Why Are You Using Picture Exchange Communication System (PECS)?

30 years ago, in 1991, I became a speech pathologist.  Very quickly, I became heavily interested in Augmentative and Alternative Communication (AAC) and started to work more and more with individuals with complex communication needs.

In the beginning, I wasn’t very good at implementing AAC – partly because I was such a new graduate and partly because the field was still learning an enormous amount about what we should do!!

The first big change that happened in my AAC world, changed me forever as a clinician.  Carol Goossens’ visited Australia and talked about aided language displays and that we needed to “speak AAC for individuals to learn AAC.”  This was a completely new concept for me – and for those I worked with.  We started making aided language displays, using them in multiple situations – and we saw many individuals with complex communication needs really blossom as communicators.  Carol taught us so much – from how to engineer the environment for communication success, to thinking about the range of vocabulary needed in a situation for an individual to participate and to have multiple communication turns. (See Goossens’, C., Crain, S., & Elder, P. (1992) for more information)

The next big change that came along, was PECS.  It started being used in Australia in the mid to late 90s and quickly gained a lot of popularity in use with autistic students.  PECS came at communication from a completely different perspective to the approach introduced by Goossens’.  Where aided language displays were based in natural communication interactions, “PECS relies on the principles of applied behavior analysis (ABA) so that distinct prompting, reinforcement, and error correction strategies are specifed at each training phase in order to teach spontaneous, functional communication.” (Bondy & Frost, 2001, p.728). 

And so, in the late 90s I attended PECS training, listened to their rationale and started implementing it with some students.  Compared to aided language displays, I saw a smaller number of individuals where PECS made an impact on their communication – and for many of them the initial positive changes were often negated over time by other changes.  For example, the structure of PECS teaches an individual to request - and sometimes the ability to request and receive their favourite items reduces challenging behaviour.  Over time, you fade the requests so that students “don’t always get what they want”.  At that point, we often saw a big increase in challenging behaviour again – and the more I thought about it, the more it made sense.  At no stage in PECS do we teach an individual how to protest or reject in a socially appropriate way.*  So – once we got to the stage where we were having to tell an individual that they couldn’t get items every time they requested them, they had to revert to body language, facial expression and behaviour to express their dissatisfaction with this change!  This was not ideal. And so, due to the limited positive outcomes and the unfortunate negative outcomes, I started to use PECS less.(*Edited to add that I have now had a number of people tell me that PECS does, in fact, include a way of teaching saying "no" at Stage 3 so my apologies for being so definite that that was not in the program.)

Meanwhile, the rest of the AAC world was developing. More varied and well designed AAC systems were being developed.  We were getting more and more information about how to introduce and implement them in ways that help each individual move towards becoming an autonomous communicator.  Today we have a number of well designed systems and options and a constantly growing evidence base around using AAC.

A few years ago, when asked my opinion of PECS I would say “I only use it with a very small number of students with autism.  There’s no evidence base for using it with other groups.  The meta-analyses which review PECS are very mixed and not overwhelmingly positive.  I also believe that for any individual PECS would only ever be an introductory system because it doesn’t have a range of language available, so it doesn’t support individuals to be independent, autonomous communicators.”  And then, because people would often tell me that PECS teaches more than requesting – I would usually specifically mention that it never teaches protesting or rejecting and that that is a skill we all need to learn to do in a socially acceptable way.

(And, if you are unsure what a meta-analysis is, please read the information at this link.)

Recently, I have been thinking more about PECS again. It’s been several years since I’ve used it. I find good comprehensive AAC systems give individuals with complex communication needs better outcomes. They provide language for them to communicate with all the different communication functions that we need across the day,   and because we can implement them in a naturalistic way through aided language stimulation, individuals learn how to use that language in a range of contexts and settings – in the same way everyone else learns to communicate.

And before I go any further I want to explain that when considering any clinical intervention as a speech pathologist, we use three elements together to make a decision based on evidence based practice. These are shown below.

One of these circles is the reason that I have returned to thinking about PECS. This is the circle called "client perspectives" - and there is currently a steady dialogue coming from autistic people about ABA and programs like PECS which are based on the principles of ABA.  If you haven’t caught up with this you can read about it online – and this is one article which gives a balanced summary.

