 EEG placement system (Fig. 3a). The system also captured gyroscope data from 10–20 position “Cz” at 10.2 Hz.
Raw fNIRS data were analyzed in Matlab version R2021a (MathWorks, Inc.) using NIRS Brain AnalyzIR Toolbox (Santosa et al. 2018). The scalp coupling index (SCI) was calculated to assess the data quality of fNIRS scans (Pollonini et al. 2016). Results showed that 74.13% of the data had an SCI of more than 0.8. Results from our previous publication showed no statistical differences in head motion (i.e. 3-dimensional acceleration and rotation) or SCIs between the 2 groups (Balters et al. 2023). The in-person and virtual conditions did not differ regarding fNIRS data quality. Channels with extensive noise (i.e. SCI ≤ 0.8) were excluded from further analyses. The remaining channels were converted from raw data to optical density data. A fourth-order Butterworth band-pass filter (0.02–0.2 Hz) was applied to eliminate artifacts such as cardiac interference. Motion artifacts were subsequently corrected using a wavelet motion correction procedure (Molavi and Dumont 2012). Data were transformed into concentration changes of oxygenated hemoglobin (HbO) and deoxygenated hemoglobin (HbR) according to the Modified Beer–Lambert Law (Wyatt et al. 1986). The resulting data were converted to z-scores since HbO and HbR data are relative values. As shown in Fig. 3b, the 100 measurement channels were clustered into 32 regions of interest (ROIs) via source localization (Huppert et al. 2017; Tremblay et al. 2018). Specifically, all channels with a common fNIRS source were averaged to derive ROI data. Averaged Montreal Neurological Institute (MNI) coordinates of each ROI were used to project all ROIs onto the cortical surface using an automatic anatomical labeling method (Lancaster et al. 2000; Singh et al. 2005; see Supplementary Table 1). For single-brain analyses, the average HbO and HbR across the first 2-min duration of each familiarization task and the entire duration of the 2-min appreciation task was calculated. Only ROIs that contained more than 75% of the data across participants were included for statistical single-brain analyses. As a result, the following 6 ROIs were excluded: ROI 8—left supplementary motor area, ROI 17—right somatosensory association cortex, ROI 21—visual association cortex (V3), ROI 24—right visual association cortex, ROI 28—left somatosensory association cortex, ROI 32—left visual association cortex. We focused our analyses on HbO, given that this signal is more robust and sensitive to task-associated changes compared with HbR (Ferrari and Quaresima 2012; Plichta et al. 2006; Balters et al. 2020; Yang et al. 2020; Zhou et al. 2022). We present HbR analyses in the Supplementary Material Section 3.1.
Positions of fNIRS measurement channels and ROIs. (a) A total of 100 measurement channels were clustered into (b) 32 ROIs via source localization (Huppert et al. 2017; Tremblay et al. 2018).
IBC analysis
IBC (i.e. the similarity between HbO signals of dyad partners) was assessed via wavelet transform coherence (Cui et al. 2012) using the wavelet coherence function (“wcoherence”) in Matlab 2021a (MathWorks, Inc.). Only the 26 ROIs eligible for the single-brain analyses were considered for the IBC analysis. Wavelet coherence values between HbO signals of each ROI and the rest of the ROIs were calculated, leading to a total of 676 combinations (26 ROIs x 26 ROIs). The IBC values between the same ROI pairs were averaged, resulting in 351 ROI pairs. The average IBC values between 0.2 and 0.02 Hz were calculated. Data above 0.2 Hz were not considered to avoid systemic noise related to heart rate (0.6–2 Hz) and respiration (0.2–0.3 Hz), and data below 0.02 Hz were excluded to remove low-frequency fluctuations (Molavi and Dumont 2012). Averaged IBC values were converted to Fisher z-statistic. Our inclusion criteria that ROI pairs with a minimum of 75% of data entries across participants were considered for further statistical analyses. This processing step reduced the eligible ROI pairs to a total of 102. Similar to the single brain analyses, we focus on IBC calculated from HbO data in this paper, and present IBC analyses based on HbR in the Supplementary Material Section 3.2.
Results
The overall objective of this study was to assess whether there were task-dependent (i.e. appreciation versus average of the 3 familiarization tasks) or condition-dependent (i.e. in-person versus virtual interaction) main or interaction effects on (1) the interpersonal closeness measures (Section 3.2), (2) single-brain activation (Section 3.3), and (3) IBC (Section 3.4). Given that the appreciation task was the last task of the experiment, we conducted subsequent time-effect analyses for each measure (Section 3.5).
