Last updated 2026-07-11
TL;DR
Modeling AAC means picking up your child's device or board and using it yourself, out loud, all day, so they see language in action before anyone expects them to produce it. Kids need to watch symbols used across many routines before they start pressing them alone. There's no perfect moment to wait for. Just grab the device and talk.
What modeling AAC actually means
Modeling AAC (augmentative and alternative communication) means an adult picks up the child's communication device, picture board, or symbol system and uses it during real conversation: narrating what's happening, asking questions, commenting. You never require the child to copy you. You're showing, not drilling.
Speech-language pathologists call this "aided language input" or "aided language stimulation," and the idea comes straight from how children learn spoken language. Babies hear thousands of words before they say one. AAC users need that same flood of input in their own system, so they can watch, absorb, and eventually use the words themselves.
This is different from prompting. Prompting sounds like "press the word for more." Modeling is pressing the word for more yourself, in a real moment when more actually applies, and then moving on. No strings attached.
ASHA (the American Speech-Language-Hearing Association) describes aided language stimulation as one of the primary naturalistic intervention strategies for AAC users and lists it as a best practice in its Practice Portal on AAC [1].
Why it matters so much across the whole day
A 30-minute therapy session once a week can't build a language, and neither can drilling flashcards at a table. Language grows in context, and context happens all day. A child learns "more" when more shows up at snack, at play, in the tub, across dozens of real moments. Strip out that repetition and the symbol stays abstract.
A 2007 study by Binger and Light found that preschoolers with complex communication needs produced more multi-symbol messages after their partners used aided language input consistently during play [2]. Consistency across the day was the variable that moved. One caregiver doing it in one room isn't enough.
Researchers often estimate that a beginning communicator may need to see a symbol modeled somewhere between 100 and 200 times, across varied situations, before using it independently. There's no perfectly clean experimental number behind that figure. The closest data comes from Drager and colleagues (2006), who tested symbol acquisition under different input conditions and found that rich, varied aided input beat minimal exposure every time [3]. Frequency and variety both count.
The other reason whole-day modeling matters is blunt: it keeps the device out of the drawer. Research on AAC abandonment keeps finding that devices get shelved when communication partners don't use them [4]. If the device only appears during "speech time," the child learns it's a school object, not a way to talk.
How often to model
There's no gold-standard number, but most SLPs working in naturalistic AAC aim for at least 30 to 50 modeling opportunities a day, spread across different routines. Structured programs like LAMP (Language Acquisition through Motor Planning) and PECS-based approaches carry their own density targets, but the principle holds: more is more, as long as it stays natural and never coercive.
The trick is to hang modeling on routines you already do rather than invent new activities. Model during three meals, one bath, one play session, one transition, and one book, and you clear 40 opportunities without adding a single minute to your day.
Here's a rough guide to how those opportunities stack up:
| Routine | Approximate opportunities | Example words to model |
|---|---|---|
| Breakfast | 6-10 | eat, more, done, want, help, no |
| Getting dressed | 4-6 | help, on, off, no, done |
| Play (15 min) | 10-15 | go, stop, more, again, my turn, fun |
| Lunch | 6-10 | eat, want, done, more, drink |
| Outdoor time | 5-8 | go, fast, stop, look, uh oh |
| Bath | 4-7 | water, more, done, cold, help |
| Bedtime story | 5-8 | more, again, look, no, what, where |
That's 40 to 64 opportunities in a normal day, and none of them ask you to stop what you're already doing.
Which words to model first
Start with core words, not nouns. This is where most parents stumble at first, and it's understandable, because nouns feel concrete and teachable. But researchers including Gail Van Tatenhove and David Beukelman have shown for decades that the words we use most in conversation are small and flexible: go, stop, more, no, want, help, like, all done, look, I, you, that [5].
Core vocabulary makes up roughly 80% of what people say day to day, even though it totals only about 200 to 400 words [5]. Fringe words (specific nouns like "banana" or "Thomas the train") matter too, but a child who can say "want," "help," and "no" holds more communicative power than a child who can name 50 objects and say nothing about any of them.