So, if autistic people are telling us that ABA and programs based on ABA principles, like PECS, are restrictive practice, should we still be doing PECS?  One therapist asked this recently in a Facebook group I am in – and was deluged with a mix of positive and negative feedback.  I noticed that a lot of people had opinions – but very few people’s opinions were based on anything other than their personal experiences in which they stated that PECS did or didn’t work or how much they did or didn't like it. And while clinical expertise is one of the three components of evidence based practice, we are also supposed to use best external evidence as well.

So, as part of my inner musings about PECS, I decided it was time to look at the evidence base again.  A few years ago, when I looked at the meta-analyses which focused on PECS, I found very mixed results and opinions about outcomes and efficacy.  However, recently the National Disability Insurance Agency (here in Australia) has commissioned a report into the best available high-quality evidence about interventions for children with autism aged up to 12 years.  The report is published by the AutismCRC, and you can access the report via https://www.autismcrc.com.au/interventions-evidence.

The team of researchers involved in writing and compiling the report have thoroughly reviewed the evidence in Interventions for Children on the Autism Spectrum. As part of this, they have even done an Umbrella Review – which is basically a meta-analysis of a group of meta-analyses.  Their stated aim is “to support consumers to make informed choices about which intervention may best suit the needs of their family”.

If you download the full report, you will find pages and page of information. But the parts that interest us for this blog post, are the summary tables. 

The table above includes the information about PECS.  Since PECS is based on ABA it is included with behaviour interventions. You can see that PECS has:

  • a positive therapeutic effect on social-communication, low quality evidence
  • a null therapeutic effect on expressive language, low quality evidence; and
  • inconsistent therapeutic effect on general outcomes, low quality evidence.

And the table above includes the information on all other forms of AAC except PECS. 

You can see that AAC (except PECS) has:

  • inconsistent therapeutic effect on social communication, low to moderate quality evidence
  • positive therapeutic effect on communication, low quality evidence
  • positive therapeutic effect on motor skills, low quality evidence
  • positive therapeutic effect on social-emotional/challenging behaviour, low quality evidence
  • positive therapeutic effect on play, low quality evidence
  • inconsistent therapeutic effect on general outcomes, low quality evidence
  • positive therapeutic effect on academic skills, low quality evidence
  • inconsistent therapeutic effect on caregiver satisfaction, medium quality evidence

This list and the table clearly show that the evidence is much stronger for using all other forms of AAC - other than PECS. (Although it also shows we still have a long way to go.)

So, from today if anyone asks me about using PECS I will now reply: “I don’t use it anymore.  The evidence base for autistic children is much stronger for all other forms of AAC – and that’s what I use.”  It’s a much more definite statement than I’ve made before – but I feel very supported to say this with the weight of the evidence base - client perspectives plus clinical expertise backed up by external scientific evidence - behind me.

And to finish this blog post, I'd like to come back to my original question.

Why are you using PECS?

 

References

Bondy, A., & Frost, L. (2001). The Picture Exchange Communication System. Behavior Modification, 25(5), 725-744. doi:10.1177/0145445501255004

Goossens’, C., Crain, S., & Elder, P. (1992) Engineering the Preschool Environment for Interactive, Synbolic Communication. Birmingham, AL: Southeast Augmentative Communication Conference Publications

Whitehouse, A., Varcin, K., Waddington, H., Sulek, R., Bent, C., Ashburner, J., Eapen, V., Goodall, E., Hudry, K., Roberts, J., Silove, N., Trembath, D. (2020). Interventions for children on the autism spectrum: A synthesis of research evidence. Brisbane, AutismCRC

Spread the word. Share this post!

Comments (49)

    • jane

      Reply

      Thanks Susan! It’s been swimming around in circles in my brain for a while. And today it felt like coming out!!

  1. Laura Taylor

    Reply

    As a classroom teacher for many years we “used” PECS and didn’t see any changes in our students communicative abilities. One year in a classroom with older students with autism proved to me that robust communication is the only way to go. Thanks for all you do!

