The Edu:Social School Study 2 is a randomized controlled trial designed to evaluate the effects of an 8-week socio-emotional, partner-based empathy-compassion Dyad training program (EmCo) compared with a mindfulness-based training (MBT) and a waitlist control group (WLC) in school teachers. The primary objective is to assess the efficacy of the adapted 8-week online EmCo Dyad training program across multiple domains of teacher functioning. Seven domains capture individual-level outcomes: (1) mental health, (2) resilience, (3) attention and cognitive processes, (4) social emotions, (5) stress coping and emotion regulation, (6) listening skills, and (7) voice-based emotional expression (DYVA). Two additional domains assess broader contextual outcomes: (8) classroom teaching and (9) organizational/school level outcomes, reflecting how individual-level changes may translate into perceived changes in the teaching and school context. To this end, participants will be randomized into one of three groups: (1) an 8-week EmCo Dyad intervention (EmCo), (2) an 8-week mindfulness-based training program (MBT), or (3) a waitlist control group (WLC) that will receive the EmCo intervention after completion of the initial post-test. Both interventions are delivered online via a dedicated study application to ensure scalability and accessibility. A second aim is to validate the Egocentric Social Network Analysis Paradigm (e-SNAP) as a measure of teachers' perceived social interactions and network structures. This study represents the first randomized controlled trial in teachers to examine whether changes in internal psychological processes are associated with perceived changes in the perception of the quality of interaction between school teachers and their students in the classroom as well as between school teachers and their colleagues in school. A third aim is to introduce and validate objective measures of emotional processing based on AI-based voice analysis in educational field research. Specifically, the study will evaluate two tasks using different parameters of AI-based emotional voice analysis: (a) the Teacher Voice Assessment (TEAVA), in which teachers' voices are recorded during a teaching-related task, and (b) the Dyad Voice Assessment (DYVA), conducted during the daily partner-based Dyad practice. In both tasks, non-semantic vocal features (e.g., pitch, loudness, speaking rate, intonation) and vocal emotional expressions (e.g., arousal, valence, dominance, and basic emotional categories) will be analyzed using audEERING devAIce software. No speech content will be transcribed or analyzed. A final aim is to investigate the cognitive and affective mechanisms underlying intervention-related changes. Based on prior research, the EmCo Dyad training is expected to enhance care- and affiliation-related motivational systems associated with positive affect, social emotions and connectedness, and prosocial behavior, while reducing loneliness, social stress, and other forms of psychological vulnerability. In contrast, mindfulness-based training is expected to primarily strengthen attention- and thought-related processes, including present-moment awareness, calming the mind and executive attentional control. These mechanisms are expected to contribute to improvements in mental health and overall well-being across both intervention groups.
The teaching profession is widely recognized as highly demanding, and teachers report elevated rates of stress and burnout. If left unaddressed, burnout adversely affects teachers' health, job performance, teaching quality, and school turnover rates. Moreover, teacher burnout has important consequences for students. Teachers' emotional exhaustion has been shown to negatively affect students' academic achievement and self-concept in Germany. Given teachers' pivotal role in shaping children's socio-emotional development, addressing psychological vulnerability-including stress, depression, anxiety, loneliness, and exhaustion-is essential. Strengthening resilience enables teachers to maintain well-being and effective classroom functioning in the face of adversity by supporting adaptive stress management. In addition, strong interpersonal skills such as empathy, compassion and high-quality listening play an important role in the teacher-student communication. Addressing psychological vulnerability and strengthening resilience and social skills, is essential not only for teacher well-being but also for the functioning of educational systems broadly. In recent years, mindfulness-based interventions focusing on individual mental practices have been expanded to include partner-based social practices (Dyads), which specifically target social skills such as empathy, (self-)compassion, social connectedness and different types of listening. However, dyadic interventions, that is daily mental practices done with a partner, have not yet been systematically investigated in educational settings. Existing evidence