- Research
- Communication & Language
- Low-Tech AAC (symbol boards, visuals)
Do picture boards and communication books help autistic children communicate?
What 7 sources say about low-tech AAC — the evidence, the caveats, and what it means for your family.
Podcast·Two-voice deep dive
Listen to the discussion
Chapters & transcript
This is an AI-generated discussion of this article, transcribed automatically. The written summary below is the reviewed source — if the two differ, go by the summary.
0:00What is this discussion about?
Host A Imagine you're dropped into a foreign country, right? Like a place where you don't speak a single word of the local language.
Host B Oh, that would be incredibly overwhelming.
Host A Totally. You have all these complex thoughts, you know, immediate needs, but your mouth simply cannot make the specific sounds that the locals understand.
Host B Yeah, you're just trapped in your own head, basically.
Host A Right. Then someone hands you this beautifully complex multi-page book of pictures, or maybe just a single board with symbols on it.
Host B Which would feel like a lifeline in that moment.
Host A It really would. But here's the catch. What if the only dictionary you have is completely untested?
Host B Oh, wow.
Host A Yeah, like how do we actually know if that specific visual tool is, you know, the absolute most effective way for you to learn to communicate?
Host B I mean, it is a profound dilemma, and it's one that parents of autistic children with complex communication needs face every single day.
Host A Exactly.
Host B They're handed these tools, usually with immense hope, right? But determining whether the underlying science actually supports those specific tools is incredibly murky.
Host A Which brings us to our mission for you today. Welcome to this deep dive into low-tech AAC.
Host B And just for clarity, AAC stands for augmentative and alternative communication.
Host A Right. And we are looking at this specifically tailored for autistic children. Today, we're actually analyzing two highly specific sources for you.
Host B Yeah, we have an internal audit document and a findings report.
Host A Yep. And that findings report was generated from a demo practice run of the Spectrum Connect method.
Host B Which is really cool because having access to an internal audit is actually pretty rare.
Host A It is, yeah.
Host B It provides this highly transparent, totally unvarnished look at exactly how the evidence for these tools is evaluated. It exposes the systemic roadblocks researchers hit along the way.
Host A For sure. But before we go any further, I do need to share a mandatory but critical disclaimer for you listening.
Host B Always important.
Host A Yep. This deep dive is purely a summary of research findings. It is not advice, and absolutely nothing discussed here is a recommendation for any individual child.
Host B Right. We are just the messengers here.
Host A Exactly. Our goal is to simply look at what the data actually says about non-electronic communication tools, stripping away all the assumptions and the anxieties that usually surround them.
2:18What counts as low-tech AAC?
Host B And separating the strict data from that broader narrative is essential, especially because the source material is incredibly rigid about what actually qualifies as low-tech AAC.
Host A So let's draw those boundaries right now. Based purely on the sources, we are talking about non-electronic-aided AAC.
Host B Meaning basically no screens.
Host A Right. To be perfectly clear, no iPads, no digital tablets, and no speech-generating devices that talk for you.
Host B Furthermore, this specific data run deliberately excludes PECS.
Host A Which is the picture exchange communication system.
Host B Exactly. And it also excludes visual schedules and social stories.
Host A Oh, really? Why is that?
Host B Well, the academic literature categorizes those quite distinctly. They're just treated as different interventions entirely.
Host A Okay, I get that. So what actually makes the cut then?
Host B The sources include things like picture boards, communication books, and core word boards.
Host A Ah, core words.
Host B Yeah.
Host A So those are the boards that use really high-frequency, versatile words like more, stop, go.
Host B Yeah, exactly. And it also includes P-O-D-D books.
Host A P-O-D-D. That stands for Pragmatic Organization Dynamic Display, right?
Host B That's the one. They are these highly organized, often really thick binders filled with just like many, many pages of symbols.
Host A I've seen those. They look pretty intense.
Host B They are. Oh, and the intervention also encompasses manual or keyword signs, but only when they're used alongside these physical boards.
3:48What is aided language input, or modeling?
Host A Okay, so the physical tool is just the baseline, though, right? Because the intervention also heavily emphasizes who is delivering it and, like, the method they use.
Host B Right. The sources describe adult communication partners utilizing something called aided language input.
Host A Or modeling, right?