Conditions were matched on individual difference variables
Results from our previous publication (Balters et al. 2023) demonstrated that there were no statistical differences in age, inter-dyad age differences, personality traits (i.e. NEO-FFI-3 T scores, Adult Attachment style, emotion intelligence), creative ability (i.e. AUT fluency, AUT originality), and familiarity with Zoom® video conferencing (i.e. experience and proficiency) between the 2 conditions (i.e. virtual versus in-person interaction). Results are presented in Supplementary Table 2. Thus, the in-person and virtual conditions were matched on age, inter-dyad age differences, personality traits (i.e. NEO-FFI-3 T scores, Adult Attachment style, emotional intelligence), creative ability (i.e. Alternate Uses Task AUT fluency, AUT originality), and familiarity with Zoom® video conferencing (i.e. experience and proficiency).
Main effect of task on interpersonal closeness
A 2-way analysis of variance (ANOVA) was run to assess whether there were main effects of task (i.e. problem-solving, creative-innovation, socio-emotional, and appreciation tasks), main effects of condition (i.e. in-person and virtual), or interaction effects of task and condition on the interpersonal closeness metric (i.e. connectedness). Two outliers were identified and removed from the analysis (i.e. both outliers were lower than 3 standard deviations from the mean across both tasks). A main effect of task for the connectedness metric was observed (F(3, 66) = 10.958, P < 0.001, Wilks’ Λ = 0.668, partial η2 = 0.332, Fig. 4). Post hoc analyses with FDR-correction for multiple comparisons showed that the appreciation task was characterized by increased levels of connectedness compared with the problem-solving task (mean difference = 0.515 [95% CI, 0.312 to 0.791], adjusted P < 0.001) and the creative-innovation task (mean difference = 0.626 [95% CI, 0.374 to 0.878], adjusted P < 0.001). The level of connectedness was also higher during the appreciation task than the socio-emotion task (mean difference = 0.140), but this difference was not significant after FDR-correction (adjusted P = 0.150). The main effect of condition, and the interaction effect of task and condition, on connectedness were not statistically significant (P > 0.568).
Results for the interpersonal closeness metric. The level of connectedness was higher during the appreciation task relative to the problem-solving and creative-innovation tasks, independent of the interaction condition. The statistically significant difference at FDR-corrected P < 0.05 is indicated by *. Standard error of the means are represented by error bars.
No differences in single-brain cortical activation
Our next aim was to assess whether there were main or interaction effects of task and condition on single-brain activation. We conducted a series of 2-way ANOVAs to assess whether there were main effects of task (i.e. problem-solving, creative-innovation, socio-emotional, and appreciation task), main effects of condition (i.e. in-person and virtual), or interaction effects of task and condition on HbO cortical activation in any of the 26 ROIs. Resulting P values were adjusted using FDR-correction for multiple testing (i.e. 26 tests). Results showed no significant main effects of task, main effects of condition, or interaction effects of task and condition.
Main effects of task on IBC
The third aim was to assess whether task and condition had main or interaction effects on IBC (HbO). Similar to the single-brain analyses, we ran 2-way repeated ANOVAs for each ROI pair to examine the effect of task and condition on IBC. Resulting P values were FDR-corrected for multiple testing (i.e. 102 tests). We observed a statistically significant main effect of task on 5 ROI pairs, including left inferior frontal gyrus-left lateral frontopolar area ROI pair (“left IFG-left lFPA”, F(3, 31) = 6.660, adjusted P = 0.050, Wilks’ Λ = 0.608, partial η2 = 0.392)), the left dlPFC-left premotor cortex ROI pair (“left dlPFC-left pMC”, F(3, 27) = 6.027, adjusted P = 0.050, Wilks’ Λ = 0.599, partial η2 = 0.401), the right dorsolateral prefrontal-left premotor cortex ROI pair (“right dlPFC-left pMC”, F(3, 27) = 7.724, adjusted P = 0.034, Wilks’ Λ = 0.538, partial η2 = 0.462), the left premotor cortex-right visual association cortex ROI pair (“left pMC-right VAC”, F(3, 24) = 6.971, adjusted P = 0.050, Wilks’ Λ = 0.534, partial η2 = 0.466), and the left premotor cortex-right premotor cortex ROI pair (“left pMC-right pMC”, F(3, 24) = 8.363, adjusted P < 0.001, Wilks’ Λ = 0.518, partial η2 = 0.482).