suggests that contemplative Dyads may be more effective than solitary mindfulness practices in reducing loneliness and social stress, enhancing social connection, and strengthening resilience. Moreover, dyadic practices appear to engage mechanisms that differ from those underlying traditional mindfulness-based interventions. Empathy, (self-)compassion, and high-quality listening are core socio-emotional skills in school contexts. However, empathy may lead to empathic distress when individuals are repeatedly exposed to others' suffering, whereas compassion is considered more protective due to its association with positive affect and prosocial motivation. To address this distinction, the Edu:Social School Study 2 implements an empathy-compassion Dyad training program (EmCo), which trains participants in empathic listening and emotion regulation during the first four weeks, followed by (self-)compassion and compassionate listening during the subsequent four weeks. The EmCo program builds on the Affect Dyad developed in the ReSource project and the online Dyad 10-week intervention program with weekly online coachings and daily app-based Dyads implemented in the CovSocial project. These two large-scale mental training studies, have shown that different types of practices produce distinct effects on psychological and biological outcomes. The ReSource project, a 9-month longitudinal study using behavioral, hormonal, autonomic and neuroimaging (fMRI) measures, compared three training modules, including a mindfulness-based "Presence" module, a socio-affective "Affect" module (including the Affect Dyad) and a socio-cognitive "Perspective" module, and demonstrated that training content differentially affects mind, behavior and brain. Building on this work, the CovSocial project implemented a 10-week fully online design comparing a low-dose purely online socio-emotional Dyad training with a mindfulness-based active control training group. Results showed both shared and distinct effects: while both interventions improved mental health outcomes such as depression, anxiety, emotion regulation, and compassion, the Dyad training showed stronger effects on resilience, positivity bias, social connectedness, and reductions in loneliness and social stress. Further, mediation analyses suggest that individual mindfulness practices and partner-based Dyadic practices influence well-being through partly distinct psychological mechanisms, whereas care-based motivational processes could account for Dyad-practice effects and thought-related process for mindfulness-practice effects. Compared with these prior interventions, the present Edu:Social School intervention program introduces a novel empathy versus compassion listening component and reduces the duration from 10 to 8 weeks, improving feasibility within applied school settings. In an initial Edu:Social School pilot study, 120 teachers participated in a 10-week socio-emotional training program (EmCo). The pilot demonstrated high compliance and provided preliminary evidence for reductions in psychological vulnerability. However, the study lacked an active control group and a reliable waitlist control group, limiting interpretation. In addition, the 10-week duration was difficult to implement within the German school calendar. The present study is a randomized controlled trial in a sample of school teachers (target N = 900). A multimethod assessment approach will be used, including validated trait and state self-report questionnaires, behavioral phone-based tasks administered via an app application, and ecological momentary assessment (EMA) delivered through smartphone notifications. These different assessments are capturing changes across the nine domains mentioned before. After providing informed consent, participants will be randomized to one of three groups: (1) an empathy-compassion training group (EmCo), based on the Affect Dyad daily mental practice; (2) a mindfulness-based training group, based on attention-based mindfulness practices; or (3) a waitlist control group (WLC). All participants will first complete a pre-test assessment phase, including baseline psychometric measures, phone-based tasks, and ecological momentary assessment (EMA). Participants in the intervention groups will then be onboarded through two group-specific online sessions (Onboarding I and II) led by trained instructors, followed by participation in their respective 8-week programs delivered via a dedicated web and smartphone application. These programs include daily partner-based Dyad practices or mindfulness meditations, respectively, as well as weekly 1.5-hour online coaching sessions with expert mental training teachers. Throughout the intervention period, participants will complete brief weekly self-report questionnaires, EMA, and daily pre-post Dyad practice ratings (DPR) to assess changes in key psychological and social processes, which will be examined as potential mediators. At the end of the