Host B Yeah, modeling is the more common term.
Host A I actually love the mechanics of this concept. It's kind of like, imagine learning to navigate a complex restaurant menu by watching someone else order first.
Host B Oh, that's a great way to put it.
Host A Right? Because instead of just handing a child a 50-page picture book and expecting them to figure out the syntax on their own.
Host B Which would be impossible.
Host A Exactly. Instead of doing that, the adult actually points to the pictures on the board while they are actively talking to the child.
Host B Yeah, they are demonstrating how the tool operates within a real-time conversation.
Host A It's so cool.
Host B They're essentially providing a bilingual environment. The adult is speaking the verbal language, but simultaneously speaking the visual language of the board.
Host A Which bridges that cognitive gap, right?
Host B Exactly.
Host A Between hearing a word and finding its visual representation.
Host B Exactly.
Host A The theory makes total sense to me.
4:56How much research is there, and what did it measure?
Host A But when we transition to looking at where the research actually came from, the picture gets incredibly complicated.
Host B It really does. Tracing the evidence base back to its roots reveals what the audit document essentially describes as an academic echo chamber.
Host A Wow. An echo chamber. How so?
Host B Well, the volume and type of studies available are highly concentrated. The research is built almost entirely on single case experimental designs.
Host A Okay, hold on.
Host B And the bulk of those primary studies span from like 2006 to 2021.
Host A Wait, single case? Meaning they're literally running an entire published study on just one child at a time?
Host B Yes.
Host A But why? Like, why not just grab 50 kids and run a standard control group?
Host B I mean, think about it. Finding 50 autistic children with the exact same communication baseline, the same motor skills, and the exact same environmental supports, it's nearly impossible.
Host A Because autism is a spectrum, right?
Host B Profoundly heterogeneous.
Host A So heterogeneous.
Host B So single case designs are actually a valid and necessary research method here. Researchers use the child as their own control.
Host A Oh, I see. So they measure the child's communication without the board, then introduce the board, take it away, and measure again.
Host B Exactly. However, while it is valid, relying exclusively on tiny sample sizes severely limits how broadly you can generalize those findings to the wider population.
Host A That makes sense. And the small sample sizes aren't even the only bottleneck, are they?
Host B No, they aren't.
Host A Because the audit actually flags a massive issue with shared authorship across the major reviews.
Host B What's fascinating here is how the same few names just keep popping up.
Host A Right. The names Ganz, Schlosser, and Light appear repeatedly across the syntheses. Like, Ganz authored three of the six captured reviews. Schlosser and Light are on two each.
Host B Yeah, and when you have a highly specialized field with really limited funding pools, the academics who design and pioneer these tools are often the only ones with the resources and the mandate to test them.
Host A But doesn't that create a huge structural weakness?
Host B Oh, absolutely. Because when the same small circle of researchers is conducting the primary studies and then writing the systematic reviews evaluating those same studies...
Host A Independent replication becomes exceedingly rare.
Host B Yes. It creates an illusion of a massive evidence base when in reality you might just be looking at overlapping data from the exact same tiny pool of unique children.
Host A The sources actually hit on a perfect example of this regarding PODD.
Host B Oh, the 2009 paper.
Host A Yeah. The primary paper introducing PODD, published in 2009, was co-authored by Gayle Porter.
Host B And she is the developer of the PODD system.
Host A Exactly.
Host B Which means that the developer is evaluating her own system.
Host A Which means the field is leaning heavily on the creators of the systems to serve as the objective evaluators of those systems. Right. Okay, let's unpack this. If the pool of unique, independently verified studies is this shallow, what on earth are they actually measuring?
Host B That's the big question.
Host A Because, I mean, you can't just run a blood test for communication. How do they actually quantify success over a few weeks with one kid?
Host B So the outcome measures primarily tracked frequency counts of target behaviors.
Host A Wait, frequency counts?
Host B Yes.
Host A So they are literally just standing there with a clicker counting gestures?
Host B Essentially, yes. Researchers counted how often a child independently asked for an item or, you know, how frequently they pointed to a picture.
Host A Wow.
Host B How many times they vocalized a word during a very specific observation window. And we have to look closely at who was actually being counted.
Host A Okay, let's look at that.