Results of the inter-brain analyses. The finding showed significant main effects of task for 5 ROI pairs (dark red lines) with higher IBC in the appreciation task compared with the problem-solving (a), creative-innovation (b), and socio-emotion task (c). A total of 31 additional ROI pairs also showed significant main effects of task (faded lines) that did not pass FDR-correction. All 31 ROI pairs had higher IBC in the appreciation task compared with the other 3 familiarization tasks. (d) Four ROI pairs showed positive correlations between IBC values and the interpersonal closeness metric during the appreciation task. However, only the right IFG-left pMC ROI pair reached significance after FDR-correction. Abbreviations: premotor cortex (pMC), adjusted P value (adj.)
Post hoc analyses with FDR-correction for multiple comparisons showed overall increased IBC for the appreciation task compared with the other 3 familiarization tasks. For the left IFG-left lFPA ROI pair, IBC was higher in the appreciation task compared with the problem-solving task (mean difference = 0.049 (95% CI, 0.027 to 0.071), adjusted P < 0.001) and the creative-innovation task (mean difference = 0.034 (95% CI, 0.004 to 0.064), adjusted P = 0.040), but not the socio-emotional task (adjusted P = 0.119). Findings for the left dlPFC-left pMC ROI pair showed higher IBC in the appreciation task compared with the creative-innovation task (mean difference = 0.038 (95% CI, 0.020 to 0.055), adjusted P < 0.001), but did not reach significance after FDR-correction for the problem- solving task and socio-emotional task (adjusted P < 0.07). Similarly, IBC in the right dlPFC-left pMC ROI pair was higher in the appreciation task compared with the creative-innovation task (mean difference = 0.042 (95% CI, 0.024 to 0.060), adjusted P < 0.001), but not the other 2 tasks (adjusted P < 0.110). For the left pMC-right VAC ROI pair, IBC was higher in the appreciation task compared with the problem-solving task (mean difference = 0.037 (95% CI, 0.011 to 0.062), adjusted P = 0.001) and the creative-innovation task (mean difference = 0.053 (95% CI, 0.028 to 0.078), adjusted P < 0.001) but not the socio-emotional task (adjusted P = 0.496). Lastly, results for the left pMC-right pMC ROI pair showed increased IBC for the appreciation task compared with the problem-solving task (mean difference = 0.033 (95% CI, 0.005 to 0.061), adjusted P = 0.023), the creative-innovation task (mean difference = 0.056 (95% CI, 0.030 to 0.083), adjusted P < 0.001), and the socio-emotional task (mean difference = 0.050 (95% CI, 0.027 to 0.074), adjusted P < 0.001). Notably, 31 other ROI pairs showed significant main effects of task, but these results were not significant after FDR-correction for multiple testing (0.004 < uncorrected P value < 0.045). In all 31 ROI pairs, IBC values were higher in the appreciation task compared with the 3 familiarization tasks. All F-statistics of the 31 ROI pairs are summarized in Supplementary Table 3. The 31 ROI pairs comprised of 9 prefrontal-prefrontal ROI pairs, 10 prefrontal-premotor ROI pairs, 5 prefrontal-motor/temporal/occipital ROI pairs, 5 premotor-premotor/motor/temporal/occipital ROI pairs, 1 temporal–temporal ROI pair, and 1 occipital-occipital ROI pair. There were no main effects of condition or interaction effects of task and condition for any ROI pair. The findings of all 36 ROI pairs are presented in Fig. 5a–c.
Lastly, we tested correlations between the 36 ROI pairs and interpersonal closeness within the appreciation task. After adjusting P values for multiple comparisons (i.e. 36 correlations tested) using FDR correction (P < 0.05), we observed statistically significant positive correlations between interpersonal closeness and the right IFG-left pMC ROI pair (r = 0.42, adjusted P = 0.033; Fig. 5d). Interpersonal closeness was also positively correlated with 3 other ROI pairs. However, these associations were no longer significant after FDR correction: left IFG-right MTG ROI pair (r = 0.33, uncorrected P = 0.007), left dlPFC-right MTG (r = 0.26, uncorrected P = 0.039), and left pMC-left pMC ROI pair (r = 0.26, uncorrected P = 0.047).
Time effect analyses
We conducted time-effect analyses for the interpersonal closeness metric and IBC to test whether the findings could be due to the appreciation task being the last task of the experiment. These analyses and results are described in detail in Supplementary Table 4 and Supplementary Table 5 and are summarized in Fig. 6. Briefly, we did not observe time-related changes in interpersonal closeness, IBC, or cortical activation, across the tasks. The findings suggest that the increased level of connectedness and increased levels of IBC were not related to the timing of the appreciation task relative to the other tasks.
