Behavioral Interventions in Face-to-face Communication to Decrease Group Stress Levels

Behavioral Interventions in Face-to-face Communication to Decrease Group Stress Levels | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Behavioral Interventions in Face-to-face Communication to Decrease Group Stress Levels Satomi Tsuji, Nobuo Sato, Kazuo Yano, Yoshihiro Miyake This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4240426/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Reducing workplace stress has become a societal challenge. Traditionally, the measurement of stress and intervention methods for it have focused on individuals. In our prior research, by contrast, we conceptualized stress as a collective phenomenon arising from group dynamics and proposed a method to estimate group stress levels quantitatively. This method involves equipping individuals within a group with accelerometers and calculating the scale exponent of the activity duration distribution of their physical movements. However, the potential for reducing group stress levels remained unverified. In this study, we hypothesized that behavioral change could effectively reduce group stress levels and intervened in workplace behaviors by offering advice through a smartphone app. The interventions targeted five categories of behaviors: (I) conversational partners, (II) methods of conversation, (III) desk work practices, (IV) arrival and departure times, and (V) time management. The results showed that changing behaviors in categories (II), (III), (IV), and (V) had no effect, but intervention with conversational partners, that is, increasing communication between certain pairs, reduced group stress levels. This finding suggests that it is possible to mitigate group stress, particularly through interventions in workplace interpersonal relationships. Biological sciences/Psychology Biological sciences/Psychology/Human behaviour Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Workplace stress management is a crucial issue, and traditional approaches to it and efforts to improve it have focused on the individuals experiencing stress. However, it is widely known that organizational culture and relationships with colleagues are strongly linked to stress and that job strain, a bullying organizational culture, and lack of autonomy in decision-making are associated with the onset of depression 1 . Also, the type of leadership exhibited by superiors can affect subordinates' well-being 2,3 , and psychological states are influenced by interactions with others, as network analyses have demonstrated 4, 5 . In particular, individuals who experience low network cohesion with their colleagues are at a relatively high risk of depression 4 . Thus, workplace stress is not merely an individual issue but a collective problem arising from interpersonal relationships and group dynamics. In our previous research 6 , we proposed the Group Stress Level, a measure estimated using data gathered from accelerometers attached to the bodies of individuals in a group. More specifically, the group stress level is calculated as the linear sum of the scale exponents, with the focus being on the fit of the cumulative frequency distribution of activity durations to a stretched exponential distribution. This value has been confirmed to be close to the results obtained using the CES-D (Center for Epidemiologic Studies Depression) questionnaire, which is a self-report depression scale for research in the general population. This confirmation has enabled the estimation of the group's state as the group stress level. However, the potential for reducing the group stress level has not been previously addressed. Therefore, in this study, we hypothesized that behavioral change can effectively reduce the group stress level. We intervened in workplace behaviors through advice provided by a smartphone app to verify the potential for reducing the group stress level. The experimental outline was as follows. The interventions were conducted using an entirely automated system. The participants were equipped with wearable sensors to detect physical movement and face-to-face interactions along with a smartphone app for displaying advice. This advice covered five categories: (I) conversational partners, (II) methods of conversation, (III) desk work practices, (IV) arrival and departure times, and (V) time management. Sensor data collected over the subsequent two weeks served to determine the effectiveness of the advice, that is, whether it was followed. Finally, the scale exponent was calculated from the sensor data, and this calculation was used to evaluate the effect of the advice by associating it with changes in the group stress level. Results The participants were employees with the sales departments of three business units in a manufacturing company in Japan. The first phase involved collecting sensor data to generate the advice. In the second phase, the intervention was implemented, with personalized advice based on the statistical processing of data from the first phase being presented through a smartphone app. By comparing the behavioral changes in the office and the group stress level determined by the scale exponent from the first to the second phases, we verified that the behavioral intervention could, in fact, reduce the group stress level. The advice provided to the participants through the smartphone app during the intervention phase covered the five categories just listed (i.e., conversational partners, methods of conversation, desk work practices, arrival and departure times, and time management). Figure 1 presents an example of the interface for the intervention application. The participants were encouraged to read the daily advice and change their behavior accordingly. [Insert Figure 1] The group stress level was calculated as in our previous research 6 . Specifically, the activity level of the individuals in the group were estimated based on the scale exponent (β) of the cumulative frequency distribution of the activity durations (Equation 1) of the time-series data of accelerations, which reflected their physical movements (Equation 2). The group stress level was calculated as the weighted average of the group, with the weight (w i ) being the ratio of the activity level to office stay time (Equation 3). The effect of the intervention was evaluated by comparing the group stress level before the intervention (during the first phase) and after the intervention (during the second phase). Figure 2 is an example of the distribution of the activity durations measured by the wearable sensors before and after the intervention. Figure 2(a) represents the individuals who achieved the goal of advice category (I), showing a decrease in the scale exponent β from 0.87 to 0.75 that indicates a less steep slope post-intervention. Figure 2(b) represents the individuals who did not achieve any of the advice categories (I through V), showing virtually no change in the value of β. [Insert Figure 2] Next, we classified the participants as either “achievers” who had followed at least one piece of advice in each of the five categories and achieved the goal in that category or “non-achievers” who did not follow any of the advice. We then calculated the stress level for each of these groups. The application conditions for the group stress level in the previous study 6 included the requirements that the group not be divided into clusters within the face-to-face network and that the data be from the same phase. After confirming that these conditions had been met, we performed the evaluation. Figure 3 presents a comparison of the group stress levels before and after the intervention for the achiever and non-achiever groups. The results indicate that, in the category of conversational partners (I), the achiever group showed a more positive trend of change than the non-achiever group, specifically, with p<0.05 for Organization 1, p<0.1 for Organization 2, and p<0.05 for Organization 3. We observed no significant differences in categories (II) through (V). [Insert Figure 3] Discussion Impact of the interventions in workplace behavior on group stress levels This study involved an empirical experiment conducted to verify that the proposed interventions in workplace behaviors have the potential to reduce group stress levels. We observed a significant reduction in the stress level of the group consisting of the participants who followed the advice provided to them. We provided five categories of advice related to measurable and intervenable workplace behaviors, specifically, conversational partners, methods of conversation, desk work practices, arrival and departure times, and time management Among these categories, the interventions that encouraged conversations with individuals proved effective. Notably, the conversations between pairs of individuals within the group reduced their group stress level. There is extensive empirical evidence that stress is influenced by interpersonal relationships 1-5 , this study is the first to suggest that interventions in interpersonal relationships can be effective in reducing group stress levels. All of the pieces of advice used in this experiment were formulated based on behavioral characteristics that correlate with activity levels. As discussed in the Results section, the implementation of the advice relating to conversational partners (I) contributed to changes in group stress levels while the implementation of the advice relating to categories (II) to (V) was not associated with significant changes in group stress levels. This difference in the effectiveness of the advice is likely attributable to differing causal relationships. Thus, the behaviors related to desk work and time management did not seem to influence the group stress level directly but, rather, seemed to be correlated owing to a third factor affecting both behavior and the group stress level. The interventions that we conducted in this study allowed us to examine the causal relationship between behavioral characteristics and group stress levels, and we found that the advice for categories (II) to (V) was less effective than the advice for category (I). Relationship Between Group Stress and Network Structure Our findings suggest that the presence and extent of communication between the individuals who received the advice and specific others with whom they formed pairs of employees within their organizations affected the group stress level. Previous research has demonstrated that the structure (topology) of interpersonal connections within an organization, conceptualized as networks, influences human stress. An example is Christakis’s theory of social contagion 7,8 . Generally, “contagion” is a term used in the context of infectious disease to describe the transmission of material substances such as viruses directly among individuals. According to Christakis, in the context of the structure of human connections revealed by epidemiological data, a similar phenomenon occurs with subjective states such as obesity, loneliness, and happiness in the absence of any material intermediary. These states spread among the individuals in a network in a process that he termed “social contagion”. Christakis and his colleagues also conducted research on depression using the same CES-D employed in the present study and demonstrated that the structure of connections influences the risk of depression 9 . Further, Lee et al. analyzed the network structure of employees using face-to-face communication data from Japanese workplaces obtained with name-tag wearable sensors similar to those used in this study as well as the results of the CES-D 4 . The findings revealed that the employees with a higher clustering coefficient, which is a measure of the degree to which an individual’s immediate contacts are themselves connected, tended to have lower stress levels. This result suggests that individuals within a dense network in the workplace experience lower stress because information transmission is smoother, and they incur lower costs for reconciling differences in perspectives. Therefore, it is plausible that enhancing or increasing the connections between specific pairs could reduce stress for third parties in topological proximity, consistent with the findings of this study and the research by the teams of Christakis and Lee. In addition, the research described here provides new insights in terms of the intervention conducted and the finding that altering the structure of connections can reduce the overall stress risk within a group. The communication partners were selected mechanically based on their statistical association with activity levels. An intriguing question for future research is how these specific pairs are positioned within the broader organizational network from a macro perspective and the influence of the designation of the pairs on the clustering coefficient, as the research by Lee et al. suggests. Limitations and Directions for Future Study In this study, a reduction in the group stress level was observed only when those who followed the advices were considered as a virtual group while no significant difference was observed a whole of the organizations including those who did not followed adovicesit. This result suggests that, while effective outcomes could be achieved when all of the members of a group followed the advice in category (I), variation in the type of advice and the frequency with which it was followed by individuals diluted the overall effect. To reduce the stress level reliably across