A randomized controlled trial on the digital socio-emotional competence training Zirkus Empathico for preschoolers

Socio-emotional competence (SEC) is crucial for preschoolers' social and emotional development, impacting their ability to adapt to social situations and assert their needs. Key components of SEC include empathy (cognitive and affective), emotion recognition, and prosocial behavior. Deficiencies in SEC can contribute to mental health issues later in life. Traditional classroom-based SEC training programs, while effective, face challenges in individualization, cost, and infrastructure. The COVID-19 pandemic highlighted the need for home-based interventions. Digital training offers a viable solution, providing individualized, engaging learning environments. Existing digital interventions primarily focus on children with neurodevelopmental conditions. *Zirkus Empathico*, a touchscreen application initially developed for children with autism, addresses SEC through modules on emotion awareness, facial expression recognition, understanding emotion-eliciting contexts, and prosocial responses. A previous RCT showed moderate effects on empathy and emotion recognition in autistic children. This study aimed to assess *Zirkus Empathico*'s effectiveness on typically developing preschoolers, focusing on empathy (primary outcome) and emotion recognition, prosocial behavior, and behavioral problems (secondary outcomes). Neurobiological changes in event-related potentials (ERPs) related to facial expression processing were also examined. The study hypothesized that *Zirkus Empathico* would improve empathy and other SEC aspects, accompanied by changes in ERP amplitudes reflecting enhanced processing of facial expressions.

The introduction thoroughly reviews existing literature on socio-emotional competence (SEC) in preschoolers. It discusses the multifaceted nature of empathy, encompassing both cognitive and affective components, and its link to emotion recognition and prosocial behavior. The literature establishes the importance of early intervention for mental health, with SEC as a protective factor. It highlights the limitations of traditional classroom-based SEC training programs, such as difficulty in individualization and resource constraints, contrasting them with the potential of digital interventions. The literature review emphasizes the scarcity of studies on digital SEC training for typically developing preschoolers and the limited investigation of neurobiological correlates in evaluating treatment efficacy. Finally, the introduction specifically cites relevant prior work on the *Zirkus Empathico* application and other digital empathy training initiatives.

This randomized controlled trial (RCT) involved 74 Central European preschoolers (5.1 years old, 34 females) randomly assigned to either the *Zirkus Empathico* training group (n=36) or an active control group (n=38) receiving a digital foreign language acquisition training. SEC was assessed at three time points: pre-training (T1), immediately post-training (T2), and at a 3-month follow-up (T3). The primary outcome was empathy, measured using the Griffith Empathy Measure (GEM) and the Inventory to survey of emotional competences for 3- to 6-year-olds (EMK 3-6). Secondary outcomes included emotion recognition, prosocial behavior (EMK 3-6 and Strengths and Difficulties Questionnaire (SDQ)), and behavioral problems (SDQ). Neural sensitivity to facial expressions was assessed using EEG, measuring P1, N170, and P3 components in response to happy, angry, and neutral faces. The *Zirkus Empathico* application comprises modules focusing on emotion awareness, facial expression recognition, understanding emotion-eliciting contexts, and prosocial responses. The control application was a digital foreign language acquisition program. Data analysis included intention-to-treat (ITT) and per-protocol (PP) analyses of covariance (ANCOVA) for behavioral measures and analyses of variance (ANOVA) for ERP data. Exploratory analyses examined follow-up effects and correlations between behavioral and neural measures. The study followed CONSORT guidelines and data were made openly available.

Significant findings emerged from this study. Directly after the six-week training (T2), the *Zirkus Empathico* group demonstrated significant improvements compared to the control group in several key areas: * **Empathy:** Increases were noted in the EMK 3-6 parent ratings (d = 0.28), but not in the GEM parent ratings. * **Emotion Recognition:** Child EMK 3-6 assessments revealed significant gains in emotion recognition (d = 0.57). * **Prosocial Behavior:** SDQ parent ratings indicated significant increases in prosocial behavior (d = 0.51) and reductions in problem behaviors (d = 0.54) for the *Zirkus Empathico* group. * **Neural Sensitivity (EEG):** The *Zirkus Empathico* group exhibited larger P3 amplitudes to happy vs. neutral and angry facial expressions, indicating increased sensitivity to positive emotions. The three-month follow-up (T3) showed a continued increase in empathy (GEM parent ratings) for the *Zirkus Empathico* group. However, other improvements were not consistently maintained at the long-term follow up. Training fidelity was high for both groups, with similar training engagement and parental involvement.

The results of this RCT support the effectiveness of the *Zirkus Empathico* digital SEC training for preschoolers. Improvements in empathy, emotion recognition, and prosocial behavior align with previous research on classroom-based SEC programs. The observed increase in P3 amplitudes for happy faces suggests enhanced attention towards positive emotions and potentially improved emotion regulation. The longer-term effects, though less consistent across all measures, suggest the potential for lasting benefits, especially for empathy. The study's high training fidelity indicates the feasibility of implementing *Zirkus Empathico* in home settings. The findings suggest that digital platforms are a viable and accessible method for enhancing SEC in young children, offering flexibility and potential reach beyond the constraints of traditional classroom settings. The findings suggest the potential for improved emotion regulation as a mechanism for training-induced effects. More research is needed to understand the long-term effects and optimal levels of parental involvement.

This study demonstrates the effectiveness of the *Zirkus Empathico* digital SEC training in improving brain and behavioral measures of socio-emotional skills in preschoolers. The intervention's accessibility and engagement potential make it a promising tool for preventing mental health issues. Future research should explore the optimal duration of training, the role of parental involvement, and the underlying neural mechanisms of *Zirkus Empathico*'s effects. Longer term studies are needed to assess the sustainability of effects.

This study has several limitations. The reliance on parental evaluations introduces potential rater bias, although efforts were made to mitigate this. The GEM measure showed lower internal consistency at baseline, suggesting a need for alternative empathy measures in future studies. EEG measurements were taken only post-training, limiting the capacity to fully map the neural trajectory of changes. The sample size, while sufficient for power analysis based on behavioral data, may not have been large enough to detect all meaningful effects in the ERP analysis. The relatively homogenous sample of middle-to-upper-class families may also limit generalizability. Finally, correlations between behavioral and neural data were not significant, potentially due to insufficient power.