intervention, all participants will complete a post-test assessment phase comparable to the pre-test (T0) and mid-test (TM), with the addition of a final feedback questionnaire. Following completion of the initial post-test (T1), participants in the waitlist control group will subsequently receive the same 8-week socio-emotional Dyad intervention (EmCo) and complete an additional mid-tests (TM1) and post-test assessment (T2). After both training groups (EmCo and MBT) have completed their respective interventions, participants will be offered the opportunity to continue their practice via the study app. Participants in the EmCo group will be invited to continue the Dyad practice, whereas participants in the MBT group may choose to continue either the mindfulness-based practice or switch to the EmCo Dyad intervention. This phase of voluntary continued practice will take place in parallel with the 8-week EmCo Dyad training offered to the WLC following post-test. The study was preregistered on the Open Science Framework (OSF) and is publicly available at https://osf.io/x59pk
Study Type
INTERVENTIONAL
Allocation
RANDOMIZED
Purpose
BASIC_SCIENCE
Masking
DOUBLE
Enrollment
900
1. Participants engage in a structured 13-minute partner-based contemplative exercise. Each dyad reflects on two experiences from the previous 24 hours: one involving a difficult emotion and one involving gratitude. Partners take turns speaking while the other listens non-judgmentally. During weeks 1-4, the practice emphasizes empathic listening; during weeks 5-8, compassionate listening. Participants are instructed to attend to bodily sensations associated with the emotions described. The practice aims to improve coping with difficult emotions, empathic and compassionate listening, (self)acceptance, compassion, gratitude, resilience. 2. Participants also attend eight 1.5-hour online group sessions led by Expert Dyad teachers. The coaching sessions help deepen the Dyad practice and educate teachers about body language, coping better with difficult emotions/stress, the benefits of empathy versus compassion and the act of listening from a mindset of empathy versus compassion.
1. In this intervention, participants will practice 13-minute basic attention-based mindfulness meditation such as the Breathing Meditation (BM). BM is a 13-minute individual exercise that requires participants to focus their attention on the sensations of breathing. Participants have to sustain their attention to breath for long stretches of time, and have to return their attention to their breath when their mind wanders. The key focus is on training attention and interoceptive body awareness. Other practices participants will be taught is mindfulness on sounds (here the object of attention is not the breath but sounds in the environment) as well as open presence meditation. 2. Participants also attend eight 1.5-hour online group sessions led by Expert mindfulness teachers. The coaching sessions help deepen the mindfulness practice and educate teachers about bodily sensations, attention, rumination.
1. Initially, participants in the WLC group will not receive the intervention. They will complete pre- and post-test procedures consisting primarily of self-report questionnaires, and behavioral tasks, as well as ecological momentary assessment (EMA) conducted on four days within two weeks at pre-test and post-test 1 \& 2. 2. The WLC will be offered the EmCo training only after both intervention groups have completed their trainings
Depression Anxiety Stress Scale (DASS-21)
A scale measuring depression, anxiety, and stress (Henry \& Crawford, 2005; Nilges \& Essau, 2021). Higher scores indicate more depression, anxiety, and stress.
Time frame: Assessed at baseline (pre-test) and after the 8-week intervention period (post-test 1 & 2)
Maslach Burnout Inventory-Educators Survey (MBI-ES)
A scale measuring burnout (Maslach \& Jackson, 1981; Schwarzer et al., 2000). Higher scores indicate more burnout.
Time frame: Assessed at baseline (pre-test) and after the 8-week intervention period (post-test 1 & 2)
UCLA Loneliness Scale (UCLA)
A scale measuring loneliness severity (Döring \& Bortz, 1993; Russell et al., 1980). Higher scores indicate more loneliness.
Time frame: Assessed at baseline (pre-test) and after the 8-week intervention period (post-test 1 & 2)
Connor Davidson Resilience Scale (CD-RISC)
A scale measuring psychological resilience (Connor \& Davidson, 2003; Sarubin et al., 2015). Higher scores indicate more resilience.
Time frame: Assessed at baseline (pre-test) and after the 8-week intervention period (post-test 1 & 2)
Types of Positive Affect Scale (TPAS)
A scale measuring the frequency and intensity of three distinct types of positive emotions: Activated PA (excitement, energy), Relaxed PA (calm, peacefulness), and Safe/Contentment PA (safety, security) (Gilbert et al., 2008). Higher scores indicate a greater frequency and intensity of experiencing specific types of positive emotions.