the entire group, it is necessary to reconsider the motivation for following the advice and explore ways to provide advice that make it easier for individuals to follow it. Moreover, this study is based on three organizations in a single corporation in Japan and is not a randomized controlled trial, so future research could involve randomized controlled trials to assess the effectiveness of interventions more rigorously and evaluate the approach in organizations with diverse backgrounds. Conclusion This study verified that group stress levels can be decreased through interventions in workplace behaviors. In particular, the interventions encouraging conversations between specific individuals were effective. While it is well known empirically that interpersonal relationships influence stress, this study is the first to suggest experimentally that interventions focusing on the network structure of connections can be effective in reducing the group stress level. Methods Data Collection and Group Stress Level s In prior research, we proposed a method for estimating group stress levels based on physical movements 6 . To collect the data on group stress levels and workplace behavior for the present study, we used name-tag wearable sensors similar to those used in our previous study 6 . As Figure 4 shows, the wearable sensors were equipped with accelerometers to capture physical movements and infrared transceivers to detect face-to-face communication between individuals 10,11 . Thus, we were able to capture the participants’ authentic movement data during office work without interfering with their activities. We used the method developed in our previous research to calculate group stress levels 6 . It has been demonstrated that the cumulative distribution of the activity duration fits a stretched exponential distribution 12 , as Equation 1 represents, and, in the present study 6 , the linear sum of the scale exponent ( b i) for the group, the group stress level, closely approximated the average stress level of the same group obtained using a CES-D questionnaire (Equations 2 and 3). Equations 2 and 3 indicate that the individual βi values reflect the individual i’s impact on the group stress level. In other words, to reduce the group stress level, it is necessary to decrease b (that is, make the distribution slope less steep). Additionally, we employed the constants a and b in Equation 2 from our previous study 6 in this research. [Insert Figure 4] Stu dy Design The study consisted of two experimental phases. During the first, the data were collected from the participants’ name-tag wearable sensors as they conducted their usual work without receiving any feedback. During the second, intervention phase, the participants received personalized advice based on the statistical processing of data from the first phase through a smartphone app. The data collection was performed during the second phase in the same manner as the first. In evaluating the results, we assessed the effect of the intervention on the group stress level by determining whether the participants followed the advice during the second phase and assessing any changes in the group stress level across both periods. Hypotheses for the Five Categories We hypothesized that behavioral changes could reduce group stress levels. We categorized the behavioral characteristics that we identified as either (I) conversational partners, (II) methods of conversation, (III) desk work practices, (IV) arrival and departure times, or (V) time management. We defined the characteristics that could be calculated based on the sensor data for each category. Table 1 provides examples of the corresponding advice. Our first hypothesis concerns conversational partners (I). We considered the possibility that conversing with specific individuals could influence the overall stress level of the group. For example, the amount of social support from supervisors or colleagues is related to workplace stress 1-3,5,13 . Since interactions between individuals in a group, even if they are unconscious, can affect group stress, we reasoned that communication is a critical aspect of workplace stress. The characteristic for this category, denoted Characteristic C, was calculated using Sato's method 14 of counting the number of face-to-face interactions between any two individuals. Our second hypothesis concerns changing the methods of conversation (II), differing from (I) in providing advice that does not specify a conversational partner. Since enriching workplace communication is associated with productivity and engagement 15-18 , the frequency of greetings among employees and the number of meetings also have the potential to affect the workplace culture and the risk of stress. Additionally, the movements of the participants in conversations have been used to assess the quality of communication 14,19 . Specifically, when the bodies of both parties are moving, the conversation is considered interactive; when only one participant is moving, that individual is the speaker while the less active participant is the listener 14 . Furthermore, the levels of stress in organizations in which conversational roles are not fixed and all members are equally empowered to speak appear to be lower than in organizations that do not thus empower their members 14 . Based on these considerations, we reasoned that providing advice to change the frequency of greetings, the duration of a single meeting, and the conversational roles in meetings (interactive, speaker, listener) can impact the group stress level, defining characteristics (D1–D7) in Table 1. The third hypothesis concerns desk work practices (III). It has been shown that time is required to refocus on the original task following interruptions in the form of speech and other distractions during desk work 20 . Conversely, the ability to maintain concentration in the absence of external interruptions for extended periods is believed to enhance cognitive productivity. Additionally, the demonstration that moving the body at the appropriate frequency influences motivation 21 suggests that there may be an optimal duration for continuous desk work. We used Sato’s definition of the behavioral characteristics of office workers 14 , considering them engaged in desk work when they were not facing someone and had minimal body movement (frequencies of less than 2Hz). The feature (E1–E4) in Table 1 is identified by counting the occurrences of this state across different durations. The fourth hypothesis concerns arrival and departure times (IV). Some people, referred to colloquially as “morning larks,” tend to be alert and productive in the early hours, while others, “night owls,” achieve peak performance during the evening or at night, though it remains unclear which if either behavioral pattern is more productive. Therefore, we included arrival and departure times in our hypothesis and prepared the features (F1–F2) in Table 1. Our last hypothesis concerns time management (V) in relation to the scheduling of conversations and desk work. Productivity peaks in the early afternoon 22 while short-term memory tasks are more effectively performed in the morning 23 , possibly because certain cognitive processes are influenced by the circadian arousal level. Thus, since the scheduling of members can impact the group stress level, we defined time-specific features (G1-G8) in Table 1. [Insert Table 1] Advice Derivation Procedure Personalized advice based on the sensor data collected during the first phase of the research (which lasted four weeks) was provided to each participant through a smartphone app during the second phase. For individual i, we calculated the activity level i (Equation 2) and the characteristics (Cij to G8i). In computing the correlation coefficients, the activity level served as the dependent variable and the characteristics served as the independent variables. Within each category, the advice corresponding to the characteristic with the highest absolute value of the correlation coefficient was displayed on the app. The advice text was predetermined, but, when the correlation coefficient was positive, the phrase “increase ~” was chosen; when the coefficient was negative, “decrease ~” was selected. In this way, we identified behaviors that are characteristically more (or less) prevalent on the days in which the individuals’ activity levels were high and intervened to reinforce these behaviors. Intervention App Smartphones are considered an effective intervention method for individual stress in the context of digital improvement programs because employees can easily access the programs. Ease of access facilitates early preventative intervention better than traditional methods such as consultation with a physician. Additionally, digital platforms allow employees to engage with the chosen content at their own pace, further fostering active participation and increasing the likelihood of desirable behavioral change 24 . Our design insights for the intervention application include presenting the intervention program after self-monitoring to gain an objective view of an individual’s state 24 , providing concrete advice such as meeting people, walking, or taking a three-minute breathing break 25 , and incorporating gamification based on achievement points 26 . In the intervention phase of this study, the prior findings were expanded to reduce group stress, and the intervention UI(user interface) was designed as a smartphone app. The app included features for daily activity log feedback for self-monitoring, the presentation of personalized concrete behavioral advice, and the display of achievement results as scores. Figure 5 shows the interface of the intervention appl, which provides users with the opportunity and motivation to consider the appropriate actions to take and when to do so during their work activities. The top left of Figure 5(a) displays the activity level, and Figure 5(c) shows the previous day’s activity log collected by the wearable sensor, thus encouraging users to self-monitor their past behaviors and states and motivating improvement. Additionally, Figure 5(b) presents the text of the advice and the specific target values. The application automatically determines whether the advice has been followed based on the sensor data and displays the number of pieces of advice followedas scores on the bar graph at the top of Figure 5(a). Tapping on the advice in Figure 5(a) opens a detailed screen in Figure 5(b), which includes a verbal description of the advice and presentation of the conditions related to the user, such as the time slots and durations defined in the condition categorization shown in Table 1. [Insert Figure 5] Participants and Settings An empirical experiment was conducted using the intervention app and name-tag wearable sensors discussed previously. The participants were, as mentioned, employees with the sales departments of three business units within a Japanese manufacturing company. The experiment took place over six weeks in June and July 2016. There was no remote work; all of the employees were in the office. The collection of the sensor data continued over the six weeks, with the first four weeks of data serving for analysis and the intervention app being provided to the participants during the fifth and sixth weeks. A total of 245 individuals from the three organizations participated in the experiment. Specifically, Organization 1 had 168 participants with 120 valid , Organization 2 had 125 with 69 valid, and Organization 3 had 79 with 56 valid, with the participants who collected a sufficient amount of data being considered valid. The determination was as follows. Since our previous research6 showed that at least 1,000 counts (1,000 minutes = approximately 16.6 hours) of accelerometer data per individual are required to calculate the group stress level for a given period, in the present study, only the data from the individuals who had more than 1,000 counts during the third to fourth weeks and fifth to sixth weeks, when the participants were wearing the sensors, were considered for the evaluation of group stress levels. Evaluation For the relevant behavioral characteristics, an increase over the average of past data served to indicate that the advice had been followed. The group stress level during the third to fourth weeks, before the intervention, served as the baseline for assessing the changes observed during the fifth to sixth weeks. Ethical Approval and Informed Consent In this study, which involved collecting workplace behavioral data and delivering advice via participants' smartphones, it was necessary to obtain participants' names, email addresses, departmental affiliations, and behavioral data through wearable sensors. Prior to the commencement of the experiment, the researchers provided a clear explanation of the study's purpose and the nature of the data to be collected. Informed consent was obtained from all participants, ensuring their understanding and voluntary agreement to participate. Furthermore, the study was conducted under the review and approval of the Hitachi Group's Ethics Review Committee. This committee verified that the research aims and methodologies were justified and rational, and that the rights and dignity of the subjects and human groups involved were protected. All experimental protocols were thus executed following these ethical standards. Declarations Competing Interests This study was conducted with research funding provided by Hitachi, Ltd. Author Contribution S.T. designed the study, analyzed the data, and wrote the manuscript. N.S. was involved in the data collection and provided feedback. K.Y. and Y.M. designed and supervised the study. All of the authors contributed to the revision of the manuscript and approved the final manuscript. Data Availability The datasets used and/or analysed during the current study available from the corresponding author on reasonable request. Additional information This study was conducted with research funding provided by Hitachi, Ltd. Cod e availability The Python code that supports the findings of this study is available from the authors upon request. References Theorell, T. et al . A systematic review including meta-analysis of work environment and depressive symptoms, BMC Public Health , 738 (2015). Inceoglu, I. et al. Leadership behavior and employee well-being: an integrated review and a future research agenda, Leadership Q. 29 , 179–202 (2017). Nappo, N. Job stress and interpersonal relationships cross country evidence from the EU15: a correlation analysis, BMC Pub. Health 20 , 1143 (2020). Lee, J. H., Sato, N., Yano, K. & Miyake, Y. Universal association between depressive symptoms and social‑network structures in the workplace. Sci. Rep. 12 , 10170 (2022). Kalish, Y., Luria, G., Toker, S. & Westman, M. Till stress do us part: on the interplay between perceived stress and communication network dynamics. J. Appl. Psychol. 