Host B A major 2006 review by Millar, which is foundational to a lot of this data, looked at 27 cases.
Host A 27 cases. Okay.
Host B But those 27 cases were individuals with developmental disabilities in general. They weren't exclusively autistic children.
Host A I have to throw a flag on the play here.
Host B Go for it.
Host A We are talking about profound, lifelong communication interventions. If they are only measuring frequency counts with a clicker over a few weeks, how can the literature draw any sweeping conclusions?
Host B It's tough.
Host A Because the sources explicitly state there is zero long-term follow-up captured for low-tech boards alone. Zero.
Host B It is a glaring limitation. And to be fair, it's one the researchers themselves acknowledge in the texts.
Host A No, that's something at least.
9:12Will a picture board stop my child from learning to talk?
Host B However, despite the short-term narrow scope of counting behaviors, the data does reveal specific consistent trends.
Host A Okay, I'm listening.
Host B And these trends directly confront the single biggest anxiety parents express when they are introduced to AAC.
Host A You're talking about the fear that introducing a picture board will act as a crutch, right? Like the child will just stop trying to talk entirely. Does the data validate that fear?
Host B The data strongly refutes it, actually.
Host A Really?
Host B Yeah. That 2006 Millar review analyzing those 27 best evidence cases, it found that zero out of 27 individuals experienced a decrease in speech production after being given a visual communication tool.
Host A Zero out of 27 decreased. Wow. That is a Massive data point for a hesitant parent to hear.
Host B It is. Furthermore, 89% of those cases actually demonstrated measurable gains in speech.
Host A That's incredible. And didn't the Schlosser 2008 review back this up, the one that did look specifically at autistic children?
Host B Yes, it did. It noted that AAC, quote, does not impede speech production, though it cautioned that the actual gains in spoken words were, quote, rather modest.
Host A Okay, so the fear of speech suppression is a ghost, basically.
Host B Exactly. The data absolutely does not support it.
10:31How well do picture boards help children communicate?
Host B But when we pivot to the actual expressive benefits of low-tech specifically, the results are positive, yet maybe less pronounced than you might expect.
Host A Let's look at the math on those benefits. The 2012 Ganz review measured expressive communication using a metric called an improvement rate difference, or IRD.
Host B Right, the IRD.
Host A And for the picture-based low-tech systems, the IRD was 0.61.
Host B Let's break down the mechanics of an IRD so it's not just this like floating number for the listener.
Host A Please do.
Host B So improvement rate difference mathematically compares a child's baseline phase before the board to their intervention phase with the board.
Host A Okay.
Host B An IRD of 0.61 basically means that if you map out all the data points, there is a 61% non-overlap where the communication was clearly quantifiably better while using the board compared to when they didn't have it.
Host A So a 61% clear improvement overlap. I mean... That sounds great.
Host B It does sound good, yes.
Host A But here is the catch. In that exact same review, PECS and speech-generating high-tech devices both scored an IRD of 0.99.
Host B Yeah, that's a huge difference.
Host A Right. So the low-tech boards are trailing significantly behind the alternatives in the data.
Host B They are trailing, yes.
11:44Does pointing to pictures while you talk help, and what about PODD?
Host B However, the data strongly supports the delivery method we discussed earlier.
Host A You mean the aided language modeling?
Host B Exactly. An O'Neill 2018 review found that aided language input, you know, the adult pointing while talking, was highly effective at boosting both comprehension and expression.
Host A And that was across quite a few studies, right?
Host B Yeah, across 26 small studies and two larger ones.
Host A Okay, but what about those massive PODD binders we mentioned? The 2009 article categorizes PODD as, quote, a promising practice.
Host B It does say that.
Host A But... The internal audit document highlights a rather shocking reality here. The article provides absolutely no captured outcome test proving whether it actually works. None at all. It literally just describes the architecture of the book and asserts that it holds promise.
Host B It's wild.
12:35Why is the evidence rated so low?
Host B So we have this really complicated mosaic of findings. We know it doesn't harm speech.
Host A Right.
Host B We see a moderate 0.61 improvement in communication. And we know adult modeling is effective.
Host A But?
Host B But when the Spectrum Connect audit finalized its analysis of this entire data set, the overall confidence level assigned to this evidence base was critically low.
