A meta-analysis of 34 studies and 188 effect sizes shows the link between pretend play and social competence is real — but the heart of it is how children play, not how much.
Introduction: We Know Play Matters — But How Exactly?
“Children grow through play” has become common sense. Early childhood experts, parenting books, and online guides all agree. Pretend play in particular — playing house, playing doctor, role-playing fantasy scenarios — is seen as central to social development.
But is it really? The claim that “pretend play builds social skills” traces back to Vygotsky’s developmental theory, but whether the empirical evidence is as strong as we assume is another question. A large 2013 review by Lillard and colleagues argued that the evidence for pretend play’s effects is weaker than commonly believed, sparking debate in the field.
A decade on, more studies have accumulated. A team at Vrije Universiteit Amsterdam decided to tackle the question head-on by statistically synthesizing the existing research through a meta-analysis.
The Core Research Question
“To what extent, and in what ways, is pretend play associated with social competence in young children?”
More specifically:
- Is there an overall positive relationship between pretend play and social competence?
- Which aspects of social competence (peer relationships, cooperation, emotion understanding) are involved?
- Does play’s quantity (frequency, duration) or its quality (richness, complexity) predict social competence more strongly?
- Does the relationship change with age?
How the Study Was Done: What a Meta-Analysis Is
Meta-analysis systematically gathers existing studies on a topic and statistically combines their results. A finding replicated across dozens of studies is more trustworthy than any single study of 30 children.
The researchers searched three databases — PsycINFO, ERIC, and Web of Science — for relevant papers.
The Selection Funnel
- Initial search hits: 4,785 papers
- After deduplication: 4,535
- After title/abstract screening: 146
- After full-text review: 34 studies selected
- Statistical associations extracted: 188 effect sizes
The age range was 3–8 years. Various forms of pretend play (house, role-play, sociodramatic play) and multiple dimensions of social competence (peer relationships, cooperation, emotion regulation, empathy) were covered.
Results: The Relationship Exists, but the Effect Is Small
The Overall Effect
Pretend play and social competence showed a statistically significant positive relationship.
- Effect size: g = 0.166 (95% CI [0.042, 0.290], p < .01)
- Based on 34 studies and 188 effect sizes
This held regardless of how social competence was measured — peer relationships, social skills, emotion understanding, cooperation.
What “Small” Means
In the social sciences, g = 0.166 counts as a small effect. How should we read that number?
It means pretend play is not a “magic key” to social competence. A child’s social skills emerge from many factors working together — temperament, family environment, peer experiences, cultural context. Pretend play is one meaningful contributor among many, not the single determinant.
That said, a small effect that replicates consistently across diverse studies in a meta-analysis is a real relationship.
The Most Interesting Finding: Quality Over Quantity
Studies That Measured Quantity vs. Studies That Measured Quality
The meta-analysis’s most notable result came from moderator analysis.
- Studies measuring quantity of pretend play (frequency, duration) showed relatively weaker correlations with social competence.
- Studies measuring quality (complexity, richness, elaboration) showed stronger correlations.
This is both intuitive and important.
“Just let them play a lot” is half right. A child spending two hours doing the same role is qualitatively different from a child spending 30 minutes inventing new scenarios, negotiating roles with peers, and building storylines together.
What “High-Quality Pretend Play” Looks Like
The studies’ “quality” markers included:
- Role complexity — beyond “I’m mommy, you’re the baby,” multiple roles interacting
- Story elaboration — play that has setup, conflict, resolution
- Decontextualization — replacing absent objects or situations with imagination (e.g., using a block as a phone)
- Negotiation with peers — jointly deciding roles and how the story unfolds
Play rich in these elements correlated more strongly with social competence.
Does the Relationship Weaken With Age?
Another interesting finding was an age effect. As children grew older, the link between pretend play and social competence weakened slightly.
The researchers didn’t propose a definitive explanation, but some possibilities:
- Pretend play is the primary channel of social interaction for younger children; as they age, other forms (rule-based games, sports, conversation) grow.
- Around school age, pretend play itself declines in frequency, weakening measurement sensitivity.
- As social competence matures, factors other than pretend play (peer experiences, classroom environment) may play a larger role.
This does not mean “older children don’t need pretend play.” The relationship weakens but doesn’t disappear, and it remains positive.
Practical Implications
Play Time Alone Isn’t Enough
“Increase free play time” is the right direction, but the message of this study is that it isn’t sufficient. Children cycling through the same play pattern benefit less than children in environments and relationships that grow the complexity and richness of their play.
The Adult’s Role Is “Extension,” Not “Intervention”
Related research (not covered directly here) shows that when an adult participates in a child’s pretend play appropriately — suggesting new roles, extending storylines, asking “what would happen if…?” — the quality of play rises. The point is not to direct the play but to enrich the story the child is leading.
Don’t Overinterpret
Pretend play is one among many contributors to social development. A child who doesn’t enjoy role-play isn’t destined to have social problems. Social skills develop through many channels; pretend play is one of them.
Limitations
- Most of the 34 studies were cross-sectional (measured at a single point), so causal direction can’t be established. Does pretend play build social competence, are socially skilled children better at pretend play, or does a third factor (temperament, parenting) influence both? This meta-analysis can’t decide.
- Included studies varied in how they measured both pretend play and social competence. The team tried to control for this heterogeneity, but comparisons aren’t fully apples-to-apples.
- A small effect size means at the individual-child level, the impact of pretend play may be hard to perceive directly.
Final Thoughts
The relationship between pretend play and social competence is real, but the key isn’t “how much” a child plays; it’s “how richly” they play.
This study sharpens the vague claim that “play is good.” What matters isn’t more play time but the quality of experience — trying on new roles, co-creating stories with peers, building more elaborate imaginary worlds. And an adult who asks the right questions and helps extend the story makes a quiet but meaningful difference in that quality.
Source: Smits-van der Nat, M., van der Wilt, F., Meeter, M., & van der Veen, C. (2024). The value of pretend play for social competence in early childhood: A meta-analysis. Educational Psychology Review, 36, 46. https://doi.org/10.1007/s10648-024-09884-z