Wednesday, October 22, 2014

One Size (with some tailoring) Fits Many: Key Considerations in AAC Selection

“There is no one-size-fits-all approach to AAC.”

In my part-time non-paying job (AAC advocating online) I’ve heard this a lot. I’ve said it a few times myself, but always halfheartedly, because I’ll tell you a secret: while one size doesn’t fit all, I believe that there are nearly universal principles that must be considered and incorporated when setting up an AAC system for any person.

Nearly universal. As in, applying to nearly all people. Not one-size-fits-all, I guess, but one-size-(appropriately tailored)-fits-many.  I’m not going to suggest a specific device/app/system, but I’m going to provide you with organizational/planning points that you may want to carefully consider as you purchase a system, or configure and personalize a system that you already own.

Please note that I am not suggesting you overhaul your system overnight. If you feel like your current AAC system is a great fit and working well, this post probably isn't for you.  I don’t suggest switching things up (unless in reading this, you realize that even their “proficiency” leaves them unable to communicate many things).

An AAC user deserves a system that will provide them with the maximum amount of language (referred to as SNUG---spontaneous novel utterance generation—the ability to put together an utterance that says exactly what they want it to say) with the minimum amount of time and effort. They need to be able to say exactly what they want to say, to whomever they want to talk to, whenever they want . . . and in a way that can happen as quickly and automatically as possible.

In order to help make this happen, here are the things that you need to think through:

1. Context-based communication pages are not the best idea. A context-based page groups together “all of the words” that you would need to communicate in a certain context. For example, an art page might contain these words (with icons):

I
paint
crayon
glitter
yellow
want
you
draw
marker
pencil
green
need
more
cut
brush
red
blue
like
all done
glue
scissors
orange
purple
bathroom

Why this is used: For someone new to AAC, this seems to make sense. Think of words that relate to art, put them all on a board, pull the board out during art. This board allows the user to request certain colors or items, to comment on the colors, to indicate what they like, to direct their communication partner “you draw,” and to also have a few commonly used words (more, all done, bathroom).

Limitations: Spontaneous novel utterance generation is low. Sure, the user can request red paint. But they can’t tell anyone that what they painted is a red school bus. They can’t ask to mix colors together, or request that the teacher draw a space shuttle so that they can paint stars next to it. They can’t point to their scribble and say “That’s my dad!” or tell their teacher “My pencil isn’t sharp enough” or “I don’t like the paint under my fingernails.” They can’t say “I’m thirsty!” They can’t comment on a friend’s painting or ask questions about what someone is drawing. They can’t say that they stopped painting because “my belly hurts” or “I feel tired” or “I don’t want to sit next to Sam because he keeps sticking his tongue out at me when you’re not looking.”

All of the words. When you wake up in the morning, you have the ability to use all of the words that you’ve ever heard. Your AAC user needs to be on a system that gives them access to all of the words, all of the time---not just food words at the dinner table and art words at the art table. 

2. Motor planning needs to be a part of the system. It just needs to be.  “Motor planning” is what happens when you need to make your body do something, like step up onto a ladder. You think about what you want to do (lift your foot to step up) and (very, very quickly) your nerves and muscles zap information around: contract this muscle, relax that muscle, shift your balance back a little, change the angle of your foot, lift, lean forward, move at hip and knee joint, lower foot down at the appropriate speed, contact!. The more that you execute a given motor plan, the better your body does it---it becomes automatic. A good example of this is typing the password on your computer. When you have a new password, you type it slowly for a few days, as your fingers learn the motor pattern of the new word. If you have an old password your fingers probably do it automatically, zipping through the letters without you even paying attention. That’s motor planning. (I wrote a whole post on it, with pictures and videos, here.)

Many individuals who are in need of AAC systems also have issues with muscle movements and motor planning, so the ability to organize a system in a way that allows for movements to access words to become automatic is essential. Imagine how much time and cognitive energy can be saved when an AAC  user’s hand just knows what little path to tap to say the word “go” instead of having to scan pages or search through folders because that word might be in a different spot.

AAC systems that honored and fostered motor planning principles used to be hard to find (limited to one company, really), but more and more companies are realizing how essential it is and attempting to allow their products to be configured in a way that will allow for the automaticity that following a motor plan to find a word can bring.

Here are some examples of when motor planning is not incorporated:


Motor Planning Foil #1: Movement of words with increasing vocabulary size.

GRID 1:

red


orange

yellow

green


blue

purple

black


brown

grey

GRID 2:

red


orange

yellow

green

blue


purple

black

brown

grey


white

paint

brush


pencil


glue

scissors

glitter

In Grid 1, each word has a spot. The user learns the motor plan to say these words. Then, the vocabulary increases (Grid 2) and the user is left having to scan and search for words (or images). And every time the vocabulary increases, the words move. Not good.