Host A critically low. The formal status in the findings report actually reads, not supported, not enough research yet.
Host B Which is pretty heavy.
Host A It is. Why does an intervention with decades of clinical use score a critically low in an evidence audit?
Host B Well, the fatal flaw lies in how the academic community categorized the data.
Host A Okay, walk me through that.
Host B One of the most significant limitations cited in the sources is that low-tech was almost universally treated as a residual category. It was defined entirely by what it was not.
Host A I have to use an analogy here to explain the mechanical problem this creates for researchers.
Host B Go for it.
Host A Relying on a residual category is like trying to scientifically prove that strawberries cure a cold. But the only data you are allowed to analyze comes from patients who drank a mixed fruit smoothie every day.
Host B Yes, exactly.
Host A You know the patients got better. But mathematically, you have absolutely zero way to isolate the variables and prove the strawberry did the work rather than, I don't know, the banana, the spinach, or the apple juice.
Host B That captures the methodological nightmare perfectly. The major reviews heavily mixed low-tech boards into the exact same data sets as digital tablets and PECS.
Host A It's all just one big smoothie.
Host B Right. Or they lumped simple core boards and complex PODD books into one undifferentiated bucket. Because the data is blended like that, isolating the specific effect size of a non-electronic picture board becomes mathematically impossible.
Host A And the structural weaknesses just compound from there. We already mentioned the tiny sample sizes and the lack of independent replication.
Host B Yeah, the echo chamber.
Host A But these studies also lack control groups entirely. Researchers rarely compared a group of children using picture boards to a matched group of children receiving no intervention.
Host B And because single-case designs are usually only published when they actually succeed, the literature is inherently skewed by publication bias.
Host A It's just layer upon layer of issues.
Host B We also have to analyze the limitations of the demo run itself that produced this internal audit.
Host A Right, because the sources are highly transparent that this was a scoping run.
Host B Yeah, because of massive scientific paywalls and access barriers, the auditors had to rely on a professional body summary of the 2012 Ganz review rather than the full original text.
Host A I mean, they couldn't even bypass the CAPTCHAs on PubMed to scrape the data.
Host B It's so frustrating.
Host A And they couldn't open the specific PODD studies they found. This isn't just a tech glitch. It highlights how privatized paywall databases actively prevent rigorous, broad-scale auditing of medical research.
Host B Completely. When you cannot access the raw data and the data you can access is blended into a residual category smoothie.
Host A It leaves massive fundamental gaps in the science.
Host B It really does.
15:46What don't we know yet?
Host A Which brings us to what the evidence absolutely does not answer. And to be clear for you listening, these are strictly gaps in the scientific literature, not warnings to avoid the tools.
Host B Very important distinction.
Host A Gap number one. We lack long-term data entirely. Yeah, zero. We do not have captured outcomes tracking how a toddler using a core board transitions into a teenager or adult using these specific low-tech systems.
Host B And gap number two, we do not know how well low-tech boards alone impact receptive language.
Host A Receptive language being how they understand what's said to them, right?
Host B Right. Can a board help a child better understand the complex instructions spoken to them? We just don't know. We also lack isolated data on how these boards impact challenging behavior.
Host A And gap number three is perhaps the most crucial for daily family life. Generalization.
Host B Oh, this is a big one.
Host A Right. Because we have no isolated data proving that a child who successfully points to more in a quiet, highly structured clinical room with a therapist can generalize that exact same skill at a chaotic kitchen table or on a noisy school playground.
Host B Exactly. And finally, gap number four. Despite its immense popularity in schools, we have no captured outcome evidence testing if PODD actually works for autistic children. The data simply wasn't present in the literature captured by this run.
Host A So we have critically low confidence, mixed residual categories, no long-term follow-up, and missing generalization data.
Host B It's bleak from a data perspective.
Host A So what does this all mean? How does a speech-language therapist actually function on the ground with a child when the science is this incomplete?
Host B Well, they have to shift to a deeply individualized clinical lens. An AAC specialist doesn't just read a flawed systematic review and base a medical decision on a residual category. They use the available research as a baseline, sure, but they ultimately evaluate the specific physical and cognitive profile of the child sitting in front of them.
Host A So they are essentially forced to build their own N-of-1 data set for every single patient.
