Psychological needs satisfaction predicts practice time, retention, and self-esteem in high school orchestra

Every director has watched a student drift. Practice logs come back blank or suspiciously round. Participation in rehearsal gets quieter. Then one spring, they don’t re-enroll. By the time the chair sits empty, the decision was made months earlier, quietly, at home.

What drives that decision? And more pointedly, what could have changed it?

Paul Evans (University of New South Wales) and Mark Y. Liu (Boston University) took on this question directly in a 2018 study published in the Journal of Research in Music Education. They surveyed 704 high school string students across three schools in a single midwestern US district and used structural equation modeling to test whether students’ psychological needs, as defined by self-determination theory, predicted three outcomes directors care deeply about: how much students practice, whether they intend to stay in the program, and how they feel about themselves. The results make a strong case that how students experience orchestra, not just what they’re taught in it, shapes what happens next.

Why this paper matters for directors

Research on student motivation in music has accumulated steadily over the past two decades, but much of it focuses on university musicians or uses samples from countries outside the US. This study is notable for several reasons. The sample is large (704 students, high response rate of 99.7%), it covers an actual ensemble program rather than a general music class, and one of the authors (Liu) is himself an orchestra director. The study doesn’t just ask what motivates students in the abstract. It asks what predicts the specific behaviors and decisions that directors watch play out every year.

The three outcomes it tracks are not academic abstractions. Practice time connects directly to skill development and performance quality. Intentions to continue map onto the enrollment decisions that determine whether a program grows or shrinks. Self-esteem is included because music education has long claimed broad developmental benefits, and this study tests whether those benefits are real or are just side effects of a good social environment.

What self-determination theory actually says

Self-determination theory (SDT), developed by Ryan and Deci, is a framework for understanding what sustains motivated behavior over time. It distinguishes between intrinsic motivation (doing something because it’s genuinely interesting or enjoyable) and extrinsic motivation (doing something for an external reason, like grades, parental pressure, or avoiding punishment). Decades of research across many domains show that intrinsic motivation produces better learning, more persistence, and greater well-being than extrinsic motivation, which can actually undermine the intrinsic kind over time.

The part of SDT that Evans and Liu focus on is basic psychological needs theory. This mini-theory within SDT identifies three psychological needs that humans require for well-being and sustained motivation. When a social environment satisfies these needs, people tend to thrive. When it frustrates them, people tend to disengage, experience negative emotions, and sometimes exit entirely.

The three needs are:

Competence

The feeling that you’re effective, that your efforts lead somewhere, that you can actually do the thing. In orchestra, competence satisfaction means students believe practice improves their playing. Competence frustration looks like helplessness: the student who puts in the time but can’t hear the connection between effort and result.

Relatedness

The sense of belonging and connection. This doesn’t require close friendship. It means students feel they are part of something, that their presence matters, that they are not isolated. In ensemble music this might seem automatic (you’re literally playing with other people), but a student can sit in an orchestra for three years and still feel socially peripheral to the group.

Autonomy

The sense that your participation reflects your own values and choices, not just external pressure. A student can choose to be in orchestra while still feeling controlled if every decision about what they play, how they practice, and how they’re evaluated is handed down without any input. Autonomy support doesn’t mean students get to do whatever they want; it means teachers explain the reasons behind structures and give students genuine choices where possible.

What makes this theory useful for music educators is that these needs are shaped by the environment. Teachers are the main architects of that environment. A director who gives students some voice in repertoire selection, who provides feedback that connects effort to progress, and who builds an ensemble culture where students feel genuinely connected is doing needs-satisfaction work, whether they call it that or not.

The three outcomes the study measured

Practice time

Measured by asking students how many days per week they practice and how many minutes per session. The two numbers were multiplied to get weekly minutes. The sample average was about 131 minutes per week, with substantial variation. This is a behavioral measure, not an assessment of quality.

Intentions to continue

Used three items: how long a student intends to participate (from “until the end of the year” to “for the rest of my life”), how often they think about quitting, and how likely they are to enroll next year. These are cognitive measures, not actual enrollment data, which is a real limitation the authors acknowledge.

Global self-esteem

Used four items from the Rosenberg Self-Esteem Scale, one of the most widely validated measures in psychological research. Self-esteem here is not domain-specific (not “I’m a good musician”) but global: a general sense of self-worth.

The inclusion of self-esteem is particularly interesting because it tests a claim music educators often make, that participation in a music program builds character and self-confidence. Evans and Liu’s framing is more precise: they’re not asking whether being in orchestra improves self-esteem, but whether the motivational climate in which orchestra happens predicts self-esteem. That’s a different and more actionable question.

How the study was designed

All three high schools in a single midwestern school district participated. The district served about 30,000 students, with roughly 30% participating in music programs at each school. The string programs at all three schools were well-established and had received state and national recognition, meaning this is not a study of struggling programs.

Students completed surveys during regular rehearsal time. Only two students declined to participate, giving a response rate of 99.7%. After removing statistical outliers, 692 cases were used in the analysis.

The analytical approach was structural equation modeling (SEM), a statistical technique that accounts for measurement error, tests multiple relationships simultaneously, and estimates how well the overall model fits the data. The model fit was good by conventional standards (CFI = .934, RMSEA = .063). The authors also tested whether the model held equally for boys and girls (it did) and for students with more versus less prior musical experience (mostly, with some meaningful differences described below).

Needs satisfaction: the findings in detail

The structural model showed that psychological needs satisfaction was a significant positive predictor of all three outcomes, with effect sizes that are hard to dismiss.