Results of the time analyses. (a) the interpersonal closeness metric (i.e. connectedness) was higher in the appreciation task compared with all 3 preceding tasks. (b) the trajectories of IBC (HbO) throughout the experiment (i.e. 2-min intervals with the appreciation task being last). The 5 ROI pairs that showed significant results in the main analysis are presented along with the additional 31 ROI pairs that did not pass FDR-correction (lower right corner). Results showed increased IBC in the appreciation task compared with the 12 preceding time points in all 5 ROI pairs and the additional 31 ROI pairs. The statistically significant difference at FDR-corrected P < 0.05 is indicated by *, and uncorrected P < 0.05 is indicated by (*). Note: For ease of reading, we did not include (*) in (b). Standard errors of the means are represented by error bars. Abbreviations: premotor cortex (pMC), visual association cortex (VAC).
Discussion
Expressing appreciation is essential for positive social relationships and has consistently been shown to facilitate interpersonal closeness (Schramm et al. 2005; Lambert et al. 2010; Algoe et al. 2013; Rusk et al. 2015; Williams and Bartlett 2015; Kong and Belkin 2019). However, as video conferencing technology becomes more prevalent, many social interactions are moving to virtual platforms. It is unclear whether expressing appreciation is an effective tool for eliciting interpersonal closeness in virtual interactions. Further, prior studies have not considered the neural correlates of appreciation at the individual or dyadic levels. This study explored the single-brain and inter-brain signatures of expressing appreciation and assessed whether these signatures varied between in-person and virtual interaction conditions. We found that expressing appreciation was related to increased subjective interpersonal closeness (i.e. connectedness) compared with other joint, cooperative tasks such as problem-solving and creative-innovation tasks. In addition, the appreciation task was characterized by increased IBC in brain areas implicated in socio-cognitive processing. Increased IBC in ROI pairs that included socio-cognitive regions was associated with greater interpersonal closeness. Our findings (1) extend prior research on expressing appreciation and interpersonal closeness to virtual interactions and (2) identify inter-brain correlates of expressing appreciation within dyads.
Across conditions, we found that the appreciation task increased interpersonal closeness as assessed by Wiltermuth and Heath’ (2009) connectedness scale compared with other joint, collaborative tasks such as problem-solving and creative brainstorming. This was the case for in person and virtual settings. These findings align with previous studies that showed increased levels of interpersonal closeness associated with expressing appreciation in previously unacquainted persons (Williams and Bartlett 2015; Kong and Belkin 2019). To our knowledge, this study is the first to find evidence that expressing appreciation during virtual interactions can increase interpersonal closeness. As hypothesized, expressing appreciation increased interpersonal closeness between dyad partners, both in person and over Zoom®.
Results further showed increased IBC in about 35% of the ROI pairs in the appreciation task compared with the 3 familiarization tasks in both interaction conditions. Increased IBC spanned multiple functional regions, primarily consisting of connections involving prefrontal ROIs with all 6 functional areas of the human cortex (i.e. prefrontal cortex, premotor cortex, primary motor cortex, temporal cortex, parietal cortex, and occipital cortex). These findings suggest widespread dyadic coupling of multiple socio-cognitive brain regions during appreciation. Concerning our second hypothesis that predicted increased IBC in the dlPFC in association with expressing appreciation, we found increased IBC in the dlPFC during appreciation compared with the other familiarization tasks. Our findings are consistent with those from prior fNIRS hyperscanning research that found increased IBC in ROI pairs that involve the dlPFC after gift exchange (as a sign of appreciation; Balconi and Fronda 2020). We found increased IBC between the dlPFC and other socio-cognitive regions such as the aFPA, IFG, pMC, MTG, smG, and VAC; these regions are implicated in mentalizing processes (e.g. Theory of Mind), attention, and perception (Tai et al. 2004; Iacoboni and Dapretto 2006; Hung et al. 2011; Kaller et al. 2011; Brunoni and Vanderhasselt 2014; Gomez-Ramirez et al. 2016; Rohe and Noppeney 2016). Notably, we found that IBC in ROI pairs containing these socio-cognitive areas (i.e. bilateral IFG, bilateral pMC, and the right MTG) to be positively associated with interpersonal closeness during the appreciation task. These findings align with prior studies that have found increased IBC, particularly in socio-cognitive ROI pairs, to be positively associated with prosocial behaviors (Cui et al. 2012; Liu et al. 2016; Pan et al. 2017; Dai et al. 2018; Xue et al. 2018; Miller et al. 2019; Lu et al. 2019a; Balters et al. 2023).