Host B Exactly. Clinicians evaluate a child's fine motor skills, like can they physically point to a one-inch square? Right. They evaluate visual processing. Does a 50-page book overwhelm them compared to just a single board? Because the long-term data doesn't exist, the clinician has to engineer a custom plan for how progress will be checked over the coming years.
Host A So they are actively building out strategies for home and school to force the generalization that the studies fail to measure.
Host B Precisely.
Host A It is a heavy reliance on specialized observation to bridge the massive canyons left by the academic literature.
Host B It is. But this raises an important question, and we must emphasize a core tenet of research. An absence of evidence is not evidence of failure.
Host A That's a great point.
Host B Just because a specific low-tech board lacks a massive 500-person randomized control trial does not mean it is ineffective. It simply means the academic machinery hasn't possessed the funding, the sample sizes, or the focus to look closely enough at low-tech boards in isolation.
Host A That is perfectly stated.
18:55What are the key takeaways?
Host A Let's distill all of this down for you. We have covered a tremendous amount of structural science today.
Host B We really have.
Host A Here are the five key findings you should take away from this deep dive. Number one, low-tech AAC does show some help for communication in small studies. However, the data reveals a moderate effect that is less pronounced than the results seen for PECS or high-tech speech devices.
Host B Number two.
Host A Number two. The data strongly challenges the fear of speech suppression. Studies checking for harm did not find that AAC stops a child from learning to talk.
Host B Which is huge. Number three. Aided language input, you know, the bilingual method of an adult pointing to the pictures while talking, was shown to effectively help comprehension and expression.
Host A Okay, and number four is our primary limitation. The academic reviews heavily mix low-tech tools with tablets and PECS into a residual category. This made it mathematically impossible to isolate the specific results for boards and books alone.
Host B The smoothie problem.
Host A Yes, the smoothie problem. And finally, number five, the open questions. The science still completely lacks long-term data, and we are missing evidence on how well these tools generalize from clinical settings to everyday environments like the home and school.
Host B So I want to leave you with a final provocative thought to consider.
Host A Let's hear it.
Host B We spent a lot of time today dissecting how the academic research community is caught in a loop of small, short-term studies restricted by funding and shared authorship.
Host A The echo chamber.
Host B Right. But we are now living in an era where parents are utilizing apps, digital journals, and video tracking to log their children's daily communication progress over years, not weeks.
Host A Oh, wow.
Host B So if the traditional academic machinery cannot overcome its structural bottlenecks to capture long-term generalized data, how might this rising wave of massive parent-led data tracking eventually shatter that echo chamber and force a total reevaluation of how we validate communication tools?
Host A That's a brilliant question. The future of the evidence base might not come from a university lab, but from millions of living rooms.
Host B Exactly.
Host A As we close out today, it is essential to reiterate that decisions about any individual child belong strictly in a conversation with that child's pediatrician or care team, and that this deep dive cannot make that call.
Host B Absolutely.
Host A Research on this topic continues to evolve, and Spectrum Connect will update its coverage as new evidence appears.
Host B Spectrum Connect translates autism research into plain language for parents and checks findings for accuracy against misinformation.
Host A If you are a credentialed expert, MD, PhD, PsyD, or DO, and would like to help validate our findings, email research@spectrumconnect.us.
Host B Thank you for joining us on this deep dive.
Host A We'll see you next time.
Key Takeaways
Mostly very small studies. Picture boards and communication books (no electronics) have mostly been studied one child at a time. We did not find any study that compared children who got a board with similar children who did not.
Some help, less clear than other picture systems. A 2012 review of 24 small studies scored picture systems other than PECS at 0.61 on a 0-to-1 scale of how clearly things got better. PECS and speech devices each scored 0.99. We do not know how many of the 24 studies were about the other picture systems.
No sign pictures stop children from talking. In a 2006 review of the strongest studies (27 children and adults with developmental disabilities, most using hand signs or picture systems without electronics), none talked less and 89% talked more. That is not only autism, and not only boards.
PODD books are described, not tested. We found one 2009 article that calls PODD a promising practice. It explains how PODD is built; it does not report a test of whether it works.
A speech-language therapist who knows AAC can look at how your child communicates now. Talk with your child’s care team before you change a communication plan.