100 , 1737–1751 (2015). Tsuji, S., Sato, N., Ara, K. & Yano, K. Estimating group stress level by measuring body motion Front. Psychol. 12 , 634722 (2021). Christakis, N. A. & Fowler, J. A. Social contagion theory: examining dynamic social networks and human behavior. Statics in Med. 32 , 556–577 (2012). Christakis, N. A. Connected: the surprising power of our social networks and how they shape our lives (Little, Brown and Company, 2009). Rosenquist, J., Fowler, J. & Christakis, N. Social network determinants of depression. Mol. Psychiatry 16 , 273–281 (2011). Wakisaka, Y., et al. Beam-scan sensor node: reliable sensing of human interactions in organization in 2009 Sixth International Conference on Networked Sensing Systems (INSS) , Pittsburgh, PA, USA, 1-4 (2009). Hayakawa, M., Ohkubo, N. & Wakisaka, Y., Business microscope: development of a practical human behavior measurement system. IEICE Transactions on Communications, J96-D , 2359–2370 (2013) (in Japanese). Nakamura, T., Kiyono, K., Yoshiuchi, K., Nakahara, R., Struzik, Z. R. & Yamamoto, Y. Universal scaling law in human behavioral organization. Phys. Rev. Lett. 99 , 138103 (2007). Mahan, P. L., et al. Work environment stressors, social support, anxiety, and depression among secondary school teachers. AAOHN J. 58 , 197–205 (2010). Sato, N., et al. Knowledge-creating behavior index for improving knowledge workers' productivity in 2009 Sixth International Conference on Networked Sensing Systems (INSS) , Pittsburgh, PA, USA, 1-4(2009). Ditta, A., et al. Analyzing the relationship between productivity and human communication in an organizational setting. PLoS ONE 16 , e0250301 (2021). Santalla-Banderali, Z. & Alvarado, J. M. Incidence of leader-member exchange quality, communication satisfaction, and employee work engagement on self-evaluated work performance. Int. J. Environ Res. Pub. Health 19 , 8761 (2022). Bakker, A. B. & Xanthopoulou, D. The crossover of daily work engagement: test of an actor-partner interdependence model. J. Appl. Psychol. 94 , 1562–1571 (2009). Zoonen, W. V. & Sivunen, A. E. The impact of remote work and mediated communication frequency on isolation and psychological distress. Eur. J. Work Organ. Psychol. 31 , 610–621 (2021). Higo, N., et al. Interpersonal similarity between body movements in face-to-face communication in daily life. PLoS ONE 9 , e102019 (2014). Mark, G., Gudith, D. & Klocke, U. The cost of interrupted work: more speed and stress. in Proc. SIGCHI Conf. Hum. Factors Comput. Syst. ( CHI '08) 107–110. (2008). Signorini, G., et al. Enhancing motivation and psychological wellbeing in the workplace through conscious physical activity: suggestions from a qualitative study examining workers' experience. Front. Psychol. 13 ,1006876 (2022). Gaggero, A. & Tommasi, D. Time of day and high-stake cognitive assessments. Econ. J. 133 , 1407–1429 (2023). Folkard, S. Diurnal variation in logical reasoning. Br. J. Psychol. 66 , 1–8 (1975). Weber, S., Lorenz, C. & Hemmings, N. Improving stress and positive mental health at work via an app-based intervention: a large-scale multi-center randomized control trial. Front. Psychol. 10 , 2745 (2019). Morrison, L. G., et al. The effect of timing and frequency of push notifications on usage of a smartphone-based stress management intervention: an exploratory trial. PLoS ONE 12 , e0169162 (2017). Parks, A. C., Williams, A. L., Tugade, M. M., Hokes, K. E., Honomichl, R. D. & Zilca, R. D. Testing a scalable web and smartphone based intervention to improve depression, anxiety, and resilience: a randomized controlled trial. Int. J. Wellbeing 8 , 22–67 (2018). Table 1 Table 1. Correspondence Between Advice Categories and Behavioral Characteristics. For each individual i, the daily aggregated behavioral characteristics were calculated, and the median value for each individual during the data collection period served as the baseline for determining the achievement of the goal associated with the advice. Behavioral Characteristics Definitions Example Advice Statements (I) conversational partners C i,j Number of Conversations with Individual (j) count of conversations between individual i and individual j i “Increase the frequency of conversations with Mr./Ms. J.” (II) methods of conversation D1 i Short Conversations Count number of conversations ending in less than 5 minutes ii “Increase/Decrease the number of short conversations and greetings within 5 minutes.” D2 i Mid-short Conversations Count number of conversations lasting between 5 to less than 15 minutes ii “Increase/Decrease the number of 5-15 minute stand-up chats and consultations.” D3 i Mid-long Conversations Count number of conversations lasting between 15 and less than 30 minutes ii “Increase/Decrease the number of conversations that are less than 30 minutes.” D4 i Long Conversations Count number of conversations lasting more than 30 minutes ii “Increase/Decrease the number of lengthy conversations more than 30 minutes.” D5 i Bi-directional Conversation Time total time in face-to-face state with any other individual during which both the individual and at least one other person are moving iii “Increase/Decrease the time you and another person are actively engaging in conversation.” D6 i Speaker’s Conversation Time total time in face-to-face state with any other individual during which the individual is moving and no one else iii “Increase/Decrease the time you spend talking as the speaker in conversations.” D7 i Listener’s Conversation Time total time in a face-to-face state with any other individual during which the individual is not moving but at least one other individual is iii “Increase/Decrease the time you spend listening as the listener in conversations.” (III) desk work practices E1 i Short Desk Work Count number of desk work sessions that ended in less than 5 minutes iv “Frequently get up and walk around, and try to complete desk work within 5 minutes.” E2i Mid-short Desk Work Count number of desk work sessions lasting between 5 and less than 15 minutes iv “Increase/Decrease the number of desk work tasks that can be completed within 5-15 minutes.” E3i Mid-long Desk Work Count number of desk work sessions lasting between 15 and less than 30 minutes iv “Increase/Decrease the number of desk work tasks that can be completed within 15-30 minutes.” E4i Long Desk Work Count number of desk work sessions lasting more than 30 minutes iv “Try to concentrate on desk work for more than 30 minutes continuously.” (IV) arrival and departure times F1 i Arrival Time time of the first accelerometer Data entry of the day v “Start your workday earlier/later.” F2 i Departure Time time of the last accelerometer data entry of the day v “Finish your workday earlier/later.” (V) time management G1 i Conversation Time (Before Work Hours) total time detected in face-to-face interaction with any individual between 6 am and 9 am “Try to increase/decrease conversations before starting work hours.” G2i Conversation Time (am) total time detected in face-to-face interaction with any individual between 9 am to 12 pm “Have more/fewer conversations during the morning hours.” G3i Conversation Time (pm) total time detected in face-to-face interaction with any individual between 12 pm to 5 pm “Engage in more/fewer conversations during the afternoon hours.” G4i Conversation Time (After Work Hours) total time detected in face-to-face interaction with any individual between 5 pm to 10 pm “Increase/decrease conversations after work hours end.” G5i Desk Work Time (Before Work Hours) total time not in face-to-face interaction and no movement detected between 6 am to 9 am “Do more/less desk work before starting work hours.” G6i Desk Work Time (am) total time not in face-to-face interaction and no movement detected between 9 am to 12 pm “Conduct more/less desk work during the morning hours.” G7i Desk Work Time (pm) total time not in face-to-face interaction and no movement detected between 12 pm to 5 pm “Increase/decrease the amount of desk work in the afternoon.” G8i Desk Work Time (After Work Hours) Total time not in face-to-face interaction and no movement detected between 5 pm to 10 pm “Do more/less desk work after work hours end.” i) Count of Conversations Between Two Individuals : Count each period in which person i and person j are detected in a continuous face-to-face state as one conversation. ii) Conversation Count : One comversation is a continuous face-to-face state with two or more people of the same member composition iii) Movement Determination : When the frequency of acceleration due to body movement is above 2Hz, it is considered “moving”; below 2Hz is considered “not moving. iv) Desk Wor k: A state that is not face-to-face and lacks movement (the acceleration frequency is below 2Hz) is determined to be desk work. v) Arrival and Departure Times : Since the name tag wearable sensor does not monitor acceleration sensing while connected to the charger, the first and last timestamps of the acceleration data are considered as the arrival and departure times. Additional Declarations Competing interest reported. This study was conducted with research funding provided by Hitachi, Ltd. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4240426","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":294271348,"identity":"c9236fa7-0471-4ae8-bfda-2dfda835065a","order_by":0,"name":"Satomi Tsuji","email":"data:image/png;base64,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","orcid":"","institution":"Tokyo Institute of Technology","correspondingAuthor":true,"prefix":"","firstName":"Satomi","middleName":"","lastName":"Tsuji","suffix":""},{"id":294271350,"identity":"2b9aee93-e3fd-4af1-a169-162f800ef08c","order_by":1,"name":"Nobuo Sato","email":"","orcid":"","institution":"Happiness Planet, Ltd","correspondingAuthor":false,"prefix":"","firstName":"Nobuo","middleName":"","lastName":"Sato","suffix":""},{"id":294271352,"identity":"c110f269-1979-4ed9-bf76-6fa6afcf7c13","order_by":2,"name":"Kazuo Yano","email":"","orcid":"","institution":"Hitachi, Ltd","correspondingAuthor":false,"prefix":"","firstName":"Kazuo","middleName":"","lastName":"Yano","suffix":""},{"id":294271354,"identity":"7805b2f2-816d-4d60-9e06-7fd5c46f97c0","order_by":3,"name":"Yoshihiro Miyake","email":"","orcid":"","institution":"Tokyo Institute of Technology","correspondingAuthor":false,"prefix":"","firstName":"Yoshihiro","middleName":"","lastName":"Miyake","suffix":""}],"badges":[],"createdAt":"2024-04-09 07:50:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4240426/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4240426/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":55516157,"identity":"ad39f315-da76-4330-91ef-7b998b8559eb","added_by":"auto","created_at":"2024-04-29 13:13:11","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":314752,"visible":true,"origin":"","legend":"\u003cp\u003eWearable sensor and smartphone app for tje intervention. The wearable sensor is equipped with an infrared transmitter and receiver to detect face-to-face interactions as well as an accelerometer to measure body movement. The application presents behavioral goals tailored to individuals across five categories: (I) conversational partners, (II) methods of conversation, (III) desk work practices, (IV) arrival and departure times, and (V) time management. Tapping a button displays more detailed advice, such as (I) “Have a conversation with Taro Hitachi for more than 5 minutes”, (II) “Engage in short conversations with more than three people”, (III) “Concentrate on desk work for more than 45 minutes”, (IV) “Leave the office by 7 pm”, and (V) “It is recommended to have meetings in the morning”.\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-4240426/v1/d11ac27c753917d07c6f09c1.jpeg"},{"id":55517099,"identity":"2fafdad2-58cb-454c-8f0a-c8803826542b","added_by":"auto","created_at":"2024-04-29 13:21:10","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":331423,"visible":true,"origin":"","legend":"\u003cp\u003eChange in the active duration distribution and scale exponent. (a) achievers, (b) non-achievers.