Most AAC devices and apps put core words front and center for exactly this reason. When you first start, focus on maybe 10 to 15 core words that come up across routines, and add more once those feel familiar. A solid starter set: more, done, stop, go, help, no, want, like, look, that, up, down.
Working modeling into routines you already have
Narrate the moments that already exist rather than manufacturing new ones. Mealtimes are loaded with language: hand your child a fork and model "eat." They reach for the bowl, say "more?" and model it. Plate's empty, model "all done." You're not quizzing them, you're just talking in their language while you talk in yours. Play is the best AAC classroom there is. Get on the floor, grab the toy they're already holding, and model words about what's actually happening. Car rolls: model "go." Car crashes: model "uh oh" and laugh. Game stops: model "stop," then "more" or "again." Keep it playful, because the second it feels like a lesson, the moment dies. Transitions get overlooked, but getting in the car, leaving the park, switching from a screen to dinner are heavy with emotion, which means they're heavy with language. Model "go," "done," "no," "help" here and you're stocking vocabulary for the hardest social situations your child faces. Books work well because you control the pace. As you read, point to the picture, model the word that matches, and keep going. You don't stop and ask your child to repeat it. In every routine, it helps to model a word, pause, and genuinely wait: not a loaded, eyebrow-raised wait that turns into a demand, but a natural beat that leaves your child room to respond if they want to. The AAC literature calls this "expectant waiting," and it's one of the few strategies with steady support across studies [2].
Mistakes that quietly undo the effort
The biggest one is turning modeling into prompting. Modeling means you use the device; prompting means you tell your child to use the device. People mix these up constantly. If you model "more" and then say "now you say more," you've flipped from input to demand, and the research suggests that shift lowers spontaneous communication instead of raising it [6].
Second most common: only pulling the device out during "AAC time." If the board lives in a bag and surfaces for 20 minutes, it's a homework assignment, not a communication tool. The device has to be within reach at all times, whether that means a printed low-tech board in each room, the tablet parked at the table during meals, or a small travel board in the car.
Third: naming things to death. "This is a dog, say dog" is not modeling. Modeling is pressing "look" when the dog runs by, pressing "go" when you call it, pressing "no" when it jumps on someone. Context carries the meaning.
Fourth: expecting fast results and quitting when they don't come. Most children take weeks to months of steady modeling before they start using the device on their own. Stopping at week three because "it isn't working" is one of the most common reasons AAC fails, and one of the saddest, because the child was almost certainly building understanding the whole time you couldn't see it.
None of this is medical advice, just patterns drawn from the research and from what tends to work in real households. If you have concerns about your child's communication, a speech-language pathologist can look at what's actually going on.
Does your child need to watch or respond when you model?
No, and that surprises a lot of parents. You don't need eye contact, you don't need a response, and you don't need your child to even look like they're paying attention.
The evidence backs this up. Children learn a huge amount through incidental, ambient exposure. The same kid who seems to ignore the TV in the next room can quote the dialogue later. Peripheral learning is real. Model the word, say it aloud at the same time, move on, and do it again tomorrow.
That said, being genuinely with your child beats narrating at them. The goal is real conversation in their modality, not a performance. Pressing buttons while staring at the screen yourself is one thing; being present, sharing a look, and pressing the button as part of your actual response to them is another. The second one is richer.
When a child is brand new to AAC, some SLPs suggest keeping input to one word at a time, matching or landing just above the child's current level. This is the "one above" principle, borrowed from spoken language research. If your child isn't using AAC expressively yet, model single core words. Once they start combining words, model two-word phrases, and keep climbing from there [1].
What if you're not the child's therapist?
You don't need a therapy degree to model AAC well. You need access to the device, a basic grip on core words, and a willingness to feel awkward for a while. Ask your child's SLP for a "communication partner training" session. This is standard practice, and most SLPs will run it with parents, teachers, and paraprofessionals. ASHA names communication partner training as a component of AAC intervention because the evidence is clear that partner behavior drives child outcomes [1]. If your child is in an early intervention program, services must include family training under IDEA (Individuals with Disabilities Education Act, 20 U.S.C. § 1400 et seq.) [7].