  2. Barbara Coven-Ellis

    Reply

    Why do I use PECS? Because in our clinic we have limited access to a variety of other more robust AAC's for initially working with these non-verbal kids. But with PECS we can get them starting to understand the idea of requesting, direct interaction with someone in order to gain access to said item/activity, the need to actually notice the other person in the room in order to accomplish functional communication, and a decrease in frustration at home as it is simple to set up for families that have limited funds, which is most of our patient base. Here in the US, I may wait over a year for a device to be approved and delivered. With that in mind, I want to start using something that can kickstart direct communication, increase communication at home, and decrease frustration. I will often transition to an SGD, but not always. Some of my kids become way too interested in trying to find ways to get to other programs/shows/apps/the web than using the device for any kind of communication so it "goes away" while we work on communication through some sort of icon exchange or other noncomputer based communication. PODS are used a lot around here and frankly, I hate them. They have way too many choices and are just too complex for most of the kids we see in our clinic. We are rural. We are hospital based. We are non-profit. We get the kids that the private clinics don't want because there is no insurance or it's state insurance so it barely pays, or because they don't want to deal with kids that have multiple disabilities. As a result, I do not remove anything from my toolbox because I never know what's going to walk through that door and I never know what will be best for any individual child and family. If PODS is what is being used at school, I find a way to work with it. If they've been given a device which they have no idea how to use, I try to teach to the device/software. Sometimes I have to say that the school's choice was not a good one and we have to look at different programs/devices.
    I am the first to admit that AAC is not my specialty. I specialize in CAS and phonological disorders, as well as pediatric feeding disorders, but I work in a clinic and whatever is placed on my caseload is who I am going to see. With this in mind, I want my toolbox to be as deep and to have as many options as possible.

    And THAT is why I use PECS. 🙂

    • jane

      Reply

      HI Barbara, thanks so much for your response. Hopefully you can move to other forms of low tech AAC, such as Project Core, which are also inexpensive but which fall into the category of other AAC. If you haven't seen Project Core (I just mention this because you say you aren't an AAC specialist) please check it out at http://www.project-core.com/. It has lots of free resources to get people going and would be a very valuable resource for your toolbox if you aren't already using it.

    • Mary

      Reply

      I agree with much of what is said here.PECS still has a place in my mind at least for beginning communicators. PECS continues to be the best method we have to teach initiation of communication and to help complex communicators understand that communication ONLY takes place when a message is delivered or "exchanged" to a communication partner.

      • jane

        Reply

        HI Mary, I used to agree with you. But my journey is that I discovered that bringing communication to them via aided language stimulation has been even more successful, especially because it builds receptive communication so well - and it feels more respectful as well since I can join them where they are at. But I do understand what you are saying since I used to think it too!

      • Erin Troost

        Reply

        Yes but that places the onus on the sender and negates the equally important role of the receiver. For emerging nt communicators the message is often in the interpretation of vocalizations, gestures, physical manipulation (a baby pushing away food that is then removed - they do it again the next time) - future initiation is likely the byproduct of successful interpretation by receivers. If we undermine non-symbolic potential messages in order to force something so specific as a picture exchange we may be inhibiting the individual's natural initiate development.

        • jane

          Reply

          HI Erin, I’m very confused by your comment. At no point do I ever say that we would stop using good communication partner practice and attributing meaning and assuming competence with emergent (receptive) communicators. We absolutely need to continue to do this. Jane

    • Lilo Seelos

      Reply

      Well said, Barbara. I 100% agree with you. PECS is a tool in our tool box and it I have used it with great success with the type of client group you describe. And I hate PODS, too!

        • jane

          Reply

          Jill - PODDs stands for Pragmatically Organised Dynamic Displays. It's an AAC language organisation system that is becoming increasingly popular - really supports individuals in developing communication autonomy. Lots of information online if you Google it.

  3. Emma

    Reply

    Once again, you’ve said it better than I ever could. Thank you for such a detailed and reference heavy post. I will be referring to this post when I tell people why I don’t use PECS. Thank you!

  4. Harriet Korner

    Reply

    Thanks Jane! I really like the clarity you bring with this explanation - this new umbrella review will be very valuable for helping us to find out overall what the evidence is saying! And it helps answer your question. I don’t use PECS, but when I get asked about it, I’ll explain the “why” more clearly now too !