\u003cstrong\u003e \u003c/strong\u003eUsing the same method as in our previous study\u003csup\u003e6\u003c/sup\u003e, we counted the frequency of durations in which the frequency measured by the accelerometer exceeded a certain threshold (active duration) and displayed the cumulative distribution before and after the intervention. Panel (a) represents an individual who achieved the goal set in advice category (I) four times, with the scale exponent β changing from 0.87 to 0.75 (and the activity level from 24.4 to 29.7) when fitted to a stretched exponential distribution. Panel (b) represents an individual who did not achieve the goal even once, showing no change in the shape of the distribution.\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-4240426/v1/3dff18a12b2e04e3a38916df.jpeg"},{"id":55516156,"identity":"3450e46c-c4a1-4f9b-a416-78a887bf9de7","added_by":"auto","created_at":"2024-04-29 13:13:10","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":365891,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between the achievement of the advice and group stress levels\u003cstrong\u003e. \u003c/strong\u003eThe participants were classified as achievers or non-achievers based on whether they achieved the goal set in the advice across the five categories of (I) conversational partners, (II) methods of conversation, (III) desk work practices, (IV) arrival and departure times, and (V) time management. The figure shows the change in the stress level for each group. The error bars represent the standard error of the activity level. In conducting the one-sided t-tests, ** p\u0026lt;0.05 and * p\u0026lt;0.1 denote the significance levels.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-4240426/v1/2d9ecb333b485b4606224df9.jpeg"},{"id":55516153,"identity":"568c2085-7dd1-4f33-a9c4-b13783fb3783","added_by":"auto","created_at":"2024-04-29 13:13:10","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":185408,"visible":true,"origin":"","legend":"\u003cp\u003eSpecification of the name tag-shaped wearable sensor node.\u003cstrong\u003e \u003c/strong\u003eThe nodes continuously measure the wearers’ body motions and face-to-face communications. The accelerometer in the nodes measures body motion at a frequency of 51.2 Hz. Face-to-face communication is detected by transmitting infrared signals between the sensor nodes when wearers face each other about 3 meters apart.\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-4240426/v1/6061d484fb504974d74fd134.jpeg"},{"id":55516154,"identity":"87508d99-44e1-4813-995f-029b82e2d619","added_by":"auto","created_at":"2024-04-29 13:13:10","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":496678,"visible":true,"origin":"","legend":"\u003cp\u003eIntervention application. (a) screen top, (b) advice detail, (c) activity log. The top left of (a) displays the activity level, and (c) shows the previous day’s activity log collected by the wearable sensor, thus encouraging users to self-monitor their past behaviors and states and motivating improvement. Additionally, (b) presents the text of the advice and the specific target values. The application automatically determines whether the advice has been followed based on the sensor data and displays the number of pieces of advice followedas scores on the bar graph at the top of (a). Tapping on the advice in (a) opens a detailed screen in (b), which includes a verbal description of the advice and presentation of the conditions related to the user, such as the time slots and durations defined in the condition categorization shown in Table 1.\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-4240426/v1/c8631ff0a91a4502ed4f5218.jpeg"},{"id":60879762,"identity":"7208af78-61db-43f6-8f8c-a4b0ceae0496","added_by":"auto","created_at":"2024-07-23 06:50:44","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2285355,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4240426/v1/e933a10f-755d-4fe5-b79a-2c6d659c07ac.pdf"}],"financialInterests":"Competing interest reported. This study was conducted with research funding provided by Hitachi, Ltd.","formattedTitle":"Behavioral Interventions in Face-to-face Communication to Decrease Group Stress Levels","fulltext":[{"header":"Introduction","content":"\u003cp\u003eWorkplace stress management is a crucial issue, and traditional approaches to it and efforts to improve it have focused on the individuals experiencing stress. However, it is widely known that organizational culture and relationships with colleagues are strongly linked to stress and that job strain, a bullying organizational culture, and lack of autonomy in decision-making are associated with the onset of depression\u003csup\u003e1\u003c/sup\u003e. Also, the type of leadership exhibited by superiors can affect subordinates\u0026apos; well-being\u003csup\u003e2,3\u003c/sup\u003e, and psychological states are influenced by interactions with others, as network analyses have demonstrated\u003csup\u003e4, 5\u003c/sup\u003e. In particular, individuals who experience low network cohesion with their colleagues are at a relatively high risk of depression\u003csup\u003e4\u003c/sup\u003e. Thus, workplace stress is not merely an individual issue but a collective problem arising from interpersonal relationships and group dynamics.\u003c/p\u003e\n\u003cp\u003eIn our previous research\u003csup\u003e6\u003c/sup\u003e, we proposed the Group Stress Level, a measure estimated using data gathered from accelerometers attached to the bodies of individuals in a group. More specifically, the group stress level is calculated as the linear sum of the scale exponents, with the focus being on the fit of the cumulative frequency distribution of activity durations to a stretched exponential distribution. This value has been confirmed to be close to the results obtained using the\u0026nbsp;CES-D (Center for Epidemiologic Studies Depression)\u0026nbsp;questionnaire,\u0026nbsp;which is a self-report depression scale for research in the general population.\u0026nbsp;This confirmation has enabled the estimation of the group\u0026apos;s state as the group stress level. However, the potential for reducing the group stress level has not been previously addressed.\u003c/p\u003e\n\u003cp\u003eTherefore, in this study, we hypothesized that behavioral change can effectively reduce the group stress level. We intervened in workplace behaviors through advice provided by a smartphone app to verify the potential for reducing the group stress level. The experimental outline was as follows. The interventions were conducted using an entirely automated system. The participants were equipped with wearable sensors to detect physical movement and face-to-face interactions along with a smartphone app for displaying advice. This advice covered five categories: (I) conversational partners, (II) methods of conversation, (III) desk work practices, (IV) arrival and departure times, and (V) time management. Sensor data collected over the subsequent two weeks served to determine the effectiveness of the advice, that is, whether it was followed. Finally, the scale exponent was calculated from the sensor data, and this calculation was used to evaluate the effect of the advice by associating it with changes in the group stress level.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eThe participants were employees with the sales departments of three business units in a manufacturing company in Japan. The first phase involved collecting sensor data to generate the advice. In the second phase, the intervention was implemented, with personalized advice based on the statistical processing of data from the first phase being presented through a smartphone app. By comparing the behavioral changes in the office and the group stress level determined by the scale exponent from the first to the second phases, we verified that the behavioral intervention could, in fact, reduce the group stress level. The advice provided to the participants through the smartphone app during the intervention phase covered the five categories just listed (i.e., conversational partners, methods of conversation, desk work practices, arrival and departure times, and time management). Figure 1 presents an example of the interface for the intervention application. The participants were encouraged to read the daily advice and change their behavior accordingly.\u003c/p\u003e\n\u003cp\u003e[Insert Figure 1]\u003c/p\u003e\n\u003cp\u003eThe group stress level was calculated as in our previous research\u003csup\u003e6\u003c/sup\u003e.\u0026nbsp;Specifically, the activity level of the individuals in the group were estimated based on the scale exponent (β) of the cumulative frequency distribution of the activity durations (Equation 1) of the time-series data of accelerations, which reflected their physical movements (Equation 2). The group stress level was calculated as the weighted average of the group, with the weight (w\u003csub\u003ei\u003c/sub\u003e) being the ratio of the activity level to office stay time (Equation 3). The effect of the intervention was evaluated by comparing the group stress level before the intervention (during the first phase) and after the intervention (during the second phase).\u003c/p\u003e\n\u003cp\u003e\u003cimg src=\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAjcAAABwCAYAAADi6lFTAAAgAElEQVR4Ae2ZjZUTu9JFSYEYSIEcCIEYSIEMyIAMiIAISIAEbgbkMN/a897+3kFXstvjtj1jn1qrUbdUqp9TVVJ5ePdUKgJFoAgUgSJQBIrAHSHw7o58qStFoAgUgSJQBIpAEXhqc9MkKAJFoAgUgSJQBO4KgTY3dxXOOlMEikARKAJFoAi0uWkOFIEiUASKQBEoAneFQJubuwpnnSkCRaAIFIEiUATa3DQHikARKAJFoAgUgbtCoM3NXYWzzhSBIlAEikARKAJtbpoDRaAIFIEiUASKwF0h0ObmrsJZZ4pAESgCRaAIFIE2N82BIlAEikARKAJF4K4QaHNzV+GsM0WgCBSBIlAEikCbm+ZAESgCRaAIFIEicFcItLm5q3DWmSJQBIpAESgCRaDNTXOgCBSBIlAEikARuCsE2tzcVTjrTBEoAkWgCBSBItDmpjlQBIpAESgCRaAI3BUCbW7uKpx1pggUgSJQBIpAEWhz0xwoAkWgCBSBIlAE7gqBNjd3Fc46UwSKQBEoAkWgCLS5aQ4UgQGBb9++PX39+nWYfX2f//zzz9OnT5+e3r179/Tx48enP3/+PBv548ePp8+fPz+xXioCReB6CHz58uV6yt6gJs6mX79+XcXyNjdXgblK3goCNAsU4FsgbKURg3jPQ+P3799PHz58eGIsFYEicHkEaGzeytlxeTTWGsazas153kqbm/Pw27ybvwTwC9vn/fv3/38xbRHCr3Bk/Pz5cwv7bjz8RUCbx5HLlGKmqP2rwW6KbyCIRoG/eNyaxlxJ3G1myAeaF4g48D7GgNhwkFySLpmX+ENu6f9b+GvaIazvqVYO+fmIa8TW2sR/c3Y15g+RW+DFueD5cUn91C/PSNfQ3eZmRP1C3xzUJLrNCSPfW35Zw0OT8dL/ZhgvvVNc1Ebtxxbeac60hzmK5Rw9p9h0KV582hKPS+lXLlh+//79+ZO4+848hyjEyGXPwXEoNzhELnWQnpuX+rsa8Rf/yCt0kYuvlWiKt+B8L7XyWuNwK7tmuUmNzi5x5qzja9iLri25uact1C6Y8MyaG2yaze9pw+s9Lfb08hXIohEg0NkA8L0l6UiULXwrNzlQubi5DHnfSvCagDRjyJDGv3BwAcsrz1sa8W92EN3CB5oViFwhR4wZvwx9J5Y2yuBuAzTaawM0zu/xfW5eHrKBfCffrBfr59CeW60Rk7G2D9ny1mvlkG+PuMZFTS2MxHyeKZ6P5DZr16JL1ukxH/BZv5P3GvXc5iYRv+A7yeylhRqD62W1Uk0hZIHIZ7PB5QUhnwMWuTPiktAG7ODdi2PGP84d+y8bL+JTZI46tn6DCT7gL9ioExuZy0tf7CgwLkvw4XLhPePB3tkBtdUm+WyStIP45iUt35bRC37Gi+3GmkYT+2dkzGdr58yt8hKZ+AyWYGB+ggF7tlI2ZcYbX7ZQ6jcO7gMnbGE+D93MD/SM+cF+bGIfD3mFHnLJOUd1rcZr1srKhs7vhwA5lbmkZPLI84c58mNG7DV3qGlrmf3Os481v7OWko91zwV45HdkztpM25DPt3zIhFIG365vPSvxbYaNsrT1WdnO/8zR3llJxf3nYDTIBJTLaUuCkOheEIkjczY4HLLKTp7VO/rZTzKzLwtltQdbLbpDPBbFiod55Fgkq3G1H1vzosQH5vCJiwa/uOx51xZGLzz8BTcvGH3f4t/KJueNBzKRrx3HcHP/OLJvlSNg4CUNjw3eKANbwHhvwrZZXhIH7AJ/CAywYeXHyi74iRsjMUZukjmUc7yP+tlvjLHXphC8kIuOMT/4Rk7ihgz42ceT/uBrfo82zb7hR0/p7SPAmTI7f4nveL6N3sJDXknkRX7P6heZ5rR5mfszF81j+eWb6U0f0GF+KsN68HuUqewcV9jAk37knr3e9z/19rLszuRwqBJMHg5/gr66kNJ1EvXQ5Yi8TMrcu+WdvVuSbAvPMVu32HOMh8K3OaH4wJJikyz2GWbs9dKFP31a2e4lCu/qUfcoH5lc7itC3sxO+Vc24ePWmMOLnr1pZRt2pc/4Z0Nxig3YbFyRgb4k45JzvI/6XZ8dyKMPxM940AyTW5LfswMdne6T/9g46j7G3/XXiwCxn9Xj2EDM6pCcs4nAQ2QxJ83qFznmoeurWpnlPbLTNnmUoR3W3Gx9tFt7xxFcZtjAlzU+7tvje/9Tbw+r7kwGTUwm5CnuHTsEWc/i2CKbZD3lLzcc7LPCHHUds3XkP/XbIsMWLh4uUWwbaVY0xkBev91/ru3alg3rsQMAOw9divjoIabdjOzJJi3XxncPv3H+3O8VXvhM8ykdOtzkGUdszgOe9VlMx318j/rlycPcueQd4wd/Nmns8S9z42GNnFmc1DMbV/jNeDv3uhEgV4jnSGPOjXkDP7kDnzTumdUvtTDmG/qZ50lbzOuRP/XMdGSTpQxGabTb+XFc1b8yR/49v9vc7InmQpaH4mL54PShQxC5/Cqe/ffAKJRLl4SGn4f3vIhH/vze+mf3Q7amPArHQlyNye/7rAhdc7Qo84Jlje8s+ixu1rfarp5xHG0DM3x7KW1tKI/JH+06xr91fYUXPqMT0ge/t8oGu7wI9GFLvqb+1EdeZPxHmdaSe9A/ayCxIZsivl8S5xV+6u/4thCY5cB4xqRH5vfYJHh+yWue+s2YOY6O/B7z3CZirMG0TZ5sXlLObH20O+3L91Vzg34xSP49319++u5pxZ3L4hcgDcWMOBwJMgk6KxCSbJYEXBzMcyiTaCTg7DBGJ7z8FYAmiPdTiYM4L4bVfnhI2kuRl6VFiL/pM7r5xk+w4Vt/wdEYUOjgkQ1QFvNL7PcQIp68YwPx5BvZpxJ7Zvlwqhww0O9T9x7ix75ZXmIzuBIj1tEtHsqDx7g4lyP1wgN2yEHGmH8rfIirOZj5wRx1kjLzR4E5ox3ow270YKt/xWE/csydjDs61M3+0WZlM7Imb873/W0iQCzJlSTmyJWRiL21w57k4T2/yX/qhTyD2JffKYt11sa8S371jrbN5Jif2sAoYaPrzs1G9Kkz19PHnN/zvc3NnmhOZHn4kWBj8sOewZ9dQuwfE8FmiYOWhwOdvbzPaDU/4x3nsBnbeQ4lMzrgyQIYZe3xbdOAz2Cnb36jQ8yzqCxe+HhGX2wSz7ERHWDgpUncmNPGrbK1H1npw9b9yTde2rl2zvssL5Fnvug39oODzQxYgP8hTOBnH3xgYKOT9qon53gnrsbYOMhjfFgfa5G1bJTRCR9281Bf2JLxRS5+uIY+/UIez4zgQc6la2Wmu3OXQ4CcNabmymrMujYv4bVe0krmlEN+Ux98855nBXOQvMrI/cylPniRASlX2cyN8sfvsY6eBcU5oC2M8jKqU/5LjG1uLoHqCTI5QI8Rh+c5yTAmZCac7+fIx34SNgv2mE/XXsdPL56Vbi+z1fpbnOfAOje2K79fkpc0EIfs8eJf6byH+ddeK/eA8a18oN7OIXLjXBnn6L/0XhoyfLwGtbm5BsoHdJDIHOgc+AR+RvxyhM9fBTOeW85hH78GjjUPt7IR+2hujhFF538/HON9C+vk0+qvB3vYf4m85C9ol7R5D7/PkfHaa+Uc37r3fATuvbk5H6HtEo6f+NtllfMFCHCZcvHO/vSe4mhs+LO3/9efa7d85wLlLzavtbEBG/5aAMZcLMeIi3XVZB7b+5rWbTy2+HyO3XvnJYf7+N9J59j3mva+hVp5TXg9mi3530dbfow9Gj6n+tvm5lTEyn/3CLz1C5ZLlGb5tf6l7+4TqA4WgSJwcwTa3Nw8BDWgCBSBIlAEikAR2BOBNjd7ollZRaAIFIEiUASKwM0RaHNz8xDUgCJQBIpAESgCRWBPBNrc7IlmZRWBIlAEikARKAI3R6DNzc1DUAOKQBEoAkWgCBSBPRFoc7MnmpVVBIpAESgCRaAI3ByBNjc3D0ENKAJFoAgUgSJQBPZEoM3NnmhWVhEoAkWgCBSBInBzBNrc3DwENaAIFIEiUASKQBHYE4E2N3uiWVlFoAgUgSJQBIrAzRFoc3PzENSAIlAEikARKAJFYE8E2tzsiWZlFYEiUASKQBEoAjdHoM3NzUNQA4pAESgCRaAIFIE9EWhzsyealVUEikARKAJFoAjcHIE2NzcPQQ0oAkWgCBSBIlAE9kSgzc2eaFZWESgCRaAIFIEicHME2tzcPAQ1oAgUgSJQBIpAEdgTgTY3e6JZWUWgCBSBIlAEisDNEWhzc/MQ1IAiUASKQBG4BwS+fPlyD25czIcfP348/fr162LyU3Cbm0Sj70WgCBSBIlAEXoAAjQ2Xd+kwAp8+fbpKg9Pm5nAcXrz67t27J54/f/68WMaWjdfSs8WWPXjAi0NCv75+/bqH2KMy+DWBzg8fPhzlPcaAD9++fXv6/v37MdaT1ve