Actionable companion · Assessment & reasoning
Learning with agency: Do Shy Preschoolers Interact Differently When Learning Language With a Social Robot? An Analysis of Interactional Behavior and Word Learning
AI-assisted editorial implementation guide, separate from the original research. Proposed actions are not claims or recommendations attributed to the source authors. No independent peer review of this guide is asserted.
Evidence and context
Title and abstract screening; stage labels describe reported settings, not verified numerical ages. Design labels require full-text appraisal. Reported stage(s): Early childhood · preschool, Children · stage unspecified. Verify the participant population and task before transferring the findings.
Original authors’ abstract, reproduced under the linked source license:
Temperamental traits can decisively influence how children enter into social interaction with their environment. Yet, in the field of child–robot interaction, little is known about how individual differences such as shyness impact on how children interact with social robots in educational settings. The present study systematically assessed the temperament of 28 preschool children aged 4–5 years in order to investigate the role of shyness within a dyadic child–robot interaction. Over the course of four consecutive sessions, we observed how shy compared to nonshy children interacted with a social robot during a word-learning educational setting and how shyness influenced children’s learning outcomes. Overall, results suggested that shy children not only interacted differently with a robot compared to nonshy children, but also changed their behavior over the course of the sessions. Critically, shy children interacted less expressively with the robot in general. With regard to children’s language learning outcomes, shy children scored lower on an initial posttest, but were able to close this gap on a later test, resulting in all children retrieving the learned words on a similar level. When intertest learning gain was considered, regression analyses even confirmed a positive predictive role of shyness on language learning gains. Findings are discussed with regard to the role of shyness in educational settings with social robots and the implications for future interaction design.
What kind of evidence is this?
Embodied learning agents · Comparative / pre-post study reported
Stage: Early childhood · preschool; Children · stage unspecified. Screening labels describe reported context and study-design language. They are not a risk-of-bias rating or proof of effectiveness for an age band.
Participant ages · source passage
The present study systematically assessed the temperament of 28 preschool children aged 4–5 years in order to investigate the role of shyness within a dyadic child–robot interaction.
Study methods · source passage
Thirty preschool children participated in the study. The data from two children were excluded because they did not participate in all sessions. This left 28 children (11 female, 17 male) aged 4.00–5.83 years (mean age = 4.98, SD = 0.48) for the final analysis. The children and their parents came from relatively high socioeconomic status backgrounds and were recruited from the wider Paderborn region (North Rhine-Westphalia, Germany). Recruitment was conducted in local kindergartens and libraries or through newspapers and our database of families willing to participate in our research studies. In addition, we assessed the level of parental education. None of the children or the parents had ever seen the robot before the experiment. Prior to their children’s participation, parents provided written consent and filled out a questionnaire on their child’s general and language development. Criteria for inclusion in the sample were: a) age between 4 and 6 years; b) normal general development such as normal sensory and cognitive skills; and 3) no developmental language delays. Additionally, detailed information regarding children’s language skills was obtained by measuring their receptive language abilities with a subtest of the standardized SETK 3–5 (Grimm et al., 2001) and their expressive vocabulary with the AWST-R (Kiese-Himmel, 2005). Parents were present during all interactions, but did not participate actively in the interaction. Children also provided verbal assent prior to taking part in the interaction, and the interaction could be discontinued at any time at no disadvantage to the child. Moreover, each child received stickers and a toy to thank them for their participation.
Limitations and uncertainty · source passage
In conclusion, the results presented here offer new input for not only research with social robots in educational settings and the design of future learner–robot interactions but also for evaluating and measuring children’s achieved learning outcomes. In fact, most past research in child–robot interaction has tested hypotheses by comparing average effects across the sample but ignored that effects may vary across individuals depending on existing intrinsic factors such as shyness. In this vein, the present study provides evidence that shy children do indeed demonstrate a distinctive behavior in terms of their interactions with a robot as well as their language learning. Our findings show that it is important to include the learner’s temperamental characteristics, such as shyness, during child–robot interactions to inform the use of robots in the educational field. In this regard, current research strongly suggests the need to consider children’s temperament in everyday practice in institutional settings such as kindergarten or school and provide a supportive climate for a variety of children and temperament types (Ann Sealy et al., 2021). Whereas approaches to automatically assessing a child’s personality based on predefined behavioral traits have made substantial progress (Abe et al., 2017; Abe et al., 2020; Sano et al., 2020), the present study has taken first steps to determine which behaviors can actually be observed in shy children and how they develop in the long term over a period of several sessions.
Design the system before the pilot
Compare the embodied interaction with an equivalent human-led or screen-based activity. Separate the effects of novelty and social presence from learning. Review physical access, sensory needs and adult supervision.
- Use common rubrics, blind scoring where feasible and a fresh unaided reasoning task. Document permissible assistance and avoid treating AI-detector scores as proof of misconduct.
A 90-day district pilot
- Days 1–15 · Appraise. Have the curriculum lead and a teacher read the methods and limitations. Record the study design, sample, comparator, outcomes and evidence gaps. Define a local question that the study can help inform.
- Days 16–30 · Design. Use a brief adult-mediated story, turn-taking or language activity. An educator controls every interaction; the agent never becomes an unsupervised companion. Preserve hands-on play and human conversation. Name an accountable owner and set the baseline, success threshold and budget before implementation.
- Days 31–60 · Test. Start with a small voluntary cohort, preserve an accessible non-AI route and require teacher review. Use synthetic or de-identified examples; approve data handling and tool terms through district processes.
- Days 61–90 · Decide. Measure independent naming, explanation or turn-taking in a human-led activity after the tool is removed. Record frustration, overstimulation and the amount of adult help. Compare with baseline, document uncertainty and stop or revise if learning worsens, access is unequal or critical errors persist.
The board decision brief
Report the learning problem, evidence quality, participating population, total cost including review time, measured benefit, errors, access differences and remaining uncertainties. Make a dated recommendation to stop, revise or expand. A positive perception survey alone does not establish learning improvement.
Attribution and reuse
Nils F. Tolksdorf; Franziska E. Viertel; Katharina J. Rohlfing. Do Shy Preschoolers Interact Differently When Learning Language With a Social Robot? An Analysis of Interactional Behavior and Word Learning. Frontiers in Robotics and AI (2021-05-31). https://doi.org/10.3389/frobt.2021.676123
Companion prepared 2026-10-30. Original authorship and DOI are preserved. The abstract is an attributed source extract; the implementation plan is new AI-assisted editorial material, offered under the same CC BY 4.0 license. This guide has not undergone independent peer review.