Ideally, a grid should have as many buttons as possible (a lot---even for people with fine motor challenges)  and the ones that aren’t currently in use could be blacked out (which will make the motor learning and target hitting more possible for people with motor challenges as well), reserved to hold more words later.

As a simple example, this would be a revision of Grid 1, which holds all of the same words but can grow to become Grid 2 without changing the motor plan for the original words:


red


orange

yellow

green

blue


purple

black

brown

grey


          
















-Motor Planning Foil #2: Predictive Text.  Predictive text walks a user through a sentence by “predicting” the type of things that a user might want to say. One issue with apps that use predictive text is that the app is not fully able to predict the types of spontaneous, novel things a person might want to say.


                                                                                         to
                                                                       want          (people)
                                                          I   à      like   à     (toys)
                                                                       feel            (food)


For example, if you type the choices like, want, and feel might appear. Tapping like might lead you to categories toys, food, people, and probably the word to, which could connect to other verbs (to eat, to drink, etc). But what if the user wanted to say “I like jumping on my bed” or “I like the smell of farts” (I mean, you never know).  That’s one problem with predictive text.

Another problem with predicative text is that, because the screen reloads after each selection, everything is always moving. If the word moves around (in the upper left corner if you get to it this way, but down at the bottom if you get to it a different way), then the user can never internalize the motor plan to get to a word, and they have to spend a lot of time scanning and searching. It’s a waste of energy.  The last issue with predictive text is that it's sometimes difficult to find a word if you’re not walked to it (if you wanted to start a sentence with the word caterpillar, it might be dicey). There are typically “category” folders in these systems that you can also access, but they may have many layers (which means many button pushes to find a word) or require scrolling (which wastes time and can present a challenge to those with fine motor impairment).

-Motor Planning Foil #3: Putting one word on a bunch of screens. When you say a word with your mouth (like “cut”) your muscles always move in exactly the same way to produce that word, no matter what the context is. You can “cut an apple” or “cut down a tree” or “cut that deck of cards”, but cut = cut = cut. If your system is made of a lot of screens that could function as context boards, then you may have one word in a lot of places (“want” on the eat screen, the drink screen, the art screen, the toy screen, etc). This is a lot of clutter and potential confusion (especially if the screens don’t have duplicate words in the same exact place, which should be the minimum requirement for duplicating words within a system).


3. Minimal work for maximum SNUG An AAC user shouldn’t have to navigate through multiple layers of folders to find a word. One of my favorite sentences that Maya ever said on her device was “Lightning---scary!” as the thunder cracked outside of our window. I loved it because it was exactly what she wanted to say, and she was able to quickly get to those two words, lightning and scary, which could easily have been buried in layers in another system. In one system the default to find the word lightning would require this navigation:

(more) --- (school) --- (weather) --- (more) --- lightning

No user should have to think their way through multiple categories to find a word. By the time Maya reached “lightning” in that example (if she managed to get there without giving up) the moment would have been long gone. And that’s only a two word sentence! It would take a very long time to say something more substantial. Systems need to be organized in a way that will minimize the amount of navigation that it will take to get to each word. Each additional step of navigation takes extra time and cognitive energy. 

A word of warning, though: notice that I said the goal is maximum spontaneous novel utterance generation----not maximum words. Programming each button to say several words (“I want” “I like” “I see” “I like that” “Pick up” “put down” etc) is not actually increasing language. It’s just relabeling buttons. It makes novel utterances more tricky (what about “You like” or “He likes” or “Mom likes”---do they all get buttons too?).  It might make it faster and easier to say simple things, but this approach will clutter the field (or fill more folders to tap through) and make it more difficult to combine individual words and build new sentences. This multi-word-button approach will also make it difficult to approach instruction about grammar, verb tenses, syntax, etc in a simple, direct way. Another consideration is that for children with some diagnoses, speech segmentation (the spaces where one word stops and a new word begins) is a challenge. This is part of why some children are prone to scripting---using entire sentences “Go to the store now?”  as one “word”, building sentences like “go to the store now yesterday.” Using a single-word vocabulary allows the users to see a word as the smallest unit of language and learn how to manipulate those pieces.

4. There is no “starter” AAC.   Do not waste your time (or money) on a “starter” AAC system, one that you plan to use for a little while until they are ready for a robust system. This is your child’s (or client’s) ability to communicate---it is not fair to waste months of their life teaching them a system that you fully intend on abandoning when/if they become “good enough” at using it to prove that they’re ready for something else. Imagine the uprising that would happen in an office setting if months were spending training on and incorporating a major software system, and just when the employees had all mastered it the boss called them in to congratulate them and introduce them to the “real system that we’re going to use, now that you’ve all proved your competency.” Unfair.

You need to find a system that can grow with the user---and you need to evaluate how the growth will happen in terms of  considerations mentioned above.