Time frame: Assessed at baseline (pre-test) and after the 8-week intervention period (post-test 1 & 2)
Executive control (Flanker Task; FT)
This task assesses executive control with three indices reaction times (RT), error rates, and the flanker effect (FE) (Trautwein et al., 2016; 2020). Faster RT, lower error rates and FE indicate higher cognitive control.
Time frame: Assessed at baseline (pre-test) and after the 8-week intervention period (post-test 1 & 2)
Mind-Wandering
These two questions embedded in the flanker task assess mind-wandering during attentional performance (Smallwood, 2015). Lower rates indicate less mind-wandering.
Time frame: Assessed at baseline (pre-test) and after the 8-week intervention period (post-test 1 & 2)
Socio-Affective Video Task (SoVT)
This task assesses behavioral empathy and compassion using emotional video clips (Klimecki et al., 2014, 2013). Higher scores indicate more empathy or more compassion.
Time frame: Assessed at baseline (pre-test), after 4 weeks of empathic listening training (mid-intervention) and after the 4 weeks of compassionate listening training (post-test 1 & 2)
Sussex-Oxford Compassion Scale for Self and Others (SOCS)
A scale measuring self-compassion (SOCS-S) and compassion for others (SOCS-O; Gu et al., 2020). Higher scores indicate more compassion.
Time frame: Assessed at baseline (pre-test), after 4 weeks of empathic listening training (mid-intervention) and after the 4 weeks of compassionate listening training (post-test 1 & 2)
Prosodic Feature Pitch
Acoustic assessment of the prosodic speech feature pitch (measured in Hz) during participants' daily Dyad practice, analyzed using audEERING devAIce software.
Time frame: Assessed weekly from week 1 to week 8, as part of the Dyad Voice Assessment (DYVA)
Prosodic Feature Loudness
Acoustic assessment of the prosodic speech feature loudness (unitless) during participants' daily Dyad practice, analyzed using audEERING devAIce software.
Time frame: Assessed weekly from week 1 to week 8, as part of the Dyad Voice Assessment (DYVA)
Prosodic Feature Speaking Rate
Acoustic assessment of the prosodic speech feature speaking rate (measured in syllables per minute) during participants' daily Dyad practice, analyzed using audEERING devAIce software.
Time frame: Assessed weekly from week 1 to week 8, as part of the Dyad Voice Assessment (DYVA)
Prosodic Feature Intonation
Acoustic assessment of the prosodic speech feature intonation (unitless) during participants' daily Dyad practice, analyzed using audEERING devAIce software.
Time frame: Assessed weekly from week 1 to week 8, as part of the Dyad Voice Assessment (DYVA)
Affect Dimensions of Vocalized Emotional Expressions
Assessment of continuous affective dimensions of vocalized emotional expressions during participants' daily Dyad practice using audEERING devAIce software. The following parameters will be assessed: arousal, valence, and dominance (each ranging from -1 to 1). Based on arousal-valence scores, the following affect-quadrant values will be calculated: high-arousal-high-valence; low-arousal-high-valence; low-arousal-low-valence; and high-arousal-low valence.
Time frame: Assessed weekly from week 1 to week 8, as part of the Dyad Voice Assessment (DYVA)
Affect Categories of Vocalized Emotional Expressions
Classification of vocalized emotional expressions intro affect categories during participants' daily Dyad practice, analyzed using audEERING devAIce software. The following parameters will be assessed: angry, happy, and sad, expressed as unitless values ranging from 0 to 1 representing category likelihood.
Time frame: Assessed weekly from week 1 to week 8, as part of the Dyad Voice Assessment (DYVA)
Stress intensity
Custom item based on the Stress Appraisal Measure (SAM; Delahaye et al., 2015; Peacock \& Wong, 1990) measuring stress intensity. Higher scores indicate more intense stress.
Time frame: Assessed using an EMA design with five push-notification measurements per day, distributed across five 3-hour intervals, on four days within a two-week period, at pre-test (Baseline) and after the 8-week intervention period (post-intervention).