08STFD8LMxfT+/fvnPCAHL127rxVWao78LW1HgNxJzDy/VuO1/nqx8oAGY4+zbiWf+fQ9sWHt0rqf9R8y7i2scRiNBJCfP38ep5ffJNooZ/xebl4s/Pz58+njx4+L1f2mr6VnP4sPS6LovFh+//79XCCHdxxePeVXAnmD7nOIC5G4X+rw2sPGc/zbupfYgf216ZR4p23//PPPc65hNzHk8N3SWM/OjpT7Vt+5rM+thbfq+0vs5iIfCQxnlzhzrF2L0HWp82jlw+g3+GSDc438+ndEVta+wnl+GV+igdhDLjKucThcS881wu9F4S9mL5xzdJ9y2RGvcw8dCjiL+BzbZ3v3sHEmd+85flxsaQ721ksOEfNTLxBszR9E1NW5P3D29u3a8sDj3Hq4ts230AdGs0ae+bzkvQ/I0WviesoZuAd+nNsjHpyJObfH2X7M1lfR3PBriSaF7g4ASAKcl/wV6J+5+GsFhee3++Anmfh2P3uZswnyW6DRBb/JtpILP4edcjn82Le6yGaXEHu0z4OUizxl53de8tqFDdmFz/SI27VG7NQ+/AMT8TzFBi4Y/IHwkZi9RE7qBNvEK9fGd/SZa+BMnp1C4JA5knvxw9iPF2nyHXt/qY3EBNvIWWsBXfrLPPaJFfPwY6s5CU9iYi0xzwMvpB7nVzVyzFfXkYtMHmSpx/XZSJ2SS/i0JYfIk7QTHLD/ECEfni3y4cV+7LJxyjigx5xHJnaDOw98PLyr0zigW7wZ0xZ4zDn88+yCBx2sG0NtS39ZH23M9b7/BwHyhriMBM7gL63yyZiyTozMQ+PkPtZ457FOkZ18ymDeHHaP+8gF3tM2+M0V1pA5yuBbWcg4RPAlJrznt7LMyUOyXrp2uHpfKvWEfThHYVHwEAdXgj6uA6qB5UI1EVTJGvIk5DKnTILmgSEPgaDIpVGuhwCFnvr4zn3uZ0RfrrEPfvzhQSekPXwzjy7tY+RJfEiQtGHUkzYce0fnoSf1HJKFX9iFrdivL4f2zNaILXgw4hd4nEuZL4dkJc68k4fsPYXIM7AYCUyIoXkrXiPfse9jNoL7LGb4Yu5hg36Bb+aWzTd2sEd/jO2Ye8ixbvExdae/x/w6tJ624z8+evAe2ucaPmI3+YStyJgRvqT9+I4uYzbbwxw8WefIEV/3YC88zGOLNZ7y0ZOYYS9z2KTfYIE/yFFn6jdm6B1zjjMN3ewHB+Qyxx5IfdrsiHz2lNYIgCvPSMQd/PKZ8YC9RGzz2zx0nRF55iVj8rOfRyJ2ye88to370gf2WGfKYA7yWxuUeWhk75hHM7sOyTh17ebNDYBSZBJFlyCP6/IxEpwRYIo65cFHkDLgHAweZF4Yx+Sy7oEgb8p0jpEgclBJJgO6ePfAHtf9ZtQH/PcAwjd8ZoRGPbn/Wu8WiZfGWGxpB2uHEjoLAKxX+KbM8Z19yDn0jHv4Flv9eIn+1R5iZhOALvzy4JjZspo7ZiM+Y0OSuedlmGvkVmJsfOTBRvJYTMjJ9INvZLjuPnX6/dJROWl75sgpcpFFHEZ8lAEOuSYWqVteR/zOOmceOYmpvIxjHuALelwjvsgUd2yGxEHe58n//oP+jIlro69jbo6xR87M17RR2R3/RgBswXMkz0bnwXIk4pRnAbKYk8xDvxkzJq6bK2OcV7mTtsmjDHSknNn6aHfaN75TD7PcxY/UOe479/vfaJ8r8cT9gORlzVaSJIt1XFe8gPvtSCNAYJL4tkFgnncPZHTnYbSSyz54/WVOcswCJl/KhJdA8uDPeCEgR7nsT9nwuxeZidVou74lfsxdkvAlscV2G7NRr4U4w405fE3K5MfXldzcM76vCmvkw4fEje9T9eWBoHzzyXxjHj9nF4l7VuNLbCQemVspGztYl0beMbaj3fgERuPFiJzMf+WfOo72gOXYTByTyR78wHbim3HIvdib54a5mjzj+6z+Rh6/0Us+S36TB+YI6/gHppkfIw7KYISP+PLom/IYpVksPUvgW+GKTbN6VW7Hf/94FhPiRt5JxGCkWQ3mnlkezmJC/jLPk7VnLowxTNtmOqgF7VBG5tNo9+iX3/g86mZNmfJdYvxftV1C+gaZGSgKle8s7FxPcauDhSIfwcyDi6COl1geaiu56M6AzBJV+7iEUibvmXDyOeY6vqfslf/sHfUob+uI7ENP+rCSmdhyuJL0ie9q3zjPnvTbgvPAHvm3fmPfmA+zvWPebC3elJVxdF4//MbP1UUiz2p8iY0zm5Q/5hZYkVMS+rZcgPgEr3RuXipntJ1Lf2vDSa2ST1vjCG/WKD6x9xCd4ud4rhy7XFIvejIuueY7uMgz5hw1RM4ZS76JvbWFLStcxxxRX8e/EQCnkTLG45pn3Zif5Hzm3RhL5GRM0JHfY814Z41nYNomD6OUcmbro93uyxEf0SPpM9/M57c8e47/jsie0jfIIjAUHQACKN+8W6gUpQBx4PBA8HqgApKBYT8Ngvvh9QJm7wgoa8iHn2JfydUV7El9zueITHxSJnqxFfk8zGeymUjMsddDB5npP+tp/6gHOfBfk7BBP8Uufd9qixcXvhNL8EI2hN/Gdas8+ZCRWDufIzrNQ97B2NxKvmPv6MmDCX5tRy7vyiZHzeUtvm2xETnEIClzD1zBGVkQtmoDI7ljHanPb2wHS+wmvsjhHVLH88d/682cyEsz7dvis/mkDmQhl3rFnhlp55bDN/ezT5vwnfgnlq7lHvDIs4M15nhGSl/QBdbgCIEj8sUaPI0L69jiGZhymTeWYsM6crSXdXPOvejNPMU2HnRog7xpl3Md/40A2GW+wMFc4uwu8sNznD3Jw3t+G0vznX3ExO+UhXzWxvxLfvWOts3kwANpQ+YGNrquXzmaTzmXdqWPybPn+82bG0AwIB4qWbQAyEHAw8EmcSA4Px4EAOfBCz8BhTf5lINM1gzUSq78BGhMYtccR5n4xT78xLZxP7pZY5+HlbL0n/W8UFgf9ZB8eZko45IjByX4ibk2Jf5b9LPfOOlrYkFOvITA3YNgtZ91YmKMGFP3at84zx6wyEMAHuWCDWvweMi455i+LTaC25hbyFU/GGZcjB374Mk19GGnxBp82m2cmTP28prPzCf22rfVZ7BCBnYgE3uRwfuK8CN1rvhm82CHLnTgn7SyVwzSHvTzjMSc+a0/yYMs9LIGX+YD8xkb9yGTNR7ecw++uDbKYy3PCc5FeNMPdIxNkHo7zhEAZ2tf7FcjvFLGkXlyPok55RAj1o0Xue4aI+S3MnI/c6kPXutFucyZC6P88Xs8b5CfuactjPrMujq18RLjzZubSzh1KZkcHnkoXErPS+VyGM2S7aXyrrUPXEn+FbFOQb4F4pI6JQZcLNco9NeE3VvzeQ97ye9sPl5TPFa2UHNeciuezv+NwNiY/L16/Iuz41wZx7XcjoN8OuV8PMfS9Y1yjtQ73EtHTrG/5gOKi5WL8q1dljRlh5oXm7a34Bf5wV9IsLlUBEDAv3q9JTS4hPyl/Zbsfuu23ntzc834tLnZgDaX1fgn/Q3brs7in8lnf8K+ujEnKKSgsX1FNAr88n0rvyJpcPCJX/ylIkAekL9vpS5P/etjI7wPAvnfR4f+kr2PtvuX0ubm/mNcD4tAESgCRaAIPBQCbW4eKtx1tggUgSJQBIrA/SPQ5ub+Y1wPi0ARKAJFoAg8FAJtbh4q3HW2CBSBIlAEisD9I9Dm5v5jXA+LQBEoAkWgCDwUAm1uHircdbYIFIEiUASKwP0j0Obm/mNcD4tAESgCRaAIPBQCbW4eKtx1tggUgSJQBIrA/SPQ5ub+Y1wPi0ARKAJFoAg8FAJtbh4q3HW2CBSBIlAEisD9I9Dm5v5jXA+LQBEoAkWgCDwUAm1uHircdbYIFIEiUASKwP0j0Obm/mNcD4tAESgCRaAIPBQCbW4eKtx1tggUgSJQBIrA/SPQ5ub+Y1wPi0ARKAJFoAg8FAJtbh4q3HW2CBSBIlAEisD9I9Dm5v5jXA+LQBEoAkWgCDwUAm1uHircdbYIFIEiUASKwP0j0Obm/mNcD4tAESgCRaAIPBQCbW4eKtx1tggUgSJQBC6FwJcvXy4l+i7k/vjx4+nXr19X8aXNzVVgrpIiUASKQBG4ZwRobLi8S4cR+PTp01UanKs3N9++fXv68OHD07t3757ev3//9PXr18NIdHWKAAkChj9//pyu7zWJDp4/f/7sJbJyLoAAdUWcLl1P5+ghhw7lkrLNudkIdFwgrDF+/Pjx+Rz5559/LoDqaSKx595/uZNfxKn0NwLEPnGZ5W7OXeuvF39b+b8v7g/u4UsR9Zj+jnVxSd36dNXmBgc5jDyI3lqhfP/+/a8EFsRbjMcuir1sonkiZqX/ILDKAQr59+/fm2HicEGWtJLr+paROF262cWOl+rBx/GQS784F/jBAzYjsQ+MIXzk7ICPOmD0TBn3Xesb+/JyQy8XGP7cGz1CE3dqzMzN3AdOs0ucOdauRei6djM11jD4pM+8HzoL9sDmas0NzlDob/UvANhNUl7j8tgSWC7SWeFs2XsKz7EL6RRZb533Ujmwl1zq6xqX/Ev1bGmKyDcOwrHOwCgvEGTpK++3JC6O8TC/pT3X0P358+e/Lqtr6HytOrjbZvFnPs9oL3PyhbVrEbZdu7nB72z2sSG/qd2s50tgcbXmZnR2dIbAe2gCAu/5J3YOO2QACEB5sJEkzCVwyHKv6+wnwMhQz2gD3/CzjkxGvmkk+M7HBGUO2dqXBy02cggoKxMM+5SHTf7qn+lf2WmxuM7FIEbohbgUwEuf85t3ebBHv5EjoQObLknggI0SsTSeYMa68ZTn0JiyVnz6i8/oGuUzJx7E9FgOZCyQhc3kBOQ3vljUxAlayWUv+tMu7XneOPyDXGW6tMo/c0zZfos5+2e5xPxMj/oOjfg52rfiN1/Nz5GPeWWBKbku1vIigxhI8GVesJa1Kt9LRuSONUI+oGOcn8mHl9iCLbjzPtqG/cwhE9/BgIc5Ht7VKRbGlT3ITFvgQQ5r2I9uCB7mWDdm2pa2sz7amOuP9E7dZP3rO1iap8yB64yMG+vEwTo0Fu5jjXce8kFKPmWwBo/8jtYB32kb/OYDa8gcZfCtnKylZ8Yj/7DPHJN1NufaHuMc7T0khwwP8NE5WQASHgDjoCIABNhLlkKiwAwoPCQE8uAZk4uiY0+us8ekYd13bXAEcGyBOPzVqQ/yzWTDix6Iwwab9QF71QkfieQBlYmy0q9eR+Qpmzlk4xd28SAHAlv0mEjgwje8jBD6kcc3/GkPdorHM/PwD3IPPfo8bPvrE5yxR+I9beD7kA3uA3vsZzQvXMtRfeAkNuyTcp1YacuYA/Kzbizgh485ZPLOg0ywh8DEPOF7Jhce+I0jfHznvmdh//1Hv53Dr1n+GWPkE3OIvcjGRog1MB9zSd6VDc+bF/8gG7lbSDxWerBV2+XN+KFD7PAXQn9iybcx22LTIR7kqif5mMc+iZwwl5wzR5jHJ+xGFnvJJYgxzz9qkjnwlBd/iBdy1Jn6WVceOlIeOKOb/dYNc+yB1Pf8Ef8gnz2PTmBnPiYWxBaM8sl13uEBX4n45TcxY38S38aSMfnZzyMRn+R3fqY3fWAPPJAytMNvbVDmOGIHe3hmecL8MRmjzFO+/0btlJ0n8FIwGQAc0uk88ODJb1WM8/BkAAmKB5XFnmCyngcle2d60JcXgvoZCXTqdA3Zq3kPBw4TfGCEOHzycFEW40p/8vCOPA8xkw3feUdvNguupwzxwC95kYcv2s0+7Lk0mQ/oQSe+iSk+aesxO+BlH7m1iq9YiB0ys/hm6+pd5cCs+cKHLFzyBPsg8DVf+V7JZQ3btBWfUuazsP/+kzXAFN/Gccw/1xMj7EG2/q9yadSTNhx6J4+QvZWwDd9X/m6Rk/vNC/eRU8bDuZeM2IeekZA91g42mNcjP/mSOZG2s0YMkUmuJJbGa4YTfClTncjL2POedhHj/EaOOagMxrQx5x/tHfzAbCRiBdbSLE9Yh09CVu6Z5Vfi7rq1NcZylR9pmzzKwJaUM1sf7db+1Tjjx4/Uudr70vl/V+VLJR3Yl0DJNha6AI4HjvMJQh7cyAM4i8+DXD2MXD5Z/HzPih5e5LDOk7agEz9GQrdNS64xT/B48HXk4ZtDYyyKlf6UjV3IlUhUdaEXmWk7vuOPBL94wOteeNJHbMxDzv17jxYocsVZe7FHW1kfG4O0hQsL37CZfemLfKwrmznyijhI47rzjNqWc2MsXANTY4COtAX96dNMrnLgNXfgW9Eo81j+jTmpbPw3H2a5NOrBHuasv5l9xmS2tpoDO/Sfk3/4Ac7oB39jwlzGA3y3NtCjvchC7kin1M6YQ36DqecfOshT7EysD+Xr7CxRHqNE7G2EmQN3cw6+rA/3MIpvzj3i+yq/mQdLyRrzm5F1+KRxzyy/ZrhTJ8zzZM0Yb+QkpZ6ZDupD25WROTPanbJn76mPdWXOePea+3dV7iU55ABUAs4S4GSDQTFxSI40Ak/hU2xZfPf4L3QAAAVuSURBVARUQk8eVB4UrgtqBsq1HJGRBT871JXNONIsAUce9oGDvuT6qD/X4E88Z/gmf65z4GWRIYf1GeH/ak1+C2o1HtuPHHEkJtiW3xkDdR4bE5uRN7FgDZwzX8b13D/LgTEW8GfO4gvyGSF9OybXdfDAJg6H8YCSZybzWP7lOjq8MA/5P9OjDYdG/Mf+Uwhd1MbKZ3KFOK/W0WW8zAe/M96n2DTjzVjn+pbakX/MobwIVvLdi55jNYK/8ozywDnPU2PMCGHLCq/MIe151BEsRso4jmuewWOTQP0xJ43xYj5xR0d+j/XrfYecpLRNHkYp5czWR7vdx6jN6JBSH3N8i4E8e4//jsjeGv7rCAXkAUpTQ0D8RiXFN3NWYOGl4ODhkJIEUhkUIoEBPPa6zl6+2WuhK8ORNYs6DwTWsZdDCDvcz3cmonIY8dfgYoO+sd/DAl3sRw50SH/Kxj9sQD6ywdO9yGSNeclEZY6DXh9ZxxZtAx++XYcX25DnnDL3HsEXH7AB4jttYQ4/wPUcQgZ6IHBDxwrLEY9ZDiAP/IwFcsEZXvaDYeY5a8ScOZv7mVx9TBudG0dk4lPKXOWfe9Wp7c4fyqWZHvwDwxWBAbacmj/IBNsZIRO88RmbVoRtWV988+Qe3sEiY7SSt5o31rmOHmzM2lF/8vGeOYk94OWZgF0pn/iYN+wFA/SMxLyYm+PwgJ3+sk784ZXGMw3beNDB3qS0K+cf8R18xnxlbnY/kAeeuexJHt7z23iZs+wDd79TFrizxlxS8qt3tG0mBx5IGzL+2Oh66uId29TjGvyJT/ooz97jVZobjKbIAZnCxcksKNYBN4s2HYWfvTyAZtHCA+DI9ECgOOETePZS3IDJPHasCBvUw3vqMamYN8jKnslDPzYhD/0enoz4rp6055D+1KGP7sVO945JxD5s0ff0iTXsERvs8lBlDfn4IJZpw97v2JDxN56pB9zB8hxCBrr0Syz1MbEED+OGzlkOuN9YwIcO8M5c0WbW0M2asZjJld/LTl7nc5zJXOWf+8CAx0PS+fSfdXJcWunJuMnryBr+nULYPp4Ps/1gONqffOjNfOE94yTvFl3yzkbsMH9cn9UOfDwjMYet5IV5mTzIIp9Ygy9zgfnMUfchkzXzMPccOk/HM434IWP0j7wnP0r/Q4DYUCOQ2K9GeKWMFfMjrswphziwbkzIf9cYU7fycz9zqY891pBylQ3vKH/8zvNBfYyZY8hLPt7VmXv2fr9ac7O34VvlEci8sLfuK9/rQ4A4ZpG8Pgv3t+haB8FLLefiPXRQcWDOLt+VPi4H9nhJJB/zSdT2Id3Ju3q3mVutb5knL7HlpcThn83HS+Vccx/+jg3PNfW/Vl1jjp5qJ/V+roxTdV6Tn5y51hl+980Nv3bOPQCvGfzqWiPgRfoo8eRX87UOgjXqh1f4qwcX86yBIU6nHtTIm12aNBA0AUnwnvvDxYb53Jzi1zH5eSqB2+jXqTKuzU988Le0PwL33tzsj9ha4t+nxZrvTa54cBz6s/mbdOxBjebyoFmdXaT3BIkX+WtvbMCcBsMGZ4wBF+AptYe/XPSHHnUkzzl/NRHrWUOlrq3jSy59/9vnreQ0NfgW8nJrzF4TX/730VtreF8Tjtpy182NTnYsAkWgCBSBIlAEHgeBNjePE+t6WgSKQBEoAkXgIRBoc/MQYa6TRaAIFIEiUAQeB4E2N48T63paBIpAESgCReAhEGhz8xBhrpNFoAgUgSJQBB4HgTY3jxPreloEikARKAJF4CEQaHPzEGGuk0WgCBSBIlAEHgeBNjePE+t6WgSKQBEoAkXgIRBoc/MQYa6TRaAIFIEiUAQeB4H/A78ynb3pdbARAAAAAElFTkSuQmCC\" width=\"567\" height=\"112\"\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eFigure 2 is an example of the distribution of the activity durations measured by the wearable sensors before and after the intervention. Figure 2(a) represents the individuals who achieved the goal of advice category (I), showing a decrease in the scale exponent β from 0.87 to 0.75 that indicates a less steep slope post-intervention. Figure 2(b) represents the individuals who did not achieve any of the advice categories (I through V), showing virtually no change in the value of β.\u003c/p\u003e\n\u003cp\u003e[Insert Figure 2]\u003c/p\u003e\n\u003cp\u003eNext, we classified\u0026nbsp;the participants as either “achievers” who had followed at least one piece of advice in each of the five categories and achieved the goal in that category or “non-achievers” who did not follow any of the advice. We then calculated the stress level for each of these groups. The application conditions for the group stress level in the previous study\u003csup\u003e6\u003c/sup\u003e included the requirements that the group not be divided into clusters within the face-to-face network and that the data be from the same phase. After confirming that these conditions had been met, we performed the evaluation.\u003c/p\u003e\n\u003cp\u003eFigure 3 presents a comparison of the group stress levels before and after the intervention for the achiever and non-achiever groups. The results indicate that, in the category of conversational partners (I), the achiever group showed a more positive trend of change than the non-achiever group, specifically, with p\u0026lt;0.05 for Organization 1, p\u0026lt;0.1 for Organization 2, and p\u0026lt;0.05 for Organization 3. We observed no significant differences in categories (II) through (V).\u003c/p\u003e\n\u003cp\u003e[Insert Figure 3]\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003e\u003cstrong\u003eImpact of the interventions in workplace behavior on group stress levels\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study involved an empirical experiment conducted to verify that the proposed interventions in workplace behaviors have the potential to reduce group stress levels. We observed a significant reduction in the stress level of the group consisting of the participants who followed the advice provided to them. We provided five categories of advice related to measurable and intervenable workplace behaviors, specifically, conversational partners, methods of conversation, desk work practices, arrival and departure times, and time management Among these categories, the interventions that encouraged conversations with individuals proved effective. Notably, the conversations between pairs of individuals within the group reduced their group stress level. There is extensive empirical evidence that stress is influenced by interpersonal relationships\u003csup\u003e1-5\u003c/sup\u003e, this study is the first to suggest that interventions in interpersonal relationships can be effective in reducing group stress levels.\u003c/p\u003e\n\u003cp\u003eAll of the pieces of advice used in this experiment were formulated based on behavioral characteristics that correlate with activity levels. As discussed in the Results section, the implementation of the advice relating to conversational partners (I) contributed to changes in group stress levels while the implementation of the advice relating to categories (II) to (V) was not associated with significant changes in group stress levels. This difference in the effectiveness of the advice is likely attributable to differing causal relationships. Thus, the behaviors related to desk work and time management did not seem to influence the group stress level directly but, rather, seemed to be correlated owing to a third factor affecting both behavior and the group stress level. The interventions that we conducted in this study allowed us to examine the causal relationship between behavioral characteristics and group stress levels, and we found that the advice for categories (II) to (V) was less effective than the advice for category (I).