When Kids Meet Pepper: What Children’s Interactions with Humanoid Robots Reveal

Robots in classrooms are no longer futuristic curiosities—they’re becoming familiar companions in lessons, activities, and play. But how do children actually treat humanoid robots, and what do they believe about these machines? New research from SWPS University suggests that young students not only behave politely toward robots, but many—especially younger children and girls—readily attribute human-like qualities to them.

These findings were highlighted at the fourth HumanTech Summit in Warsaw, where researchers explored how social robots are reshaping education. Tools like Pepper, the 120-centimeter humanoid robot developed by SoftBank Robotics, are increasingly used to support learning through tailored instruction, playful engagement, and real-time feedback. But their presence also raises questions about long-term effects on social development.

Konrad Maj (Ph.D.), social psychologist, SWPS University and humanoid robot Pepper. Credit: SWPS University

The SWPS research team—Konrad Maj, Ariadna Gołębicka, and Zuzanna Siwińska—set out to understand how children interpret a robot’s behavior and in what circumstances robots become effective learning partners. Their study, published in Computers & Education, examined how 251 children aged 7–12 responded to Pepper when its communication style and assigned “gender” were varied. Pepper either spoke politely or issued firm instructions, and was introduced as either “Adam” or “Ada.”

The results were surprisingly consistent. Children almost always responded politely to Pepper, even when the robot used a commanding tone—suggesting that established human social norms override mimicry in these interactions. Younger children and girls were more inclined to see Pepper as capable of emotions, dreams, or imagination, and polite robots—particularly those programmed with traditionally “female” cues—were viewed as more human-like than commanding ones.

These patterns matter because they show how deeply children’s expectations shape their interactions with robots. The social signals a robot uses—tone, warmth, even its assigned name—can influence engagement, trust, and learning outcomes. As Konrad Maj notes, understanding these dynamics will be vital as robots become more common in educational environments. The more we understand how children perceive them, the better we can design robots that support—not hinder—healthy development and meaningful learning.