Ask for a short list of the 10 to 15 core words the SLP thinks make the best starting point, and where each one lives on the device. Then practice finding them fast, until you can hit each one without hunting.
Then just start. It's going to feel awkward. You'll forget. You'll press the wrong button and have to search. Your child might ignore you completely. All of that is normal and none of it means you're doing it wrong.
If your child is just starting out with AAC, early intervention shows how AAC fits into the larger support system, including what services your child is legally owed.
Staying consistent when life is chaotic
Consistency is the hard part, not the modeling itself, but doing it across days, people, and rotten moods.
A few things help. Tie modeling to anchors, the routines you do every single day no matter what: morning snack, bath, bedtime book. Three anchors alone give you a floor of 15 to 25 modeling opportunities on your worst day.
Placement matters more than people expect, too. Charged and sitting on the kitchen counter, you'll use it at breakfast. Zipped in a bag in another room, you won't. Figure out where your child spends the most time and put their system there.
Keep low-tech backups as well. A paper core board taped to the fridge or the bathroom wall means you can still model when the battery dies or the tablet is at school. Many SLPs help families make these, and there are free printable core boards from sources like PrAACtical AAC.
And pull in everyone who spends real time with your child: teachers, grandparents, therapists, older siblings. One consistent caregiver beats zero, but the input compounds when several people use the same system.
If you're working with a therapist remotely, online speech therapy can be a way to get regular coaching on your modeling without weekly in-person visits.
Does this look different for autistic kids?
The core approach holds across populations, but a few things shift for autistic communicators.
Many autistic children are strong visual learners and may pick up symbol associations faster than expected, while generalization (using a word in a new setting) takes longer. So you might see your child use a word at the table but not in the car, even after months of modeling in both. That's not a plateau, that's how the learning works. Keep modeling everywhere.
Some autistic children who use AAC also use echolalia. If you're unsure how the two fit together, it's worth reading up on echolalia on its own, because it often coexists with AAC and the strategies sit alongside each other rather than clash.
Motor planning is another variable. For autistic children who also have significant motor difficulties (sometimes seen alongside childhood apraxia of speech), the physical act of getting to the word matters a lot. In those cases, work closely with your SLP on the access method (touch, eye gaze, switch scanning), and let modeling account for the motor demands. For a wider view of communication approaches, autism spectrum speech therapy covers the options and how they interact.
One thing that never changes: no response and no eye contact during modeling is not a signal to stop. For autistic children especially, ambient input is often getting processed even when the child looks checked out.
How do you know it's working?
Progress in AAC rarely looks like a switch flipping. It looks like small shifts over weeks and months, and they tend to arrive in a rough order.
First, the child starts looking at or reaching toward the device more often. That's receptive recognition, not expressive use, but it counts. Then the child starts touching symbols without a prompt, even if the presses don't look intentional yet: pre-symbolic exploration, and it matters. Next, the child uses a symbol in one context where they've seen it modeled, usually a high-motivation word like "more" or "go." Eventually the child starts combining two words: "Want more," "Go now," "No stop." That's a real milestone, and it usually means the system has become theirs.
Data catches progress you can't see day to day. Keep a simple tally of how many different symbols your child starts on their own each week, not counting prompted presses. A sticky note on the fridge works. If that number climbs slowly over 6 to 8 weeks, the modeling is working.
See no change at all after 8 to 12 weeks of consistent daily modeling across multiple routines? Bring that data to your SLP. It might mean the layout needs adjusting, the vocabulary isn't motivating enough, or there's an access problem worth solving. It does not mean AAC can't work for your child.
An app like Little Words can help you track these patterns daily and build a starting-point vocabulary profile based on your child's current level, which then tells your SLP what to prioritize in modeling sessions.
What the research says about timelines
Studies vary, but they agree on one thing: children need sustained exposure across weeks to months, not days. A systematic review by Schlosser and Wendt (2008) looked at speech production outcomes for children with autism using AAC and found that naturalistic aided input was among the more effective strategies, with stronger effects in studies running 12 weeks or longer [8].