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRelationship Between Group Stress and Network Structure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur findings suggest that the presence and extent of communication between the individuals who received the advice and specific others with whom they formed pairs of employees within their organizations affected the group stress level. Previous research has demonstrated that the structure\u0026nbsp;(topology) of interpersonal connections within an organization, conceptualized as networks, influences human stress. An example is Christakis’s theory of social contagion\u003csup\u003e7,8\u003c/sup\u003e. Generally, “contagion” is a term used in the context of infectious disease to describe the transmission of material substances such as viruses directly among individuals. According to Christakis, in the context of the structure of human connections revealed by epidemiological data, a similar phenomenon occurs with subjective states such as obesity, loneliness, and happiness in the absence of any material intermediary. These states spread among the individuals in a network in a process that he termed “social contagion”. Christakis and his colleagues also conducted research on depression using the same CES-D employed in the present study and demonstrated that the structure of connections influences the risk of depression\u003csup\u003e9\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eFurther, Lee et al. analyzed the network structure of employees using face-to-face communication data from Japanese workplaces obtained with name-tag wearable sensors similar to those used in this study as well as the results of the CES-D\u003csup\u003e4\u003c/sup\u003e. The findings revealed that the employees with a higher clustering coefficient, which is a measure of the degree to which an individual’s immediate contacts are themselves connected, tended to have lower stress levels. This result suggests that individuals within a dense network in the workplace experience lower stress because information transmission is smoother, and they incur lower costs for reconciling differences in perspectives. Therefore, it is plausible that enhancing or increasing the connections between specific pairs could reduce stress for third parties in topological proximity, consistent with the findings of this study and the research by the teams of Christakis and Lee.\u003c/p\u003e\n\u003cp\u003eIn addition, the research described here provides new insights in terms of the intervention conducted and the finding that altering the structure of connections can reduce the overall stress risk within a group. The communication partners were selected mechanically based on their statistical association with activity levels. An intriguing question for future research is how these specific pairs are positioned within the broader organizational network from a macro perspective and the influence of the designation of the pairs on the clustering coefficient, as the research by Lee et al. suggests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLimitations and Directions for Future Study\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn this study, a reduction in the group stress level was observed only when those who followed the advices were considered as a virtual group while no significant difference was observed \u0026nbsp;a whole of the organizations including those who did not followed adovicesit. This result suggests that, while effective outcomes could be achieved when all of the members of a group followed the advice in category (I), variation in the type of advice and the frequency with which it was followed by individuals diluted the overall effect. To reduce the stress level reliably across the entire group, it is necessary to reconsider the motivation for following the advice and explore ways to provide advice that make it easier for individuals to follow it. Moreover, this study is based on three organizations in a single corporation in Japan and is not a randomized controlled trial, so future research could involve randomized controlled trials to assess the effectiveness of interventions more rigorously and evaluate the approach in organizations with diverse backgrounds.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study verified that group stress levels can be decreased through interventions in workplace behaviors. In particular, the interventions encouraging conversations between specific individuals were effective. While it is well known empirically that interpersonal relationships influence stress, this study is the first to suggest experimentally that interventions focusing on the network structure of connections can be effective in reducing the group stress level.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eData Collection and Group Stress Level\u003c/strong\u003es\u003c/p\u003e\n\u003cp\u003eIn prior research, we proposed a method for estimating group stress levels based on physical movements\u003csup\u003e6\u003c/sup\u003e. To collect the data on group stress levels and workplace behavior for the present study, we used name-tag wearable sensors similar to those used in our previous study\u003csup\u003e6\u003c/sup\u003e. As Figure 4 shows, the wearable sensors were equipped with accelerometers to capture physical movements and infrared transceivers to detect face-to-face communication between individuals\u003csup\u003e10,11\u003c/sup\u003e. Thus, we were able to capture the participants’ authentic movement data during office work without interfering with their activities.\u003c/p\u003e\n\u003cp\u003eWe used the method developed in our previous research to calculate group stress levels\u003csup\u003e6\u003c/sup\u003e. It has been demonstrated that\u0026nbsp;the cumulative distribution of the activity duration fits a stretched exponential distribution\u003csup\u003e12\u003c/sup\u003e, as Equation 1 represents, and, in the present study\u003csup\u003e\u0026nbsp;6\u003c/sup\u003e, the linear sum of the scale exponent (\u003cem\u003eb\u003c/em\u003ei) for the group, the group stress level, closely approximated the average stress level of the same group obtained using a CES-D questionnaire (Equations 2 and 3). Equations 2 and 3 indicate that the individual βi values reflect the individual i’s impact on the group stress level. In other words, to reduce the group stress level, it is necessary to decrease\u0026nbsp;\u003cem\u003eb\u003c/em\u003e (that is, make the distribution slope less steep). Additionally, we employed the constants a and b in Equation 2 from our previous study\u003csup\u003e\u0026nbsp;6\u003c/sup\u003e in this research.\u003c/p\u003e\n\u003cp\u003e[Insert Figure 4]\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStu\u003c/strong\u003e\u003cstrong\u003edy Design\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study consisted of two experimental phases. During the first, the data were collected from the participants’ name-tag wearable sensors as they conducted their usual work without receiving any feedback. During the second, intervention phase, the participants received personalized advice based on the statistical processing of data from the first phase through a smartphone app. The data collection was performed during the second phase in the same manner as the first. In evaluating the results, we assessed the effect of the intervention on the group stress level by determining whether the participants followed the advice during the second phase and assessing any changes in the group stress level across both periods.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHypotheses for the Five Categories\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe hypothesized that behavioral changes could reduce group stress levels. We categorized the behavioral characteristics that we identified as either (I) conversational partners, (II) methods of conversation, (III) desk work practices, (IV) arrival and departure times, or (V) time management. We defined the characteristics that could be calculated based on the sensor data for each category. Table 1 provides examples of the corresponding advice.\u003c/p\u003e\n\u003cp\u003eOur first hypothesis concerns conversational partners (I). We considered the possibility that conversing with specific individuals could influence the overall stress level of the group. For example, the amount of social support from supervisors or colleagues is related to workplace stress\u003csup\u003e1-3,5,13\u003c/sup\u003e. Since interactions between individuals in a group, even if they are unconscious, can affect group stress, we reasoned that communication is a critical aspect of workplace stress. The characteristic for this category, denoted Characteristic C, was calculated using Sato's method\u003csup\u003e14\u003c/sup\u003e of counting the number of face-to-face interactions between any two individuals.\u003c/p\u003e\n\u003cp\u003eOur second hypothesis concerns changing the methods of conversation (II), differing from (I) in providing advice that does not specify a conversational partner. Since enriching workplace communication is associated with productivity and engagement\u003csup\u003e\u0026nbsp;15-18\u003c/sup\u003e, the frequency of greetings among employees and the number of meetings also have the potential to affect the workplace culture and the risk of stress. Additionally, the movements of the participants in conversations have been used to assess the quality of communication\u003csup\u003e14,19\u003c/sup\u003e. Specifically, when the bodies of both parties are moving, the conversation is considered interactive; when only one participant is moving, that individual is the speaker while the less active participant is the listener\u003csup\u003e14\u003c/sup\u003e. Furthermore, the levels of stress in organizations in which conversational roles are not fixed and all members are equally empowered to speak appear to be lower than in organizations that do not thus empower their members\u003csup\u003e14\u003c/sup\u003e. Based on these considerations, we reasoned that providing advice to change the frequency of greetings, the duration of a single meeting, and the conversational roles in meetings (interactive, speaker, listener) can impact the group stress level, defining characteristics (D1–D7) in Table 1.\u003c/p\u003e\n\u003cp\u003eThe third hypothesis concerns desk work practices (III). It has been shown that time is required to refocus on the original task following interruptions in the form of speech and other distractions during desk work\u003csup\u003e20\u003c/sup\u003e. Conversely, the ability to maintain concentration in the absence of external interruptions for extended periods is believed to enhance cognitive productivity. Additionally, the demonstration that moving the body at the appropriate frequency influences motivation\u003csup\u003e21\u003c/sup\u003e suggests that there may be an optimal duration for continuous desk work. We used Sato’s definition of the behavioral characteristics of office workers\u003csup\u003e14\u003c/sup\u003e, considering them engaged in desk work when they were not facing someone and had minimal body movement (frequencies of less than 2Hz). The feature (E1–E4) in Table 1 is identified by counting the occurrences of this state across different durations.\u003c/p\u003e\n\u003cp\u003eThe fourth hypothesis concerns arrival and departure times (IV). Some people, referred to colloquially as “morning larks,” tend to be alert and productive in the early hours, while others, “night owls,” achieve peak performance during the evening or at night, though it remains unclear which if either behavioral pattern is more productive. Therefore, we included arrival and departure times in our hypothesis and prepared the features (F1–F2) in Table 1.\u003c/p\u003e\n\u003cp\u003eOur last hypothesis concerns time management (V) in relation to the scheduling of conversations and desk work. Productivity peaks in the early afternoon\u003csup\u003e22\u003c/sup\u003e while short-term memory tasks are more effectively performed in the morning\u003csup\u003e23\u003c/sup\u003e, possibly because certain cognitive processes are influenced by the circadian arousal level. Thus, since the scheduling of members can impact the group stress level, we defined time-specific features (G1-G8) in Table 1.\u003c/p\u003e\n\u003cp\u003e[Insert Table 1]\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAdvice Derivation Procedure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePersonalized advice based on the sensor data collected during the first phase of the research (which lasted four weeks)\u0026nbsp;was provided to each participant through a smartphone app during the second phase. For individual i, we calculated the activity level i (Equation 2) and the characteristics (Cij to G8i). In computing the correlation coefficients, the activity level served as the dependent variable and the characteristics served as the independent variables. Within each category, the advice corresponding to the characteristic with the highest absolute value of the correlation coefficient was displayed on the app. The advice text was predetermined, but, when the correlation coefficient was positive, the phrase “increase ~” was chosen; when the coefficient was negative, “decrease ~” was selected. In this way, we identified behaviors that are characteristically more (or less) prevalent on the days in which the individuals’ activity levels were high and intervened to reinforce these behaviors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIntervention App\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSmartphones are considered an effective intervention method for individual stress in the context of digital improvement programs because employees can easily access the programs. Ease of access facilitates early preventative intervention better than traditional methods such as consultation with a physician. Additionally, digital platforms allow employees to engage with the chosen content at their own pace, further fostering active participation and increasing the likelihood of desirable behavioral change\u003csup\u003e24\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eOur design insights for the intervention application include presenting the intervention program after self-monitoring to gain an objective view of an individual’s state\u003csup\u003e24\u003c/sup\u003e, providing concrete advice such as meeting people, walking, or taking a three-minute breathing break\u003csup\u003e25\u003c/sup\u003e, and incorporating gamification based on achievement points\u003csup\u003e26\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eIn the intervention phase of this study, the prior findings were expanded to reduce group stress, and the intervention UI(user interface) was designed as a smartphone app. The app included features for daily activity log feedback for self-monitoring, the presentation of personalized concrete behavioral advice, and the display of achievement results as scores. Figure 5 shows the interface of the intervention appl, which provides users with the opportunity and motivation to consider the appropriate actions to take and when to do so during their work activities.