Social cognition
Do young children trust a robot the way they trust a person?
Open access · cc by · source: Europe PMC
Children trusted a robot about as readily as a person overall, but the youngest favoured the human while seven-year-olds favoured the robot once theory of mind was accounted for.
Study at a glance
- Design
- Human experiment — Each child played a guessing-based Trusting Game once with a human and once with the NAO robot (order counterbalanced), across four age groups; trust was related to theory of mind, attachment and switching tasks.
- N
- N=94 · 94 Italian children recruited (22 three-year-olds, 24 five-year-olds, 25 seven-year-olds, 23 nine-year-olds); a few children with a missing session were dropped from analyses.
- Population
- Italian preschool and primary-school children aged 3 to 9 from Milan-area schools
- Outcome
- Number of trials before the child followed (trust), stopped following (mistrust) and again followed (restored trust) the partner's guesses; attributions of mental states to human and robot
Structured fields used in claim comparison tables when every cited study has a complete layer.
Key findings
Children's tendency to trust dropped across the game, being lowest in the restoration phase, and 3-year-olds were more trusting than every older group. Partner type had no effect on its own, but once first-order theory of mind was controlled, 3-year-olds trusted the human more while 7-year-olds trusted the robot more. Better theory of mind went with more cautious trust, more secure attachment went with more trust (mostly with the human and in the youngest children), and even 3-year-olds credited the human with more perception and knowledge than the robot.
Methodology
Children aged 3, 5, 7 and 9 played a 'guess where the doll is' game with either an adult experimenter or the humanoid robot NAO, and later with the other partner. The partner always guessed first; the game was rigged so the partner was first always right (trust acquisition), then always wrong (trust loss), then right again (trust restoration), and the researchers counted how many trials the child took to switch behaviour. Children also completed false-belief (theory of mind) tasks, an attachment-representation task, a response-switching task and a questionnaire on what mental states humans and robots have.
Limitations
Groups were small (roughly twenty-odd children per age), all from a few schools in one Italian region, and the authors themselves call for a larger, longitudinal sample. The age comparisons are cross-sectional, and the links with attachment, theory of mind and switching are many separate correlations run without correction across subgroups, so some may be chance findings. The robot versus human difference only appeared after adding theory of mind as a covariate, and the game involved pure guessing rather than a partner with real knowledge, so it may not generalise to trust in informative robots such as tutors.
How this study connects
Role on claims
Each row is a claim on a concept or method page where this paper supports, challenges, or qualifies the statement. Roles are hand-checked — not a model guess.
Not yet placed on a claim. This paper has study layers, but no concept page yet cites it as support, challenge, or qualifier.
Related papers in this topic
Same topic cluster — not a recommendation engine.