Coping strategies
Custom items based on the Brief-COPE (Carver, 1997; Knoll et al., 2005) and Cognitive Emotion Regulation Questionnaire (CERQ; Garnefski et al., 2001; Loch et al., 2011) measuring Coping Strategies (Acceptance, Positive Reinterpretation, Social Support, Rumination, Self-Blame, Distraction). Higher scores indicate a higher use of the specified coping strategies.
Time frame: Assessed using an EMA design with five push-notification measurements per day, distributed across five 3-hour intervals, on four days within a two-week period, at pre-test (Baseline) and after the 8-week intervention period (post-intervention).
Active Empathic Listening Scale (AELS)
A scale measuring active empathic listening (Bodie, 2011). Higher scores indicate more active empathic listening.
Time frame: Assessed at baseline (pre-test), after 4 weeks of empathic listening training (mid-intervention) and after the 4 weeks of compassionate listening training (post-test 1 & 2)
Copenhagen Psychosocial Questionnaire (COPSOQ)
A scale measuring psychological working conditions (stressors and coping resources) impact employee health and well-being (Kristensen et al., 2005; Nübling, 2005)
Time frame: Assessed at baseline (pre-test) and after the 8-week intervention period (post-test 1 & 2)
Egocentric Social Network Analysis Paradigm (e-SNAP)
The e-SNAP is a novel assessment paradigm based on the Affect Grid. Teachers rate their interactions with both students and colleagues in terms of frequency and pleasantness in interactions. In addition, teachers rate in both the student and colleague networks the relationships among each other, allowing to assess positive relations and network density. Higher scores indicate more frequent and more positively experienced interactions, as well as more cohesive and positively connected social networks.
Time frame: Assessed at baseline (pre-test) and after the 8-week intervention period (post-test 1 & 2)
Positive Affect (Affect Grid) (explanatory mechanism)
Assessment of emotional state (valence) and arousal (Russell et al., 1989). Higher scores on valence and arousal indicate more positive affect and higher arousal.
Time frame: Assessed weekly during the course of 8 weeks of intervention
Emotion Acceptance (EAQ) (explanatory mechanism)
A self-report questionnaire measuring emotional awareness, and acceptance of emotions (Beblo et al., 2011; Kisley et al., 2025). Higher scores indicate greater emotion awareness and acceptance.
Time frame: Assessed weekly during the course of 8 weeks
Gratitude Questionnaire-6 (GQ-5-G) (explanatory mechanism)
A scale measuring gratitude (Hudecek et al., 2021; McCullough et al., 2002). Higher scores indicate more gratitude.
Time frame: Assessed weekly during the course of 8 weeks
Self-Kindness Scale (SCS-SF) (explanatory mechanism)
A self-report questionnaire measuring self-kindness and compassionate attitudes toward oneself (Hupfeld \& Ruffieux, 2011; Raes et al., 2011). Higher scores indicate greater self-kindness.
Time frame: Assessed weekly during the course of 8 weeks
Short Loneliness Scale (SLS) (explanatory mechanism)
A short scale measuring frequency, intensity, and duration of loneliness (Hughes et al., 2004; Qualter et al., 2021). Higher scores indicate more loneliness.
Time frame: Assessed weekly during the course of 8 weeks
Perceived Stress (PSS-4) (explanatory mechanism)
A self-report questionnaire measuring the degree to which situations in one's life are appraised as stressful (Cohen et al., 1983; Klein et al., 2016). Higher scores indicate greater perceived stress.
Time frame: Assessed weekly during the course of 8 weeks
Depression (PHQ-2) (explanatory mechanism)
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.
A brief self-report screening measure assessing core depressive symptoms, including depressed mood and anhedonia (Kroenke et al., 2003). Higher scores indicate greater depressive symptom severity.
Time frame: Assessed weekly during the course of 8 weeks
Empathic Concern & Distress (IRI) (explanatory mechanism)
A scale measuring different facets of social emotions, including personal distress and empathic concern (Davis, 1980; Paulus, 2009). Higher scores indicate higher personal distress or empathic concern.