    • jane

      Reply

      Thanks Harriet. And thanks too for moderating the session about it at ISAAC Connect. David and Hannah did a great job explaining it I think.

  5. Connie

    Reply

    "At no stage in PECS do we teach an individual how to protest or reject in a socially appropriate way" except that teaching learners to say "no" or to reject is in fact one of their 9 Critical Communication Skills and it is recommended to be taught at around Phase 3. You can find that information in the PECS Training Manual.

  6. Claudia Marimón

    Reply

    Thanks Jane, always learning with your contributions.
    Your explanation is very clear. And based on evidence.

  7. Julie

    Reply

    It is wonderful to see the evidence so concisely summarised. That umbrella study is very valuable! Thank you Jane. Having said that, the evidence for other AAC isn't overly strong, is it?! Do you find this makes it hard to speak about AAC interventions other than PECS with too much confidence?

    • jane

      Reply

      HI Julie - it isn't overly strong - just stronger than the evidence for PECS with evidence it addresses more areas on their table. Two of the researchers presented on the umbrella study last week at ISAAC Connect. They talked about the fact that in AAC it's difficult to get high quality evidence because the population of people with CCN is so heterogenous. The randomised control trials that are best practice in so many areas are rarely used in the field of AAC because of this heterogenity. Since this is the nature of individuals with CCN I think this is going to be a continuing dilemma - and it would be awesome to see some other way of rating quality of evidence that takes our population into account - but that's one for the researchers. I do have a memory of Pammi Raghavendra presenting at an ISAAC Conference years ago suggesting an alternative rating but my memory could be faulty. I'll have to ask her!

      • Julie

        Reply

        Thanks for your response, Jane. This definitely helps to explain the difference between the strength of research/evidence and what we are seeing on the ground. This article was perfectly timed. Looking forward to seeing the future research that rolls out in this area!

        • jane

          Reply

          Thanks Julie! I love how rapidly AAC is changing and how much we are learning. Looking forward to changing and learning more too!!

      • Andy Bondy

        Reply

        Given the comment about PECS and other AAC strategies, let me ask a question about specific research. Is there a study focused on ANY AAC strategy involving many (about 20) very young (mean age 2.5 yrs) diagnosed as ASD using 10 or fewer spoken words at the start of intervention, using a RCT format to randomly assign to AAC or speech-focused training, yielding the development of a reasonably large vocabulary (let’s say over 80-120 spoken words) in a reasonably short time (let’s say 6-9 months) involving strong parental training and implementation? If such a study existed- pertaining to Aided communication or any other AAC strategy, would you think it important?
        Thanks Andy

        • jane

          Reply

          HI Andy, Thanks for commenting. What I would like to see, eventually, is research showing which forms of AAC lead to communication autonomy and communicative competence. I know that you know this - but communicative competence is one of the pillars of AAC intervention - and we are aiming for social competence, linguistic competence, operational competence and strategic competence. Communication autonomy is, of course, the ability to say what you want to say, wherever you want to say it, whenever you want to say it and to whoever you want to say it. It's what we're all participating in here - although within the boundaries of social competence, of course, as I'm sure there are many other things some of us would like to say that we don't!!

          I'd actually like you to consider something - I don't need a response because I'm sure we both understand that we are on very different pages in this discussion. Yesterday I put out a new blog post for book week. It has videos of shared reading. Shared reading is something that parents naturally do with their children almost as soon as they are born and that also happens at pre-school and school. Children participate in shared reading, initially very informally and then using their increasing language skills. I'd like you to consider how a student using AAC participate in that shared reading, responding to the various questions I ask people to pause and discuss. Does their AAC system support them with this participation? A system which doesn't support participation in those interactions is (in my opinion) a system which is unlikely to lead to communication autonomy. And if we don't even consider an individual's need to participate in those interactions then we are, in my opinion, indulging in a form of exceptionalism.