Binger and Light's 2007 study on modeling two-symbol combinations found that three children with complex communication needs all increased their multi-symbol use during play after a structured aided input intervention. Time to first spontaneous production ranged from 3 to 8 weeks of regular sessions [9].
For parents doing home modeling without a formal protocol, real-world timelines are harder to pin down, since how faithfully it's done varies from house to house. The honest answer: expect 2 to 6 months of consistent daily modeling before you see reliable spontaneous use, and treat anything earlier as a bonus.
That's genuinely hard to sit with. Two to six months feels like forever when you're watching your child struggle to say what they mean. The research doesn't offer a shortcut, but it does keep saying the investment pays off.
Frequently asked questions
What is aided language stimulation, and is it the same as modeling AAC?
Yes, they're the same thing. Aided language stimulation (sometimes called aided language input) is just the clinical name for what most parents call AAC modeling: a communication partner uses the child's AAC system during everyday interactions to show how it works. ASHA lists it as a core evidence-based practice for AAC users, and most speech therapists use the two terms interchangeably.
Can I model AAC even if my child isn't using their device at all yet?
You can, and honestly this is the best time to start. Kids need to take in language before they produce it, the same way they do with spoken words. Modeling during this quiet, pre-expressive stage builds the groundwork your child will lean on once they start using the system themselves. Waiting for them to "show interest" first just delays that groundwork for no real reason.
How do I model AAC without it turning into a test?
Drop any expectation that they'll respond. Press the word, say it out loud, then get back to whatever you were doing. Skip the expectant stare, the repeated prompts, the "now you try." If it feels relaxed to you, your child will feel that too. It helps to practice during easy, low-stakes play before you try it during a transition or a meltdown.
Should I still model AAC if my child can say some words out loud?
Yes. AAC and spoken language aren't competing with each other. Research shows AAC doesn't reduce or replace speech, and it often supports it by taking some pressure off communication. A child whose verbal words come out unreliably may find steadier footing with AAC in certain settings, and the two can reinforce each other over time. ASHA's guidance addresses this directly.
What if I hit the wrong button or can't find the word fast enough?
That's completely fine. Laugh it off, say "oops, wrong one," find the right word, and keep going. A messy model still counts as a model. Fear of getting it wrong is one of the biggest reasons parents hesitate, and that hesitation costs far more than any fumbled button press ever will. Your own fluency with the device builds with practice, same as your child's will.
How do I get my child's school to model AAC during the day?
Start with the IEP. AAC goals should call for communication partner training and specify that the device gets used across school settings, not just during pull-out speech sessions. Under IDEA, schools are required to provide the supports written into the IEP, and that can include staff training. Ask directly for the SLP to train teachers and paraprofessionals on aided language input. If it's still not happening in the classroom, write it down and bring it to the next IEP meeting.
At what age should a child start using AAC?
There's no minimum age. Research and clinical consensus support introducing AAC as early as infancy for children with significant communication needs. The old assumption that a child needs to hit some cognitive milestone first has been thoroughly rejected. Earlier access to a communication system is linked to better long-term language outcomes, and ASHA's AAC evidence base includes studies with children under 24 months.
Does modeling look different on a high-tech device versus a picture board?
The strategy itself doesn't change: model the symbol, say the word, move on. Low-tech boards are often easier to model from since they're quick, never need charging, and travel anywhere. A lot of families run both systems at once, using the high-tech device as the primary tool and printed boards in the kitchen, car, or bedroom as backup. Most SLPs will tell you not to wait for the high-tech device to arrive before you start modeling with paper.
My child throws or avoids the AAC device. How do I model if they won't let it near them?
Start by fully separating modeling from any expectation to respond. If your child has learned that the device means demands are coming, they may be avoiding the demand, not the tool itself. Pull back on prompting altogether and spend a few weeks just modeling with zero pressure. It's also worth checking whether the device became unpleasant through too much drilling. And it can help to try a different setup entirely: a smaller board, a new spot for it, or something your child can move away from without feeling trapped.
What's the 'one above' principle in AAC modeling?