The authors offer an interpretation of the self-esteem finding that’s worth sitting with: it may not be music participation itself that builds self-esteem, but the motivational climate in which that participation happens. This reframes several earlier studies that found inconsistent effects of music on self-esteem. If the climate is supportive, self-esteem benefits follow. If it’s controlling or isolating, they may not.

Why would needs satisfaction drive these outcomes? The paper works through the logic carefully. Competence satisfaction helps students see the connection between effort and improvement; without it, practice feels pointless. Autonomy satisfaction helps students internalize the value of practice rather than doing it only under pressure; students who have internalized that value don’t stop when no one is watching. Relatedness satisfaction gives solo practice a social meaning: students know that the work they do alone contributes to something collective and that other people are doing the same work and facing the same challenges.

The complicated picture with needs frustration

The study’s most interesting finding, and probably its most counterintuitive one, is about psychological needs frustration.

The authors expected frustration to predict all three outcomes negatively: more frustration, less practice, lower intentions, lower self-esteem. Two of those three hypotheses held. Needs frustration significantly and negatively predicted self-esteem, which makes sense given how closely self-esteem is theorized to track the quality of a person’s social environment.

But for the other two outcomes, the picture was messier.

For practice time, frustration actually predicted higher practice, not lower, with a small but statistically significant positive path coefficient (beta = .100). This seems paradoxical until you recall that controlled extrinsic motivation can drive behavior even as it undermines well-being. Students who feel pressured, controlled, or fearful of consequences may still practice, maybe more than their intrinsically motivated peers in the short run. What SDT research consistently shows, though, is that controlled motivation affects quantity without improving quality. Those students are probably going through the motions rather than practicing deliberately.

For intentions to continue, frustration was not a significant predictor once needs satisfaction was accounted for in the model (despite a meaningful zero-order correlation of -.271). The authors offer two possible explanations. Students in these well-established, recognized programs may have been resilient enough to offset negative experiences with positive ones. Or the incidence of frustration in these particular programs was low enough that it couldn’t move the needle. Either way, the implication for practice is cautious: reducing sources of frustration alone may not be sufficient to change retention outcomes. Actively increasing needs satisfaction appears to be the more consequential lever.

This distinction, increasing positive support versus reducing negative pressure, parallels findings in broader educational research. Taking the foot off the brake is not the same as pressing the accelerator.

What more experience does (and doesn’t) change

The study compared students with more versus less than two years of formal musical experience. More experienced students reported higher practice time, stronger intentions to continue, and greater competence satisfaction, which is not surprising. Students who have stuck with an instrument for longer have already crossed some of the hardest motivational hurdles.

What is worth noting is that the structural model held up at the level of relationships. Less experienced students benefited from needs satisfaction just as much as more experienced students, even though their baseline levels of practice, intentions, and competence were all lower. The mechanisms are the same regardless of experience level. A beginner who feels competent, connected, and autonomous is more likely to practice and stay than a beginner who doesn’t, to the same degree as a veteran player.

What teachers can actually do with this

The implications section of the paper is unusually specific for a research article, partly because Liu is a working director and not just an academic researcher. The recommendations are worth spelling out.

To increase practice, address all three needs

• Competence support: Give feedback that draws an explicit line between effort and specific improvement. "Your intonation in the third measure is cleaner than last week" is competence-supporting. "You need to practice more" is not.
• Relatedness support: Acknowledge shared practice experiences in rehearsal, what feels hard, what surprises people, what small breakthroughs feel like. This reminds students that they belong to a community of people doing the same difficult thing alone.
• Autonomy support: Help students build practice schedules that reflect their own priorities rather than prescribing exact amounts. Ownership of the structure makes it more likely to survive when motivation dips.

To improve retention, make the value of continuing visible

• Relatedness support: Connect students to what ensemble membership offers that individual practice does not: cooperation, shared achievement, the specific kind of social belonging that comes from making music with others.
• Autonomy support: Explain the reasons for structures and requirements rather than simply enforcing them.
• Competence support: Expose students to peers at higher levels, giving them a concrete view of what continued participation makes possible.

To support self-esteem, connect contribution to effort, not just performance

Students whose self-worth becomes tied to performance outcomes are vulnerable to the inevitable bad audition, the difficult concert, the missed note. Needs-supportive feedback separates the student’s value as a person and ensemble member from any single performance moment.

The authors also note that teacher professional development could be aimed specifically at autonomy-supportive teaching behaviors, with evidence from other disciplines showing that teachers can learn these behaviors and that students benefit measurably when they do. Music-specific training of this kind has not yet been developed and tested.

Limits of the study worth knowing

The authors are unusually candid about what the study can’t tell us.

• Self-reported practice time. High school students' estimates of how much they practice may not be accurate. The authors suggest diary methods would be more reliable.
• Intentions versus actual decisions. Intentions to continue are not the same as actual enrollment decisions. Students who report strong intentions to stay may still exit when they face real trade-offs against other classes and activities.
• Cross-sectional design. All variables were measured at a single point in time. This limits what can be concluded about cause and effect. Needs satisfaction probably drives practice and intentions, but better practice and stronger engagement might also reinforce needs satisfaction over time.
• Single-district sample. The sample comes from a single district with well-recognized programs. Programs that are struggling, under-resourced, or in districts with different demographics may produce different results.
• Collapsed needs structure. The study collapsed all three psychological needs into two higher-order factors (satisfaction and frustration) rather than examining competence, relatedness, and autonomy separately. The correlations in the paper hint that some needs may matter more than others for specific outcomes, but the study design doesn't let us say this cleanly.