\u003c/p\u003e\n\u003cp\u003eThe top left of\u0026nbsp;Figure 5(a) displays the activity level, and\u0026nbsp;Figure 5(c) shows the previous day’s activity log collected by the wearable sensor, thus encouraging users to self-monitor their past behaviors and states and motivating improvement. Additionally,\u0026nbsp;Figure 5(b) presents the text of the advice and the specific target values. The application automatically determines whether the advice has been followed based on the sensor data and displays the number of pieces of advice followedas scores on the bar graph at the top of\u0026nbsp;Figure 5(a). Tapping on the advice in\u0026nbsp;Figure 5(a) opens a detailed screen in\u0026nbsp;Figure 5(b), which includes a verbal description of the advice and presentation of the conditions related to the user, such as the time slots and durations defined in the condition categorization shown in Table 1.\u003c/p\u003e\n\u003cp\u003e[Insert Figure 5]\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eParticipants and Settings\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAn empirical experiment was conducted using the intervention app and name-tag wearable sensors discussed previously. The participants were, as mentioned, employees with the sales departments of three business units within a Japanese manufacturing company. The experiment took place over six weeks in June and July 2016. There was no remote work; all of the employees were in the office. The collection of the sensor data continued \u0026nbsp;over the six weeks, with the first four weeks of data serving for analysis and the intervention app being \u0026nbsp;provided to the participants during the fifth and sixth weeks. A total of 245 individuals from the three organizations participated in the experiment. Specifically, Organization 1 had 168 participants with 120 valid , Organization 2 had 125 with 69 valid, and Organization 3 had 79 with 56 valid, with the participants who collected a sufficient amount of data being considered valid. The determination was as follows. Since our previous research6 showed that at least 1,000 counts (1,000 minutes = approximately 16.6 hours) of accelerometer data per individual are required to calculate the group stress level for a given period, in the present study, only the data from the individuals who had more than 1,000 counts during the third to fourth weeks and fifth to sixth weeks, when the participants were wearing the sensors, were considered for the evaluation of group stress levels.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEvaluation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFor the relevant behavioral characteristics, an increase over the average of past data served to indicate that the advice had been followed. The group stress level during the third to fourth weeks, before the intervention, served as the baseline for assessing the changes observed during the fifth to sixth weeks.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical Approval and Informed Consent\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn this study, which involved collecting workplace behavioral data and delivering advice via participants' smartphones, it was necessary to obtain participants' names, email addresses, departmental affiliations, and behavioral data through wearable sensors. Prior to the commencement of the experiment, the researchers provided a clear explanation of the study's purpose and the nature of the data to be collected. Informed consent was obtained from all participants, ensuring their understanding and voluntary agreement to participate.\u003c/p\u003e\n\u003cp\u003eFurthermore, the study was conducted under the review and approval of the Hitachi Group's Ethics Review Committee. This committee verified that the research aims and methodologies were justified and rational, and that the rights and dignity of the subjects and human groups involved were protected. All experimental protocols were thus executed following these ethical standards.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003cp\u003eThis study was conducted with research funding provided by Hitachi, Ltd.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eS.T. designed the study, analyzed the data, and wrote the manuscript. N.S. was involved in the data collection and provided feedback. K.Y. and Y.M. designed and supervised the study. All of the authors contributed to the revision of the manuscript and approved the final manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets used and/or analysed during the current study available from the corresponding author on reasonable request.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eAdditional information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was conducted with research funding provided by\u0026nbsp;Hitachi, Ltd.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCod\u003c/strong\u003e\u003cstrong\u003ee\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;availability\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Python code that supports the findings of this study is available from the authors upon request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eTheorell, T. \u003cem\u003eet al\u003c/em\u003e. A systematic review including meta-analysis of work environment and depressive symptoms, \u003cem\u003eBMC Public Health\u003c/em\u003e, \u003cstrong\u003e738\u003c/strong\u003e (2015).\u003c/li\u003e\n\u003cli\u003eInceoglu, I. et al. Leadership behavior and employee well-being: an integrated review and a future research agenda,\u003cem\u003e Leadership Q.\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 179\u0026ndash;202 (2017).\u003c/li\u003e\n\u003cli\u003eNappo, N. Job stress and interpersonal relationships cross country evidence from the EU15: a correlation analysis, \u003cem\u003eBMC Pub. Health\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 1143 (2020).\u003c/li\u003e\n\u003cli\u003eLee, J. H., Sato, N., Yano, K. \u0026amp; Miyake, Y. Universal association between depressive symptoms and social‑network structures in the workplace. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 10170 (2022).\u003c/li\u003e\n\u003cli\u003eKalish, Y., Luria, G., Toker, S. \u0026amp; Westman, M. Till stress do us part: on the interplay between perceived stress and communication network dynamics. \u003cem\u003eJ. Appl. Psychol.\u003c/em\u003e \u003cstrong\u003e100\u003c/strong\u003e, 1737\u0026ndash;1751 (2015).\u003c/li\u003e\n\u003cli\u003eTsuji, S., Sato, N., Ara, K. \u0026amp; Yano, K. Estimating group stress level by measuring body motion \u003cem\u003eFront. Psychol.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 634722 (2021).\u003c/li\u003e\n\u003cli\u003eChristakis, N. A. \u0026amp; Fowler, J. A. Social contagion theory: examining dynamic social networks and human\u0026thinsp;behavior. \u003cem\u003eStatics in Med.\u003c/em\u003e \u003cstrong\u003e32\u003c/strong\u003e, 556\u0026ndash;577 (2012).\u003c/li\u003e\n\u003cli\u003eChristakis, N. A. \u003cem\u003eConnected: the surprising power of our social networks and how they shape our lives\u003c/em\u003e (Little, Brown and Company, 2009).\u003c/li\u003e\n\u003cli\u003eRosenquist, J., Fowler, J. \u0026amp; Christakis, N. Social network determinants of depression. \u003cem\u003eMol. Psychiatry\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 273\u0026ndash;281 (2011).\u003c/li\u003e\n\u003cli\u003eWakisaka, Y., et al. Beam-scan sensor node: reliable sensing of human interactions in organization in\u003cem\u003e 2009 Sixth International Conference on Networked Sensing Systems (INSS)\u003c/em\u003e, Pittsburgh, PA, USA, 1-4 (2009).\u003c/li\u003e\n\u003cli\u003eHayakawa, M., Ohkubo, N. \u0026amp; Wakisaka, Y., Business microscope: development of a practical human behavior measurement system. \u003cem\u003eIEICE Transactions on Communications,\u003c/em\u003e \u003cstrong\u003eJ96-D\u003c/strong\u003e, 2359\u0026ndash;2370 (2013) (in Japanese).\u003c/li\u003e\n\u003cli\u003eNakamura, T., Kiyono, K., Yoshiuchi, K., Nakahara, R., Struzik, Z. R. \u0026amp; Yamamoto, Y. Universal scaling law in human behavioral organization. \u003cem\u003ePhys. Rev. Lett.\u003c/em\u003e \u003cstrong\u003e99\u003c/strong\u003e, 138103 (2007).\u003c/li\u003e\n\u003cli\u003eMahan, P. L., et al. Work environment stressors, social support, anxiety, and depression among secondary school teachers. \u003cem\u003eAAOHN J.\u003c/em\u003e \u003cstrong\u003e58\u003c/strong\u003e, 197\u0026ndash;205 (2010).\u003c/li\u003e\n\u003cli\u003eSato, N., et al. Knowledge-creating behavior index for improving knowledge workers\u0026apos; productivity in\u003cem\u003e 2009 Sixth International Conference on Networked Sensing Systems (INSS)\u003c/em\u003e,\u003cem\u003e \u003c/em\u003ePittsburgh, PA, USA, 1-4(2009).\u003c/li\u003e\n\u003cli\u003eDitta, A., et al. Analyzing the relationship between productivity and human communication in an organizational setting. \u003cem\u003ePLoS ONE\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, e0250301 (2021).\u003c/li\u003e\n\u003cli\u003eSantalla-Banderali, Z. \u0026amp; Alvarado, J. M. Incidence of leader-member exchange quality, communication satisfaction, and employee work engagement on self-evaluated work performance. \u003cem\u003eInt. J. Environ Res. Pub. Health\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 8761 (2022).\u003c/li\u003e\n\u003cli\u003eBakker, A. B. \u0026amp; Xanthopoulou, D. The crossover of daily work engagement: test of an actor-partner interdependence model. \u003cem\u003eJ. Appl. Psychol.\u003c/em\u003e\u003cstrong\u003e 94\u003c/strong\u003e, 1562\u0026ndash;1571 (2009).\u003c/li\u003e\n\u003cli\u003eZoonen, W. V. \u0026amp; Sivunen, A. E. The impact of remote work and mediated communication frequency on isolation and psychological distress. \u003cem\u003eEur. J. Work Organ. Psychol.\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 610\u0026ndash;621 (2021).\u003c/li\u003e\n\u003cli\u003eHigo, N., et al. Interpersonal similarity between body movements in face-to-face communication in daily life.\u003cem\u003e PLoS ONE\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, e102019 (2014).\u003c/li\u003e\n\u003cli\u003eMark, G., Gudith, D. \u0026amp; Klocke, U. The cost of interrupted work: more speed and stress. in \u003cem\u003eProc. SIGCHI Conf. Hum. Factors Comput. Syst. \u003c/em\u003e(\u003cem\u003eCHI \u0026apos;08)\u003c/em\u003e 107\u0026ndash;110. (2008). \u003c/li\u003e\n\u003cli\u003eSignorini, G., et al. Enhancing motivation and psychological wellbeing in the workplace through conscious physical activity: suggestions from a qualitative study examining workers\u0026apos; experience. \u003cem\u003eFront. Psychol.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e,1006876 (2022).\u003c/li\u003e\n\u003cli\u003eGaggero, A. \u0026amp; Tommasi, D. Time of day and high-stake cognitive assessments. \u003cem\u003eEcon. J.\u003c/em\u003e \u003cstrong\u003e133\u003c/strong\u003e, 1407\u0026ndash;1429 (2023).\u003c/li\u003e\n\u003cli\u003eFolkard, S. Diurnal variation in logical reasoning. \u003cem\u003eBr. J. Psychol.\u003c/em\u003e \u003cstrong\u003e66\u003c/strong\u003e, 1\u0026ndash;8 (1975).\u003c/li\u003e\n\u003cli\u003eWeber, S., Lorenz, C. \u0026amp; Hemmings, N. Improving stress and positive mental health at work via an app-based intervention: a large-scale multi-center randomized control trial. \u003cem\u003eFront. Psychol.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 2745 (2019).\u003c/li\u003e\n\u003cli\u003eMorrison, L. G., et al. The effect of timing and frequency of push notifications on usage of a smartphone-based stress management intervention: an exploratory trial. \u003cem\u003ePLoS ONE\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, e0169162 (2017).\u003c/li\u003e\n\u003cli\u003eParks, A. C., Williams, A. L., Tugade, M. M., Hokes, K. E., Honomichl, R. D. \u0026amp; Zilca, R. D. Testing a scalable web and smartphone based intervention to improve depression, anxiety, and resilience: a randomized controlled trial.\u003cem\u003e Int. J. Wellbeing\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 22\u0026ndash;67 (2018).\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Table 1","content":"\u003cp\u003e\u003cstrong\u003eTable 1.\u0026nbsp;\u003c/strong\u003eCorrespondence Between Advice Categories and Behavioral Characteristics.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eFor each individual i, the daily aggregated behavioral characteristics were calculated, and the median value for each individual during the data collection period served as the baseline for determining the achievement of the goal associated with the advice.