          • Andy Bondy

            First, of course encouraging parents to read to all kids is a good idea- or otherwise increase talking time with kids via any activity. But everyone knows that exposure alone is insufficient with regard to the substantial changes needed. But I would find it polite if you attempt to answer my question- do you know of any research along the parameters I noted?
            Thanks

          • jane

            HI Andy, thanks for coming back to clarify the communication breakdown! I have already responded to your question. I understand that you would prefer if I answered using one of the two options that you scripted for me, that is "yes" or "no". However, I chose to respond with my own words in what I believe is the most appropriate way. And I'm sorry it came across as impolite as that was not my intention. Autonomous communication can be challenging when you get responses you don't expect.

          • Andy Bondy

            Just so all readers are clear about important publications, there is a study that meets the criteria noted in my question. Schreibman & Stahmer (2014) published in JADD a randomized control trial comparing PECS to PRT (a speech-based intervention) involving 39 very young children identified ASD using 10 or fewer spoken words upon entry involving clinic and home based training by clinicians and parents. The 19 children in the PECS group used on average 89 spoken words after 6 months and 130 spoken words at 9 months. This is the largest impact of any AAC strategy with a large group of very young children – that’s why I use PECS- because the next kid I work with deserves the best.

        • Catharine

          Reply

          Hi Andy. Can I just doublecheck that I'm following up on the correct research that you mention please - I'm currently researching AAC (focusing on aided language stimulation and PECS). I've read Schreibman, L., Stahmer, A.C. A Randomized Trial Comparison of the Effects of Verbal and Pictorial Naturalistic Communication Strategies on Spoken Language for Young Children with Autism. J Autism Dev Disord 44, 1244–1251 (2014). https://doi.org/10.1007/s10803-013-1972-y https://rdcu.be/dlI0p

          The above article mentions " Children in both intervention groups demonstrated increases in spoken language skills, with no significant difference between the two conditions. Seventy-eight percent of all children exited the program with more than 10 functional words. Parents were very satisfied with both programs but indicated PECS was more difficult to implement."

          The quote above is from their abstract. The equivalent raw data scores for PRT were 83 spoken words after 6 months and 113 words after 9 months (versus 89, 130 for PECS).

          I know very little about PRT to be honest, but I assume AutismCRC would have classified it under behavioural approaches? Please can you confirm if this is the correct article that you are referencing in your comment. Thank you

  8. Jane M

    Reply

    Thanks Jane! I remember reading somewhere 'nothing for me, without me' and the research that you referenced regarding autistic individuals negative experiences with ABA definitely seems to reflect this. Ive just been reading references to research that suggest individuals without spoken language rarely have the opportunity to be immersed in a language system that is accessible to them, which is integral to expressive language development - this could be an opportunity barrier preventing greater efficacy in the AAC outcomes that you mentioned? And unless people have access to robust systems we will not be giving them the opportunity to tell us what they think and want (and why). We definitely have a long way to go! A great read - thank you!!

  9. Reply

    Thank you this wonderful post, Jane. I plan on printing copies of it to keep in my work bag to hand out to staff. I am constantly being questioned about why I am NOT using PECS and I'm still being told PECS needs to be trialed first. My heart breaks for the kids who have been unable to access robust communication systems sooner. Thanks for all you do.

    • jane

      Reply

      Oh Michelle - I'm glad it was helpful! And I will never understand why PECS is seen as a starter system. I don't think (although I could be wrong) that the PECS developers perceive it like that, and there's no evidence supporting that view point. And yet, it seems to be a common one. Very odd!

  10. Reply

    Thank you so much Jane for outlining your reasoning so clearly and with great evidence. I will be using it often when replying to the same question.

  11. Brenda

    Reply

    Thanks Jane. This is a great post. I agree with your reasoning and have travelled some of the same journey with use of PECS. I really appreciate how you have clearly pulled together all of the info and evidence. This will be so helpful when I am trying to convince others about the significant limitations of PECS. We need to provide our students with the tools to communicate ideas and thoughts beyond making a request. For me, I see the some very limited use of PECS for establishing the understanding of independently signalling a request to another person for a few of my more challenging students. But I would always also be modelling and introducing more robust communication supports at the same time. It is so exciting to see what students will communicate when you support their access to supports.that allow them to speak their mind!

    • jane

      Reply

      It is exciting when you see it, isn't it Brenda! And I particularly love it when they get to say what we don't want to hear!!

Leave a comment

Your email address will not be published. Required fields are marked *