Resources
Speech Therapy Homework That Works: How to Get Students Practicing Between Sessions
Why most home practice programs fail and what the research says about designing ones that don't
April 6, 2026•9 min read
Why most home practice programs fail and what the research says about designing ones that don't
Homework compliance in speech therapy is notoriously low. Studies suggest 30% to 70% of families do not follow through on home practice consistently. At the same time, the research is clear that practice between sessions is one of the strongest predictors of progress.
The thing that matters most is the thing that happens least. This article is about closing that gap by understanding why compliance is low and redesigning home practice so families can actually do it correctly.
Most parents care deeply. Compliance failures are usually design problems, not motivation problems. The most common reasons home practice breaks down:
When a child says "rabbit," most parents cannot reliably judge whether the /r/ was correct. Without a trained ear, they are supervising blind. They know it, and the practice feels pointless.
Many home programs are designed for 20 to 30 minute sessions. For a family already juggling school, dinner, and bedtime, that is unrealistic. Families try a few times and stop.
"Practice at the sentence level" means something specific to an SLP and almost nothing to a parent. Vague instructions lead to incorrect practice or no practice at all.
When homework goes out with no follow up, it signals that it is optional. Without check ins or visible tracking, it quietly drops off the priority list.
If practice becomes a battle, parents will protect the relationship over forcing compliance. The solution is not pushing harder, it is making practice less aversive.
Motor learning research consistently shows that short, frequent practice sessions outperform long, infrequent ones. This principle, called distributed practice, has been demonstrated across dozens of motor skill domains. Speech is no exception.
Massed practice:
Distributed practice:
Maas et al. (2008) reviewed the evidence as it applies to speech therapy and found that distributed practice supports better retention and generalization. Less time per session, more often, is not lowering the bar. It is a more effective structure that families can actually sustain.
Five principles consistently improve follow through:
The single most impactful change most SLPs can make. Ten minutes fits before dinner or in the car. Thirty minutes does not. Families can sustain short sessions.
Send a specific list of 8 to 10 words at the right level. Tell parents exactly what to do with them. Specificity drives follow through.
Home practice should be at or slightly below the child's current therapy level. Practice that is too hard produces errors, frustration, and avoidance.
Habits stick when anchored to existing ones. "Right after brushing teeth" or "in the car on the way to school" works better than a vague "sometime each day."
Every point of friction reduces follow through. The ideal home program requires opening one thing and starting immediately.
Practice without feedback is dramatically less effective than practice with it. Worse, a child who repeats errors without correction may be reinforcing the wrong motor plans. This is exactly what SLPs are right to worry about.
The hard part is that most parents cannot reliably judge whether a child's /r/, /s/, or /l/ production is accurate. These are subtle distinctions SLPs spend years training their ears to detect. Asking parents to be the feedback source is asking them to do something they genuinely cannot do.
AI speech recognition can now provide immediate feedback on each production. The child says a word, and the app tells them whether it was accurate. The parent does not need to be the judge, they just need to be present.
LumaSpeech was built for this. SLPs assign targets, students practice with AI feedback at home, and SLPs see exactly what was practiced and how it went.
Without technology, you can reduce the feedback burden by sending audio or video models for reference and by choosing targets where the contrast between correct and incorrect is more obvious for an untrained ear.
Tracking creates accountability and gives you data to adjust the program. If it adds work for families, it backfires.
A home practice program should not be static. Watch for these signals:
Targets may be too hard, sessions too long, or the format too boring. Reduce time, step back to an easier level, or change the format.
Usually means errors are being reinforced. Provide better feedback tools or step back to a level where correct practice is more likely.
Novelty wore off. Refresh the targets, change the format, or have an honest conversation with the family about what changed.
Move targets up. Word level becomes phrases, phrases become sentences. Keep practice at the child's growing edge.
With a caseload of 50+ students, individualized homework for every child every week is not realistic without good systems:
Practice quality matters as much as practice frequency. SLPs are right to worry about reinforcing errors, and the goal is never to lower the bar on accuracy. The goal is to make accurate, supported practice happen more often.
That is why feedback is the linchpin of effective home practice. When children get reliable feedback on each production, errors are caught and corrected immediately. When they practice without it, errors compound. The job of a good home program is not to maximize repetitions at any cost. It is to maximize correct repetitions.
Therapy sessions are a small fraction of a child's week. The hours between sessions are where progress is built or lost. Removing friction from home practice and ensuring feedback is built in gives children more chances to practice correctly, which is what actually drives carryover.
Proven strategies to help students generalize skills from therapy to everyday conversation.
Why immediate feedback matters for motor learning and how AI is changing home practice.