What this means for you
Picture boards and communication books have been studied mostly in very small studies, often one child at a time. The one review that looked at picture systems apart from PECS found they helped, but less clearly than PECS or speech devices.
Many parents worry that pictures will stop a child from learning to talk. Where researchers checked, none of 27 children and adults talked less, and 89% talked more. Those studies mixed many kinds of AAC, so we cannot say this about boards alone.
Who was studied. Mostly autistic children with limited speech, in small single-case studies, usually counting how often a child asked for things, used a picture, or said words over several weeks. Ages were not reported in the summaries we read. Some reviews also included people with other developmental disabilities. We found no long-term results, no results for teens and adults, and no results for boards alone in everyday places like home and school. Generalizability beyond this is not established.
Where the studies landed
Not enough research yet3 Points toward it helping
4 Mixed or inconclusive
Tap any tile to read it. Almost all of these cover AAC more broadly than low-tech boards. Four more sources we found could not be opened and are not counted here.
Points toward it helping3
Mixed or inconclusive4
Spectrum Connect reviews published research on interventions parents are exploring for their autistic children — so you can see where the evidence actually stands. No agenda, no selling, no cherry-picking. Just the studies, our method, and what it means for you.
See the research behind this
For anyone who wants to check our work · 7 sources
Where we looked
This review was a web-based scoping search, not a full database search. We ran two web searches and looked up known papers by title and DOI through Crossref, a free index of journal articles. We did not search PubMed, PsycINFO, ERIC or CINAHL directly, so we have no per-database counts. The string below is the database search we plan to run; you can run it on PubMed yourself.
("Communication Aids for Disabled"[MeSH] OR "augmentative and alternative communication"[tiab] OR AAC[tiab] OR "communication board*"[tiab] OR "communication book*"[tiab] OR PODD[tiab] OR "dynamic display"[tiab] OR "core vocabulary"[tiab] OR "aided language"[tiab] OR "aided input"[tiab] OR "aided model*"[tiab] OR "key word sign*"[tiab] OR "manual sign*"[tiab]) AND ("Autism Spectrum Disorder"[MeSH] OR autis*[tiab] OR ASD[tiab])
Run on PubMed →
Planned databases: PubMed/MEDLINE, APA PsycINFO, ERIC, CINAHL, Linguistics and Language Behavior Abstracts, Cochrane CENTRAL, Epistemonikos and speechBITE, plus dissertations and a hand search of two AAC journals. No protocol was registered in advance. PECS-only and speech-device-only studies are set aside at screening, not by the search string.
What we did with what we found
We read abstracts, plus one summary from a professional body (for the 2012 review). No full papers were read. This is not a PRISMA funnel and the search cannot be reproduced exactly. The 4 sources we could not open were a PODD parent-coaching study, two studies comparing kinds of AAC, and a parent-information page.
What the strongest evidence says
Some help in small studies, but less clearly than PECS or speech devices. The number of studies behind the 0.61 score is not known.
No sign of it. None of 27 best-evidence cases talked less and 89% talked more, but the studies mixed hand signs, picture systems and developmental disabilities beyond autism.
Helped across many AAC users and tools. No test of PODD with autistic children turned up.
Not measured separately for low-tech boards. We found no long-term results.
Keep reading
What would you like to do?
Want to talk to a parent who’s been there?
We match you with families whose kid is like yours. Free, takes two minutes.
Heard of something we haven’t covered?Suggest an article
If another parent, a clinic or a video mentioned a therapy you can’t find here, tell us. We’ll read the research and review it next.
Suggest an articleThink we got something wrong?Report a factual error
We publish the whole record so it can be checked, and that only counts if we act on what you find. If a number looks wrong, a study is missing or has been retracted, or we’ve read a finding in a way the evidence doesn’t support, tell us.
You don’t need a research background to file one. “This doesn’t match what our doctor told us” is a useful report. Every one reaches a person: we reply within seven days, and within thirty we have either corrected the page or told you when we will. Substantive reports send the affected steps back through the protocol and need fresh sign-off before anything here changes.
Report a factual errorAre you a clinician or researcher?Review this article
If you have read this and the evidence holds up, you can sign off on it. We send you a personal review link where you confirm your credentials, and our team checks every review before anything is published. Once approved, your name and a link to your profile appear on this article.
Ask to review this