A child sits across from a therapist. The therapist holds up two cards and asks a question. The child looks at the therapist’s face, then at the cards, then back at the therapist. The therapist glances, almost imperceptibly, toward the correct card. The child points to it.
Data sheet: correct response.
But was it?
The child gave the right answer. But the child did not find the answer independently. The child read the adult’s cue and followed it. Take away the therapist, change the setting, or remove the subtle signal, and the response disappears.
This is prompt dependence. And it is one of the most common ways that autism programs accidentally replace learning with the appearance of learning.
What prompt dependence looks like
Prompt dependence is not always obvious. In fact, that is the problem. It often looks like competence until you change the conditions.
Common signs include:
- The child waits before responding. Instead of acting, they pause, look at the adult, and wait for some kind of signal — a nod, a gesture, a repeated instruction, a shift in body language.
- The child watches the adult instead of the task. Their eyes are on the person, not the materials. They are reading the helper, not solving the problem.
- The child performs better after a second prompt. The first instruction gets no response. The second instruction, delivered with more emphasis or a subtle cue, gets the correct answer.
- The child freezes when the adult changes. A new therapist, a substitute teacher, or even a parent asking the same question in a different tone produces a completely different result.
- The child performs well in one setting but not others. Skills that appear solid in the therapy room vanish at home, at school, or in the community.
None of these mean the child has not learned anything. But they suggest the child has learned to respond to cues rather than to the actual demand.
Why prompt dependence happens
Prompt dependence is rarely intentional. It develops because of how prompts are used, not because anyone decided to create it.
Prompts are given too quickly
When an adult prompts before the child has had time to process the question and generate a response, the child learns that waiting produces the answer. Over time, the child stops trying to respond on their own because the prompt always arrives.
The solution is simple but hard to implement: wait longer. Many children need five, eight, or even ten seconds to process and respond. Most adults prompt within two or three.
Prompts are not faded
Fading means gradually reducing the amount of help over time. A full physical prompt becomes a partial prompt. A partial prompt becomes a gesture. A gesture becomes nothing.
But fading requires planning, patience, and consistent tracking. When fading does not happen — or happens inconsistently — the child stays at whatever level of support was introduced, sometimes for months or years.
Prompted and independent responses are treated the same
If a child gets the same reinforcement for a prompted response as for an independent one, there is no reason for the child to do the harder thing. Independence requires more effort. If the reward is the same either way, waiting for help is the rational choice.
Good programs differentiate. An independent response gets stronger reinforcement than a prompted one. The child learns that doing it alone is worth more.
Success is measured by correct responses, not independence
When a program tracks how many correct responses a child gives but does not track how many were independent, the data can look excellent while the child’s actual learning stays flat.
A child who gives forty correct responses in a session but thirty-five of them were prompted is not making the progress the numbers suggest. Independence data changes the picture entirely.
The transfer test
The most useful test for prompt dependence is simple: can the child do it without the prompt, with a different person, in a different setting?
This is sometimes called the transfer test, and it has three parts.
Without the prompt. If you remove the cue, the gesture, the glance, the repeated instruction — does the child still respond? If not, the child may have learned to follow the prompt, not the skill.
With a different person. If someone else presents the same demand in the same way, does the child respond? If the skill only appears with one specific adult, it is tied to that person’s cues rather than to genuine competence.
In a different setting. If you move the task from the therapy room to the kitchen or the classroom, does the skill hold? If it vanishes when the setting changes, the child may have learned a routine rather than a transferable skill.
A skill that fails all three parts of this test is not yet a learned skill. It is a prompted performance.
Questions to ask a program
If your child is in a therapy program and you want to understand whether prompt dependence is being managed, ask these questions:
“What does your prompt fading plan look like?”
A good program should be able to describe, for any goal, what level of prompt the child currently needs and what the plan is for reducing it. If the answer is vague or the team has not thought about it, that is a concern.
“How do you measure independence?”
Ask whether the program tracks independent responses separately from prompted ones. If the data sheet only records correct versus incorrect — without noting how much help was given — the data cannot tell you whether the child is actually becoming more independent.
“What happens when the child does not respond?”
This question reveals the program’s approach to latency. Does the adult wait? How long? Or does the adult immediately prompt? A program that prompts too quickly, every time, is building the very pattern it is trying to prevent.
“How do you test for generalization?”
Generalization means the skill works across people, places, and materials. Ask whether the program checks for this. If skills are only ever tested in the therapy room with the same therapist, you cannot know whether the child owns the skill or is borrowing it from the context.
“Can you show me data on independence levels over time?”
The most revealing data point is not how many skills a child has mastered. It is how many they can perform independently. If a child has been in a program for months and independence levels are flat — even while correct responses increase — prompt dependence may be building.
What good prompt fading looks like
Effective prompt fading is systematic, not casual. It follows a plan that is documented, tracked, and adjusted based on the child’s performance.
Here is what it looks like in practice:
The prompt level is identified. For each target, the team knows whether the child currently needs a full physical prompt, a partial physical prompt, a gestural prompt, a verbal prompt, or no prompt at all.
Fading steps are planned in advance. The team has a defined path from the current prompt level to independence. Each step reduces the amount of help by a specific amount.
Data tracks the prompt level. Every response is coded not just as correct or incorrect but by the level of prompt that was used. This creates a clear picture of whether independence is actually increasing.
Fading decisions are based on data. The team does not fade based on intuition or hope. They fade when the data shows the child is consistently successful at the current prompt level.
Latency is respected. Adults wait a defined amount of time before prompting. This gives the child a real opportunity to respond independently before help arrives.
Independent responses are reinforced differently. When the child does it alone, the reinforcement is bigger, better, or more enthusiastic than when the child needed help.
Chapter 9 of The Autism Skills Handbook provides a detailed framework for managing prompts, planning fading sequences, and tracking independence. It includes specific criteria for when to fade, when to hold, and when to step back up.
What parents can do
If you suspect prompt dependence is building, you do not need to overhaul the program. Start with observation and questions.
Watch a session. Notice whether the child is looking at the therapist for cues before responding. Notice whether the therapist gives the child time to respond independently before helping.
Test at home. Try presenting a skill the program says the child has mastered. Do it without the usual cues, in a different room, with a different tone. See what happens.
Ask for independence data. Request that the team share data on how much prompting the child needs, not just how many correct responses they are producing.
Ask about fading timelines. For any skill that has been in the program for more than a few weeks, ask where the child started in terms of prompt level and where they are now. If the level has not changed, ask why.
You are not looking for perfection. You are looking for movement toward independence. If prompted responses are steady but independent responses are not growing, the program may need to adjust its approach.
The real measure of progress
Progress in autism support is not how many skills a child can perform with help. It is how many skills a child can perform without help, with different people, in different places.
A child who does three things independently is further along than a child who does thirty things only when prompted by a specific adult in a specific room.
Prompt dependence is not the child’s failure. It is a teaching problem. It is solvable with better planning, better data, and better fading. But it can only be solved if someone is looking for it.
The Autism Skills Handbook dedicates an entire chapter to prompt management, fading plans, and independence tracking. If you want a clear system for ensuring that your child’s progress is real and transferable, the handbook gives you the framework and the tools to make it happen.