\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"586\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"35.21367521367522%\" colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eBehavioral Characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.7008547008547%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eDefinitions\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.085470085470085%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eExample Advice Statements\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e(I) conversational partners\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eC\u003csub\u003ei,j\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eNumber of Conversations with Individual (j)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003ecount of conversations between individual i and individual j\u003csup\u003ei\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Increase the frequency of conversations with Mr./Ms. J.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e(II) methods of conversation\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eD1\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eShort Conversations Count\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003enumber of conversations ending in less than 5 minutes\u003csup\u003eii\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Increase/Decrease the number of short conversations and greetings within 5 minutes.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eD2\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eMid-short Conversations Count\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003enumber of conversations lasting between 5 to less than 15 minutes\u003csup\u003eii\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Increase/Decrease the number of 5-15 minute stand-up chats and consultations.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eD3\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eMid-long Conversations Count\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003enumber of conversations lasting between 15 and less than 30 minutes\u003csup\u003eii\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Increase/Decrease the number of conversations that are less than 30 minutes.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eD4\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eLong Conversations Count\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003enumber of conversations lasting more than 30 minutes\u003csup\u003eii\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Increase/Decrease the number of lengthy conversations more than 30 minutes.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eD5\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eBi-directional Conversation Time\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003etotal time in face-to-face state with any other individual during which both the individual and at least one other person are moving\u003csup\u003eiii\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Increase/Decrease the time you and another person are actively engaging in conversation.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eD6\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eSpeaker\u0026rsquo;s Conversation Time\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003etotal time in face-to-face state with any other individual during which the individual is moving and no one else\u003csup\u003eiii\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Increase/Decrease the time you spend talking as the speaker in conversations.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eD7\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eListener\u0026rsquo;s Conversation Time\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003etotal time in a face-to-face state with any other individual during which the individual is not moving but at least one other individual is\u003csup\u003eiii\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Increase/Decrease the time you spend listening as the listener in conversations.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e(III) desk work practices\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eE1\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eShort Desk Work Count\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003enumber of desk work sessions that ended in less than 5 minutes\u003csup\u003eiv\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Frequently get up and walk around, and try to complete desk work within 5 minutes.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eE2i\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eMid-short Desk Work Count\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003enumber of desk work sessions lasting between 5 and less than 15 minutes\u003csup\u003eiv\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Increase/Decrease the number of desk work tasks that can be completed within 5-15 minutes.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eE3i\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eMid-long Desk Work Count\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003enumber of desk work sessions lasting between 15 and less than 30 minutes\u003csup\u003eiv\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Increase/Decrease the number of desk work tasks that can be completed within 15-30 minutes.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eE4i\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eLong Desk Work Count\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003enumber of desk work sessions lasting more than 30 minutes\u003csup\u003eiv\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Try to concentrate on desk work for more than 30 minutes continuously.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e(IV) arrival and departure times\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eF1\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eArrival Time\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003etime of the first accelerometer Data entry of the day\u003csup\u003ev\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Start your workday earlier/later.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eF2\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eDeparture Time\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003etime of the last accelerometer data entry of the day\u003csup\u003ev\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Finish your workday earlier/later.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e(V) time management\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eG1\u003csub\u003ei\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eConversation Time (Before Work Hours)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003etotal time detected in face-to-face interaction with any individual between 6 am and 9 am\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Try to increase/decrease conversations before starting work hours.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eG2i\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eConversation Time (am)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003etotal time detected in face-to-face interaction with any individual between 9 am to 12 pm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Have more/fewer conversations during the morning hours.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eG3i\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eConversation Time (pm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003etotal time detected in face-to-face interaction with any individual between 12 pm to 5 pm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Engage in more/fewer conversations during the afternoon hours.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eG4i\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eConversation Time (After Work Hours)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003etotal time detected in face-to-face interaction with any individual between 5 pm to 10 pm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Increase/decrease conversations after work hours end.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eG5i\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eDesk Work Time (Before Work Hours)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003etotal time not in face-to-face interaction and no movement detected between 6 am to 9 am\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Do more/less desk work before starting work hours.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eG6i\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eDesk Work Time (am)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003etotal time not in face-to-face interaction and no movement detected between 9 am to 12 pm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Conduct more/less desk work during the morning hours.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eG7i\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eDesk Work Time (pm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003etotal time not in face-to-face interaction and no movement detected between 12 pm to 5 pm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Increase/decrease the amount of desk work in the afternoon.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"9.215017064846416%\" valign=\"top\"\u003e\n \u003cp\u003eG8i\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"26.109215017064848%\" valign=\"top\"\u003e\n \u003cp\u003eDesk Work Time (After Work Hours)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"34.641638225255974%\" valign=\"top\"\u003e\n \u003cp\u003eTotal time not in face-to-face interaction and no movement detected between 5 pm to 10 pm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"30.034129692832764%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ldquo;Do more/less desk work after work hours end.\u0026rdquo;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003ei) \u003cstrong\u003eCount of Conversations Between Two Individuals\u003c/strong\u003e: Count each period in which person i and person j are detected in a continuous face-to-face state as one conversation.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003eii) \u003cstrong\u003eConversation Count\u003c/strong\u003e: One comversation is \u0026nbsp;a continuous face-to-face state with two or more people of the same member composition\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003eiii)\u003cstrong\u003e\u0026nbsp;Movement Determination\u003c/strong\u003e: When the frequency of acceleration due to body movement is above 2Hz, it is considered \u0026ldquo;moving\u0026rdquo;; below 2Hz is considered \u0026ldquo;not moving.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003eiv) \u003cstrong\u003eDesk Wor\u003c/strong\u003ek: A state that is not face-to-face and lacks movement (the acceleration frequency is below 2Hz) is determined to be desk work.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" colspan=\"4\" valign=\"top\"\u003e\n \u003cp\u003ev) \u003cstrong\u003eArrival and Departure Times\u003c/strong\u003e: Since the name tag wearable sensor does not monitor acceleration sensing while connected to the charger, the first and last timestamps of the acceleration data are considered as the arrival and departure times.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4240426/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4240426/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eReducing workplace stress has become a societal challenge. Traditionally, the measurement of stress and intervention methods for it have focused on individuals. In our prior research, by contrast, we conceptualized stress as a collective phenomenon arising from group dynamics and proposed a method to estimate group stress levels quantitatively. This method involves equipping individuals within a group with accelerometers and calculating the scale exponent of the activity duration distribution of their physical movements. However, the potential for reducing group stress levels remained unverified. In this study, we hypothesized that behavioral change could effectively reduce group stress levels and intervened in workplace behaviors by offering advice through a smartphone app. The interventions targeted five categories of behaviors: (I) conversational partners, (II) methods of conversation, (III) desk work practices, (IV) arrival and departure times, and (V) time management. The results showed that changing behaviors in categories (II), (III), (IV), and (V) had no effect, but intervention with conversational partners, that is, increasing communication between certain pairs, reduced group stress levels. This finding suggests that it is possible to mitigate group stress, particularly through interventions in workplace interpersonal relationships.\u003c/p\u003e","manuscriptTitle":"Behavioral Interventions in Face-to-face Communication to Decrease Group Stress Levels","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-04-29 13:13:05","doi":"10.21203/rs.3.rs-4240426/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"57372e6c-a471-46bd-9512-3e1732990712","owner":[],"postedDate":"April 29th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":31019943,"name":"Biological sciences/Psychology"},{"id":31019944,"name":"Biological sciences/Psychology/Human behaviour"}],"tags":[],"updatedAt":"2024-07-23T06:42:37+00:00","versionOfRecord":[],"versionCreatedAt":"2024-04-29 13:13:05","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4240426","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4240426","identity":"rs-4240426","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}