How to measure AAC use (spontaneous vs prompted)

By the IEP Atlas editorial teamLast updated November 1, 2026

Short answer

In AAC data collection, a spontaneous message is one the student sends with no adult cue toward the device or the word, and a prompted message follows a cue such as a question, a point or a model. Ask: in the 10 seconds before the message, did an adult signal what to say or that it was time to say it? If yes, code it prompted.

Most AAC data sheets fail the same way: every message is a tally mark. A student who taps "more" after a para points at the device and says "tell me" gets the same mark as a student who walks across the room to say "swing go." The totals go up, the progress report looks good, and nobody can tell whether the student is communicating more on their own.

Splitting spontaneous from prompted messages fixes this. It is a coding decision, made once, written down and used by every adult who takes data. This page sets out the two categories side by side, adds a middle category many SLPs find useful, and works through three classroom cases. For goal wording, see the AAC goals examples page. The speech therapy IEP goals pillar covers the rest of the field.

The difference in one sentence

A spontaneous message starts with the student; a prompted message starts with something an adult did to get it.

The deciding question: in the 10 seconds before the message, did an adult do anything that told the student what to say, or that it was time to say something on the device? If yes, it is prompted. If no, it is spontaneous.

The 10-second window is a common choice, not a rule. Some teams use 5 seconds, some use 30. Pick one, write it in the goal's definition and keep it the same all year.

Spontaneous vs prompted AAC use, side by side

	Spontaneous	Prompted
Definition	Student sends a message with no adult cue toward the device or word in the set window	Student sends a message after a question, pause, point, verbal cue, model or physical help
What it tells you	Whether the student uses AAC on their own, when they want to	What the student can do with support, and how much support
Typical goal wording	"will initiate 3 or more messages per 30 minutes without a prompt"	"will request using the device given no more than a gestural prompt"
Basis in law or practice	The IEP must state how progress is measured;^[1] the coding rule is how you measure it	Same, plus the prompt level, which is part of the goal's condition
Who decides the definition	SLP and team, written into the goal or data sheet	SLP and team, with a prompt level list everyone uses
Data method	Frequency per timed block	Frequency or opportunities, coded by prompt level
Example	At recess, the student taps "swing go" with nobody speaking to them	Para points to the device and says "tell me what you want"; student taps "cookie"
Where it is documented	Present levels, progress reports, data sheet	Same, plus the prompt level column on the data sheet
Risk if confused	Progress looks slower than it is if light cues count against the student	Progress looks faster than it is if heavy cues count as independent

Add a middle category: responsive

Many AAC users answer questions well but rarely start a conversation. A third code captures that.

Spontaneous (S): student initiates. No adult turn in the window.
Responsive (R): student answers an open adult turn, such as "What should we do now?", that does not name the word, the device or the answer.
Prompted (P): adult cue that points to the device or the word. Code the level: expectant pause (EP), verbal (V), gestural point (G), model (M), physical (PH).

Use the same prompt hierarchy your team already uses for other goals. The prompt hierarchy examples page lists levels from least to most help.

A worked data sheet

Rosa, a fictional second grader, uses a speech-generating device. Her SLP watches a 30-minute block of centers on 3 days.

Day	S	R	EP	V	G	M	Total
Mon	2	4	3	1	2	2	14
Wed	3	5	2	2	1	3	16
Fri	3	3	4	1	1	2	14

Her median is 3 spontaneous messages per 30 minutes, out of a median of 14 total messages. That is about 21% of all messages, and it is the baseline for a spontaneous initiation goal. The total of 14 alone would have hidden how much of Rosa's talk depended on adults.

The frequency data sheet works for this if you add one column per code. Read more about counting events in a set time on the frequency recording page.

Three classroom scenarios

Scenario 1: the point and the question

At snack, the para holds Rosa's device in front of her, points to the food page and says "Tell me what you want." Rosa taps "cookie."

Code: prompted, verbal plus gestural. The para named the task and pointed to the place. This is a good teaching moment, but it is not independent use.

Scenario 2: the expectant pause

During play, Rosa's teacher holds a bubble wand, opens the jar and waits 5 seconds while looking at her. Rosa taps "blow."

Code: prompted, expectant pause, under the rule above. The adult set up a moment that signaled it was Rosa's turn. Some teams treat a natural pause as a spontaneous opportunity. That is a defensible choice, but then every adult must code it the same way. The expectant pause code keeps the lightest cue visible, and it shows the team when Rosa no longer needs it.

Scenario 3: the walk across the room

At recess, Rosa walks to the aide on the bench and taps "swing go" while nobody is talking to her.

Code: spontaneous. She chose the moment, the partner and the message. This is the behavior a spontaneous initiation goal is chasing.

A fourth case comes up often: an adult modeled "more" 2 minutes ago during a different activity, and Rosa now uses "more" at snack. Outside the 10-second window, that is spontaneous. Modeling earlier in the day is the point of aided language modeling, and using a modeled word later is what success looks like.

When you need both measures

Most AAC students need both numbers on the same sheet.

Early on, prompted data shows how much help the student needs. Track the prompt level falling.
As the student grows, spontaneous data shows whether they use AAC without being asked.
At progress reports, report both: "Rosa sent a median of 3 spontaneous messages per 30 minutes (baseline 1) and 6 responsive messages. Prompted messages fell from 9 to 5."

Track the adults, too. If adults model less than they did in the fall, spontaneous use often drops. The goal measures the student, but the data sheet can carry a tally of adult models per block so the team can see why. The IEP can name supports for school personnel, such as training on modeling and prompting.^[1]

Mistakes that blur the data

Weak: One tally column for every message. Better: Separate columns for S, R and each prompt level.

Weak: "Independent" left undefined on the data sheet. Better: A written window (10 seconds) and a list of what counts as a cue.

Weak: Only the SLP codes, in the therapy room. Better: Train the teacher and para on the codes, and collect in class, at lunch and at recess.

Weak: Counting device exploration (babbling through buttons) as messages. Better: Count a message when it has a clear communicative purpose: a request, comment, protest, greeting or answer.

AAC is a communication system without prerequisites, and use across the whole day is the aim.^[2] Baseline methods for AAC goals are on the AAC baseline page, and general session methods are on the speech therapy data collection page. Template goals are on the AAC goals page.

Frequently asked questions

Is a response to a question spontaneous AAC use?

Not under most definitions. An answer to an open question that does not name the word is responsive, a middle category. It shows the student can take a turn, but it does not show they can start one. Code it separately from both spontaneous and prompted messages.

How long should the window be for coding a prompt?

Many teams use 10 seconds, meaning a cue in the 10 seconds before the message makes it prompted. There is no fixed rule. Write your window into the goal's definition and keep it the same for the whole year.

Should paraprofessionals take AAC data?

Yes, because they see the student in more settings than the SLP. Give them a sheet with one column per code, a written window and a short practice session coding the same video with the SLP.

Use it in IEP Atlas

Sources

34 CFR §300.320: Definition of individualized education program, Legal Information Institute, Cornell Law School
Augmentative and Alternative Communication (AAC), ASHA

Written and checked by the IEP Atlas editorial team under our editorial policy. This guide describes general practice and federal rules. It is not legal advice; your state may add requirements, and the IEP team makes the decisions for each student.