It means modeling language one step above wherever your child is right now. If they aren't yet using AAC to express themselves, model single core words. If they're combining one word at a time, model two-word phrases. This mirrors the scaffolding used in spoken language development and draws on Vygotsky's zone of proximal development: challenging enough to push language forward, close enough that your child can still follow.
Does AAC modeling work for children with apraxia of speech?
Yes. For children with childhood apraxia of speech who also use AAC, modeling gives them a way to communicate that isn't dependent on precise motor control for speech sounds. The modeling itself stays the same, though the AAC layout may need adjusting for motor planning, which is where approaches like LAMP come in. A speech therapist who knows both apraxia and AAC is the right person to help coordinate this.
How do I model AAC in a two-language household?
Bilingual AAC is a real and growing area of practice. ASHA and AAC researchers generally recommend supporting both languages: model in whatever language the interaction is happening in, and stock the device with vocabulary from both languages where you can. Don't drop the home language to make AAC simpler, that's not necessary. If the device itself doesn't support the home language, work with the SLP on low-tech materials in that language instead.
What's the difference between modeling AAC and PECS?
PECS (Picture Exchange Communication System) is a structured, staged program where kids learn to exchange picture cards to request things, starting with a physical hand-to-hand exchange and building from there. Aided language modeling is a broader, more natural strategy that works across many AAC systems rather than a fixed program. Plenty of children use pieces of both. PECS has its own protocol and research base, while modeling can layer on top of PECS or any other approach.
Sources
- ASHA, Practice Portal: Augmentative and Alternative Communication: ASHA lists aided language stimulation as a primary naturalistic intervention strategy for AAC users and includes communication partner training as a component of AAC intervention.
- Binger, C., & Light, J. (2007). The effect of aided AAC modeling on the expression of multi-symbol messages by preschoolers who use AAC. Augmentative and Alternative Communication, 23(1), 30-43.: Preschoolers with complex communication needs produced more multi-symbol messages when partners used aided language input consistently during play; expectant waiting has steady support across studies.
- Drager, K. et al. (2006). AAC technologies for young children. Augmentative and Alternative Communication, 22(1), 21-35.: Rich, varied aided input consistently produced faster symbol acquisition rates compared to minimal exposure conditions.
- Johnson, J. M. et al. (2006). Perspectives of AAC users and their communication partners regarding request for repair strategies. Augmentative and Alternative Communication.: AAC abandonment research consistently identifies lack of communication partner use as a primary factor in devices being put away.
- Beukelman, D. & Mirenda, P. Augmentative and Alternative Communication (4th ed.), Paul H. Brookes Publishing: Core vocabulary of roughly 200-400 words accounts for approximately 80% of daily communication across ages and contexts.
- Snell, M. E. et al. (2010). Twenty years of communication intervention research with individuals who have severe intellectual and developmental disabilities. American Journal on Intellectual and Developmental Disabilities, 115(5), 364-380.: Prompted communication requests, when overused, are associated with reduced spontaneous communication compared to naturalistic modeling approaches.
- Individuals with Disabilities Education Act, 20 U.S.C. § 1400 et seq.: IDEA requires early intervention services to include family training as a component of the individualized family service plan (IFSP).
- Schlosser, R. W., & Wendt, O. (2008). Effects of augmentative and alternative communication intervention on speech production in children with autism. American Journal of Speech-Language Pathology, 17(3), 212-230.: A systematic review found naturalistic aided input was among the more effective AAC strategies; effects were stronger in studies lasting 12 weeks or more.
- Binger, C., & Light, J. (2007). The effect of aided AAC modeling on the expression of multi-symbol messages. Augmentative and Alternative Communication, 23(1), 30-43.: Time to first spontaneous multi-symbol production ranged from 3 to 8 weeks across three children in the aided modeling condition.
- American Academy of Pediatrics: AAP endorses early communication supports and opposes delaying AAC introduction based on cognitive prerequisites, consistent with ASHA guidance.
- ASHA, IDEA (Part B and Part C): ASHA describes the legal basis for AAC services and communication partner training under IDEA in school settings.