Choice or Structure? What 94 Studies Found About Student Motivation

Teachers are often pulled between two ideas.

One says students need more freedom.

Let them choose.

Give them ownership.

Avoid controlling every step.

The other says students need structure.

Clear expectations.

Guidance.

Feedback.

Predictable routines.

These can sound like opposites.

But a large meta-analysis suggests that may be the wrong way to think about them.

Researchers reviewed 94 studies and 110 effect sizes examining two approaches:

• autonomy support
• teacher-provided structure

The strongest conclusion was not that one beats the other.

It was:

Students appear to benefit most when teachers provide both.

Autonomy support and structure were themselves strongly related, with an overall correlation of:

r = 0.53

And both were positively associated with student motivation, engagement and feelings of competence and autonomy.

Read the full open-access meta-analysis

The short version

The review found that:

• 94 studies were included
• 110 effect sizes were analyzed for the relationship between autonomy support and structure
• autonomy support and structure were strongly positively related: r = 0.53
• both were associated with greater autonomous motivation
• both were associated with stronger student engagement
• both were associated with students feeling more competent
• both were associated with students feeling more autonomous
• when researchers statistically removed one, the effects of the other became noticeably smaller

That last result is especially important.

It suggests that the two approaches may reinforce each other.

The practical takeaway is:

Choice without guidance can become confusion. Structure without autonomy can become control. Good teaching often combines both.

What is autonomy support?

Autonomy support does not mean:

“Students can do whatever they want.”

In the research, it usually includes things such as:

• offering meaningful choices
• explaining why a task matters
• acknowledging students’ perspectives
• allowing an appropriate amount of independence
• avoiding unnecessarily controlling language
• encouraging students to take initiative

For example, compare:

Complete questions 1–20 because I said so.

with:

We’re practising these problems because tomorrow you’ll need the same reasoning in a more complex task. Start with either section A or B.

The teacher still has a learning goal.

The difference is that the student is given:

• a rationale
• some ownership
• some room to act

Autonomy is not the absence of teaching.

It is the feeling that you are participating in your own learning rather than simply being pushed through it.

What is structure?

Structure is also easy to misunderstand.

It does not mean strictness.

It does not necessarily mean more rules.

In the studies reviewed, structure usually meant things such as:

• clear expectations
• guidance
• predictable organization
• informational feedback
• appropriate challenge
• help understanding how to succeed

A structured teacher might say:

Your goal is to explain why this answer works. Here are the three criteria I’ll use. Try the first example, then check your reasoning against the model.

The student knows:

• what success looks like
• what to do first
• how they will receive feedback
• what the boundaries are

That can make autonomy easier.

It is much easier to make meaningful choices when you understand the task.

Autonomy and structure were strongly connected

Across the 110 effect sizes, the overall relationship between autonomy support and structure was:

r = 0.53

That is a large positive correlation.

In other words, teachers who tended to support student autonomy also tended to provide structure.

That is important because these approaches are sometimes presented as opposites.

For example:

Student choice vs teacher direction

Freedom vs rules

Independent learning vs explicit guidance

The research suggests that real classrooms may not work that way.

Good teachers can simultaneously say:

“You have some control over how you do this.”

and:

“Here is what success looks like and how I will help you get there.”

Structure did not reduce autonomy

This may be the most counterintuitive finding.

Teacher structure was strongly associated with students’ sense of autonomy:

r = 0.43

It was also strongly associated with students’ sense of competence:

r = 0.41

That seems strange at first.

Wouldn’t more structure make students feel less free?

Not necessarily.

Imagine two tasks.

Task A

Create a project about climate change. Do whatever you want.

Maximum freedom.

Almost no structure.

Some students will love it.

Others will immediately wonder:

• What am I supposed to make?
• How long should it be?
• What counts as good?
• Where do I start?
• Can I choose anything?

Task B

Choose one of three climate questions. Create either a short video, infographic or written explanation. Your work needs to include two sources, one piece of data and a conclusion. Here are examples.

There are more boundaries.

But students may actually feel more capable of making choices because they understand the space in which those choices operate.

Structure can make autonomy usable.

Autonomy support also depended partly on structure

The researchers performed another useful analysis.

They asked:

What happens to the relationship between autonomy support and student outcomes if we statistically remove the effect of structure?

The relationships became smaller.

For example, autonomy support was associated with students’ sense of autonomy at:

r = 0.47

But after controlling for structure, that dropped to:

r = 0.29

For competence, the relationship dropped from:

r = 0.36

to:

r = 0.16

That suggests that some of the benefits we attribute to autonomy-supportive teaching may actually come from the fact that those teachers are also providing good structure.

The same happened in the other direction

The researchers then reversed the analysis.

They looked at structure while statistically removing autonomy support.

Again, the effects became smaller.

For autonomous motivation, structure had an ordinary correlation of:

r = 0.28

After removing the influence of autonomy support, it dropped to:

r = 0.15

The strongest remaining relationship was with competence:

r = 0.25

So neither approach appears to operate completely independently.

They seem to work together.

Choice alone is not necessarily motivating

This matters because “give students choice” has become a common recommendation.

Choice can be useful.

But not all choice is meaningful.