We observed increased IBC in the IFG, MTG, and pMC during appreciation compared with the other cooperative tasks. These regions are posited to be part of the extended mirror neuron system (MNS) in humans (Iacoboni and Dapretto 2006; Cattaneo and Rizzolatti 2009). The MNS is vital for social processes, including social cognition, empathy, language, and theory of mind (Schulte-Rüther et al. 2007; Oh et al. 2019). While the MTG (and superior temporal sulcus) is the primary visual input to the MNS, the pMC is thought to be involved in understanding motor actions performed by others (Allison et al. 2000; Iacoboni and Dapretto 2006). Together with the rostral portion of the inferior parietal lobule, the MTG and pMC form the main circuit of the MNS and facilitate understanding other’s actions and, potentially, intentions (Iacoboni 2005; Cattaneo and Rizzolatti 2009). Thus, one interpretation of our findings is that relative to other collaborative contexts (i.e. problem-solving, creative brainstorming, and emotion sharing), expressing appreciation involves increased coordination across interaction partners in neural processes implicated in representing motor actions in self and others. In addition, our findings suggest that increased dyadic coordination in the activity of these regions is accompanied by increased feelings of interpersonal closeness. Researchers have theorized that in the absence of prior experience with another person, an expression of gratitude can help to increase closeness between individuals by providing an important signal about future social affiliation (Williams and Bartlett 2015). Our findings provide support for a dyad-level neural mechanism by which expressing appreciation facilitates interpersonal closeness. However, whether increased appreciation-related IBC in socio-cognitive areas causes or is a consequence of interpersonal closeness is unclear from our findings and requires future research.
We note 3 limitations of this study. First, participants wore clear, anti-fog facemasks due to COVID-19 regulations in both conditions, which might have obstructed some facial features and impacted the results. Future research should replicate our findings without using face masks. Second, given our sample size, the current findings should be considered preliminary. As a related point, the number of dyads in our sample did not allow us to assess whether findings varied as a function of dyad composition (e.g. sex composition of the dyad) or individual characteristics (e.g. personality) (Baker et al. 2016; Miller et al. 2019; Kruse et al. 2021). Future studies should consider dyad- and individual-level characteristics as potential covariates and moderators. Lastly, the appreciation task was always last in the experiment. It is important to note, however, that the control tasks were counterbalanced, and we did not observe effects of time on interpersonal closeness and IBC throughout the experiment. Thus, our findings did not appear to result from increasing interpersonal closeness and IBC over the duration of the experiment. Our findings are consistent with social psychology research suggesting that relationship-building tasks build closeness, whereas structured joint tasks throughout an experiment do not (Aron et al. 1997). In addition, subjective assessments and IBC measured in 2-min epochs suggest that a 2-min duration of the appreciation task was sufficient in increasing interpersonal closeness and IBC. Nonetheless, future studies with longer task durations will be necessary for capturing dynamic aspects of IBC during appreciation (Li et al. 2021).
In conclusion, this study provides evidence that expressing appreciation positively affects interpersonal closeness and IBC between relative strangers in both in-person and virtual settings. Specifically, the appreciation task was associated with increased interpersonal closeness and enhanced IBC in socio-cognitive areas of the cortex. One implication of these findings is that expressing appreciation for as short as 2 min can serve as an effective intervention for establishing increased interpersonal closeness between 2 individuals, both in person and over Zoom®. There are potential applications for this intervention. For example, eliciting states of interpersonal closeness may have positive effects in collaborative contexts. Indeed, prior research suggest that interpersonal closeness increases team creativity and team performance (for review, please see De Jong et al. 2016). Future studies could assess whether increased interpersonal closeness during appreciation serves as a mediator for improved team performance during in-person and virtual interactions.
CRediT for author contributions
SB: Conceptualization, data curation, formal analysis, methodology, visualization, writing – original draft, writing-review & editing JGM: Methodology, Writing – Original Draft, Writing – review & editing ALR: Conceptualization, Funding acquisition, Supervision, Writing– review & editing.
Funding
This work was supported by a Hasso Plattner Design Thinking Research Program Grant and a gift from the Kelvin Foundation.
Conflict of interest statement: All authors declare no conflict of interest.
Data availability statement
Data available on request due to privacy/ethical restrictions.
