Since 2022, Raising the Barre has run an IRB-approved study of mood change in children and adolescents during dance class, in conjunction with the University of Pennsylvania. The paper was published 11 June 2026 in Frontiers in Psychology (Performance Science). A second study, on executive function in young dancers, is underway at Stanford.
Diagnosed mental and behavioral health conditions among 12-to-17-year-olds in the United States rose 35% between 2016 and 2023. Anxiety diagnoses climbed 61%. Depression diagnoses rose 45%.
That’s the backdrop our research speaks to. We aren’t claiming dance class is a cure. We’re asking a smaller, more answerable question: does an hour in the studio change how kids feel? Across more than four thousand class sessions, the answer was yes — small, reliable, and consistent across ages, styles, and skill levels.
Our paper — Mood Changes of Children and Adolescents in Dance Classes: A Prospective Repeated-Measures Study — was published in Frontiers in Psychology (Performance Science) on 11 June 2026. The full article is open access.
Across more than four thousand paired pre- and post-class mood ratings, post-class mood was significantly higher than pre-class mood. The lift held across genres, age groups, class types, beginners and advanced, ballet and hip-hop, morning classes and evening classes.
Read this as: across more than four thousand tracked sessions, dancers left the studio in better or equal spirits than they arrived in. They didn’t leave feeling worse.
In plain English: the mood lift after dance class is real and reliable. It’s not enormous — it’s a small-to-medium effect — but it shows up clearly in the data, again and again, across thousands of class sessions.
What the analysis found. No single feature of a class — how often the dancer attended, what style, what level, who taught it — explained the variation in mood response. Individual differences between dancers accounted for most of the meaningful variation.
The mood study uses a prospective, repeated-measures design. Each participant served as their own control: mood was recorded immediately before and immediately after the same dance class session, repeatedly across the study period (February–May 2025). The design supports inference about within-session change in mood. As stated in the published paper: “the observational design precludes causal attribution.” The design does not support claims about long-term well-being, character, or development — questions we treat as separate.
The primary measure was a single-item emoji scale, administered on iPad immediately before and after each class. The brief format keeps response burden minimal for young dancers across the full age range and enables high-frequency, in-context (ecological momentary) assessment of mood.
To assess whether the emoji scale captures something meaningful, we also administered the PANAS-C (Positive and Negative Affect Schedule for Children, a validated published instrument) to a subset of participants. The emoji scale showed significant convergent validity with PANAS-C Positive Affect (r = .19, p = .011, n = 175), with the strongest correlation among students attending 4–6 classes per week (r = .51, n = 25). In plain English: the quick emoji rating tracks the kind of mood the longer, validated PANAS-C measures — especially for more-involved dancers.
The published study reports data from 256 children and adolescents at four studio locations, measured during February through May 2025. Participants completed 4,224 paired pre- and post-class mood ratings; the primary analytic sample comprised 4,059 sessions from 251 students. Participants ranged in age from 5 to 17, with parental consent on file under IRB Protocol #857840.
Participation in the research was open to any consenting dance student at a participating studio, regardless of whether they were also enrolled in the Raising the Barre youth-leadership program. Research participants and RTB program participants are two different groups.
| Measure | Value |
|---|---|
| Students measured | 256 |
| Paired pre/post ratings | 4,224 |
| Primary analytic sessions | 4,059 |
| Sessions: maintained or improved | 85.8% |
| Session-level effect (Cohen’s d) | 0.27 |
| Per-student effect (Cohen’s dz) | 0.43 |
| Statistical significance | p < .001 |
| Intraclass correlation (ICC) | .432 |
The data were analyzed using multilevel models to account for the nested structure (sessions nested within students; students contributed between 1 and 109 sessions each). The intraclass correlation (ICC = .432) indicated that a substantial share of variation was between students rather than within them — a finding that meaningfully shapes how to read the results.
A Tobit sensitivity analysis was used to confirm the results after adjusting for a 28.8% post-class ceiling rate (dancers reporting the maximum mood rating after class). Random-slopes models substantially improved model fit (ΔAIC = 201), indicating that dancers’ individual responses to dance class varied meaningfully — the same class affects different dancers differently.
No covariate — including weekly class frequency, class type, proficiency level, or instructor experience — explained significant additional variance. The full statistical methodology is documented in the accepted manuscript at Frontiers in Psychology.
The within-subject, repeated-measures design is a real strength: it controls for stable individual differences and supports inference about within-session change. Sample size and session count are both large for a study of this kind in a real-world (non-laboratory) setting. Convergent validity with a published instrument (PANAS-C) provides external grounding for the brief emoji measure.
At the same time, this is observational research conducted in real dance studios, not a randomized controlled trial. The paper’s own conclusion is direct: “the observational design precludes causal attribution.” The dancers in our data chose to attend dance class; we cannot rule out selection effects, and self-report mood scales carry known limitations. The acute mood effect we observe is an effect per session, not a claim about cumulative benefits over months or years.
We do not claim that dance class causes long-term changes in well-being, character, leadership, or academic outcomes. Those are separate questions, and answering them rigorously would require different study designs — ones we are actively pursuing through the Stanford executive function collaboration and through future longitudinal work.
In April 2026, we launched a second research collaboration — this one with Stanford — examining executive function in young dancers. Executive function refers to the suite of cognitive processes that enable goal-directed behavior: working memory, cognitive flexibility, and inhibitory control.
We welcome inquiries from researchers interested in the data, instrument, replication studies, or downstream collaborations. Data availability follows the journal’s policy; specific questions are best directed to the corresponding author.
research@raisingthebarredance.orgFor journalists covering youth wellness, dance, or applied positive psychology, please reach out for interview availability or to request the press kit. We provide accurate methodological framing and respect embargo conventions on the manuscript until publication.
press@raisingthebarredance.orgStudios participating in IRB-approved data collection access the tracking app through the Dance Positive platform. This is an instructor and parent-portal tool, not a public-facing study experience — if you’re an existing participating site, your contact at RTB has your access details.
Contact RTB