Imagine asking:

Do you want to write with the blue pen or the black pen?

That is technically a choice.

It probably does very little for motivation.

Or:

Choose any topic in history and make any kind of project.

That provides enormous freedom.

But it may also create unnecessary cognitive load.

Autonomy support is more than offering options.

It includes helping students understand:

• why the work matters
• what choices are genuinely available
• how those choices connect to the goal

A small meaningful choice may be more motivating than unlimited freedom.

Structure alone can become controlling

The opposite problem is also possible.

A highly structured classroom can provide:

• detailed instructions
• explicit criteria
• precise schedules
• constant feedback

and still make students feel that they have almost no ownership.

For example:

Sit here.

Do question 1.

Now question 2.

Use this method.

Do not try another one.

Stop.

Wait.

Continue.

The student always knows what to do.

But they may never need to make a decision.

Structure becomes controlling when it removes unnecessary amounts of agency.

The research suggests that the most supportive version of structure still leaves room for the student to participate in the learning process.

Motivation improved with both

Both autonomy support and structure were positively associated with autonomous motivation.

That is the kind of motivation where students engage because they:

• enjoy the activity
• value it
• understand its importance
• personally endorse the goal

This is different from controlled motivation:

I’m doing this because I’ll get punished if I don’t.

or:

I need the grade.

The review found that autonomy support and structure both tended to be associated with the more self-directed form of motivation.

That is useful because it challenges another simple assumption:

Structure does not necessarily make motivation less intrinsic.

Clear guidance can help students feel competent.

Competence can make an activity more rewarding.

And students often enjoy things more when they understand how to succeed.

Engagement improved too

Both approaches were also associated with greater student engagement.

That includes things such as:

• effort
• persistence
• participation
• attention
• involvement in learning

Again, this makes sense when the two approaches are combined.

Students are more likely to persist when:

I understand what I’m supposed to do.

and:

I feel some ownership over doing it.

Remove either one, and engagement can become harder.

The relationship appeared across subjects and cultures

The researchers tested whether the relationship between autonomy support and structure differed across subject areas.

It did not show meaningful variation.

The same was true when they examined cultural individualism.

The authors therefore describe the relationship as appearing fairly universal across the contexts represented in the evidence.

That does not mean every classroom should look identical.

A five-year-old and a university student need different levels of guidance.

A maths lesson and an art project require different kinds of choice.

But the broader principle appears fairly robust:

Students generally benefit from having both agency and clarity.

School level did matter somewhat

The relationship was not identical across every education level.

School level moderated the strength of the relationship between autonomy support and structure.

That makes sense.

Students’ capacity for independent planning develops over time.

Younger students may need:

• more explicit routines
• narrower choices
• more immediate guidance

Older students may be able to handle:

• broader decisions
• longer-term planning
• more open-ended tasks

The balance can change.

The principle does not have to.

There is substantial variation between studies

This is an important caveat.

The review found extremely high heterogeneity:

I² = 98.94%

That means results varied substantially between studies.

Some of that is likely because researchers measured autonomy support and structure differently.

The studies also involved:

• different ages
• different countries
• different subjects
• different classroom environments
• different measurement methods

So r = 0.53 should not be treated as a universal constant.

The authors also found some evidence that small or negative studies could be underrepresented, although their tests did not show strong statistical asymmetry.

The overall pattern is convincing.

The exact size is less certain.

What does this look like in a real lesson?

Imagine students are writing an argumentative essay.

A purely controlling approach might be:

Write five paragraphs. Use these three arguments. Follow this exact structure.

A poorly structured autonomy approach might be:

Write something persuasive about anything you care about.

A blended approach could be:

Choose one of these four questions, or propose your own. Your essay needs a clear claim, evidence from at least two sources and a response to a counterargument. Here are two examples. You can organize your argument in the way you think works best.

The student has:

choice

but also:

criteria

guidance

feedback

boundaries

That is probably close to what the research means by combining autonomy support with structure.

A simple classroom checklist

Before a task, ask two sets of questions.

Structure

Do students know:

• what the goal is?
• what success looks like?
• where to start?
• what support is available?
• how they will get feedback?

Autonomy

Do students have:

• a meaningful choice?
• room to make decisions?
• a reason for doing the task?
• permission to use different approaches?
• a chance to express preferences or ideas?

If one side is missing, the task may become weaker.

The biggest takeaway

Teaching is sometimes framed as a choice between two philosophies.

Give students freedom.

or:

Give students structure.

This meta-analysis suggests that is a false choice.

Across 94 studies, autonomy support and structure were strongly related and both were associated with better student motivation, engagement and psychological need satisfaction.

And when researchers statistically separated them, the effects of each became smaller.

So perhaps the better question is not:

“How much freedom should students have?”

It is:

“How can we give students enough structure to succeed and enough autonomy to make the learning feel like their own?”

Students may not need freedom from structure.

They may need freedom inside good structure.

Source

Patzak, A. & Zhang, X. (2025).
Blending Teacher Autonomy Support and Provision of Structure in the Classroom for Optimal Motivation: A Systematic Review and Meta-Analysis. Educational Psychology Review, 37, 17.

Read the full open-access study