Time frame: Assessed weekly during the course of 8 weeks
Coping strategies (explanatory mechanism)
Custom items based on the Brief-COPE (Carver, 1997; Knoll et al., 2005) and Cognitive Emotion Regulation Questionnaire (CERQ; Garnefski et al., 2001; Loch et al., 2011) measuring Coping Strategies (Acceptance, Positive Reinterpretation, Social Support, Rumination, Self-Blame, Distraction). Higher scores indicate a higher use of the specified coping strategies.
Time frame: Assessed weekly during the course of 8 weeks
Positive Interpretation Bias (ERT) (explanatory mechanism)
This task assesses the tendency to judge persons' facial expressions more positively using morphed sequences of facial expressions (DeBruine \& Jones, 2017; Griffiths et al., 2015). Higher scores indicate a stronger positive interpretation bias.
Time frame: Assessed at baseline (pre-test) and after the 8-week intervention period (post-test 1 & 2)
In-group-Out-group bias (ERT) (explanatory mechanism)
This task assesses the tendency to judge persons from one's own group to be more positive in facial emotion recognition using morphed sequences of facial expressions (DeBruine \& Jones, 2017; Griffiths et al., 2015). Higher scores indicate a stronger in-group-out-group bias.
Time frame: Assessed at baseline (pre-test) and after the 8-week intervention period (post-test 1 & 2)
Weekly Mind-Wandering (explanatory mechanism)
These two questions assess mind-wandering in day-to-day life (Smallwood et al., 2013). Lower rates indicate less mind-wandering.
Time frame: Assessed weekly during the course of 8 weeks
Weekly Thought Patterns (Cube of Thought) (explanatory mechanism)
These three questions assess polarities of thought orientation or quality (self-other, past-future, pleasant-unpleasant) (Petzold et al. 2023). Higher scores indicate a higher level in each respective dimension.
Time frame: Assessed weekly during the course of 8 weeks
DPR-Affect (Affect Grid; explanatory mechanism)
Assessment of emotional state (valence) and arousal (Russell et al., 1989) just before starting the Dyad. Higher scores on valence and arousal indicate more positive affect and higher arousal.
Time frame: Assessed for 8 weeks during intervention period, before the daily exercise
DPR-involvement (explanatory mechanism)
Assessment of listening involvement (1 custom item; only post-Dyad exercise). Higher scores indicate more listening involvement.
Time frame: Assessed for 8 weeks during intervention period, only in the intervention group, after the daily exercise.
DPR-Emotions
Emotion intensities (of e.g., happiness, gratitude, sadness, anger) rated by the speaker and by the listener of a Dyad directly after the Dyad.
Time frame: Assessed weekly from week 1 to week 8, as part of the Dyad Voice Assessment (DYVA)
DPR-Listening-Affect
Custom items (self-generated) measuring the self-rated affective state during listening to the Dyad partner's telling of the difficult situation and the event that they are grateful for. Higher scores indicate a more positive affect.
Time frame: Assessed weekly from week 1 to week 8, after the Dyad practice.
Dyad closeness - Inclusion of Other in Self Scale (explanatory mechanism)
Assessment of how close participants felt to their Dyad partner (post-Dyad exercise). Higher scores indicate more closeness (Aron et al., 1992; Kinnunen \& Windmann, 2013).
Time frame: Assessed for 8 weeks during intervention period, only in the intervention group, after the daily exercise.
Dyad listening quality (DPR- listening-quality)
Custom items (self-generated) measuring the quality of listening to the Dyad partner. Higher scores indicate a higher degree of active, attentive listening.
Time frame: Assessed weekly from week 1 to week 8, after the Dyad practice.
Dyad empathic and compassionate listening skills (DPR-listening-skills)
Custom items (self-generated) measuring the self-rated skill of listening empathically and compassionately to the Dyad partner's telling of the difficult situation and the event that they are grateful for. Higher scores indicate a higher degree of empathic or compassionate listening respectively.
Time frame: Assessed weekly from week 1 to week 8, after the Dyad practice.