  • Will they have access to all of the words easily as it grows? 
  • Will the motor planning remain the same, or very very close to the same? 
  • Will words become buried in layers of folders as the vocabulary increases? 
  • Will predictive text make it difficult to say exactly what they want to say?
  • Can the system hold a massive amount of words, and does it have a solid keyboard to support emergent and conventional spellers who may want to start typing their words?

In my opinion, these are core points to consider as you plan and implement any AAC system. I believe that a system that is organized in a way that takes the above into consideration will offer maximum success and maximum SNUG. If you’ve landed on this page because you’re in the process of selecting an AAC system for your child, I hope that you’ll think through each of these points as you evaluate the options----each one of these things became a bump in the road for us at some point, and I wish that we could get back the months that we wasted figuring all of this out. I wish that we could have heard Maya’s thoughts when she was younger, rather than fumbling through several options that let her effortlessly request milk, or a certain color crayon, and not say much else.




PS: And now, for my preemptive defenses, because I anticipate push back on this:
1. My kid's system doesn't do any of that, and it's perfect for him! That's great! I'm so glad your kid has something that works well for them. This post isn't for you, then, because I would never suggest that you switch a system that is working well for you. However, for most new users, I believe the points above are essential considerations that will maximize success and SNUG.

2. What about kids with vision issues who would struggle with small buttons? Good question, and one that I don't know the answer to. I do believe that the motor planning is especially important for those with vision issues, who should be forced to use their vision to scan for words even less than a user with normal sight. I've seen great use of high contrast symbols and heard about success with visual/tactile markers in the field. When Maya was at the eye doctor she had her eyes dilated, and she was using her talker after the dilation. The doctor said to me: "You know that she can't see those pictures anymore, right? They are just bright blurry squares." I had no idea, but it made sense that the motor plans she had learned for each word, combined with the vague coloring of the icons she knew, was enough to keep her going. (And I'm certainly not saying that my kid's post-dilation blurry vision  is the same as someone with a serious visual impairment . . . it was just interesting for me to see her motor planning in action when her vision was reduced.) I don't know how to create systems for all users, but I believe that the above features can be considered when planning any system.

3. What about kids who struggle to hit small targets? They might need a few words on one page. Again, good question, and one that I don't know the answer to. Access is tricky. One thing that I know is that many kids who seem unable to hit a small target (or even isolate a finger) can often develop that ability with a keyguard, or fingerless glove, combined with creative positioning of the device and some stabilizing hand-under-hand (or wrist) support. I've seen an AAC user who uses his toe for direct access. If I were the AAC user and I had the following two choices: say exactly what I want to say, but it will take longer and more effort or say a few things (eat, drink, more, stop) really fast . . . well, I would choose the former. I don't know how to create systems for all users (and I'm not even going to wander into the waters of non-direct access here), but I believe that the above features can be considered when planning any system.

4. Who are you, anyway? You aren't even qualified to make a list of what needs to be considered.  If a professional's response to this list is to defensively point out that I am not a professional, rather than to consider these points with an open mind (and then provide well-argued critiques, if necessary), well . . . I would have reservations about working with that professional. Feel free to come back with a point-counterpoint rebuttal. I welcome debate and feel like we can all learn and grow from it. (If you ended up here following links and are sincerely wondering who I am, I'm an AAC parent who has spent a few years doing a lot of reading, research, immersion, and training about all things AAC. And I'm in school now to become an SLP.)




4 comments:

Michael said...

Awesome writeup! I've seen most if not all of the considerations elsewhere as I believe you have. Thank you.

Dana said...

I'm not fully sure what you mean, but I definitely haven't invented anything here! Motor planning is, well, life. Everything is motor planning.

Combining new words to form sentences is the basic generative aspect of language, which is kind of intrinsically known to anyone who has spoken, but is also in every basic SLP class. "SNUG" is a common AAC term that I think I heard first at the ISAAC conference in 2012, but it's one that definitely stuck, because it makes good sense (and it's kind of cute). Here's an ASHA "information for the public" page that delves more deeply into SNUG:
http://www.asha.org/public/speech/disorders/CommunicationDecisions/

My purpose here was to assist people who are looking for systems to consider features that someone unfamiliar with AAC might not have thought through. We wasted time by not knowing some of this stuff going in, and I want to save other the time. I didn't even think to include a disclaimer like "I didn't invent motor planning or the acronym 'SNUG'" because, well, I thought that would be obvious. (The ASHA article that I just linked also doesn't cite the originator of the SNUG acronym---I only know it as part of the general AAC lexicon, along with "access method" and "dynamic display.")

Anyway, thanks!

Michael said...

Sorry, I wasn't implying anything negative. I was applauding you taking the time to collect all of this into one place. I too feel these are some of the most important things to consider. Thanks for putting this information all in one place.

I agree motor planning is uber important. I also share when explaining to others that it is so important that we should try our best to have motor planning remain the same for *life*. I don't expect my typical son to learn the spoken motor plan for the word "go" more than once. Once he learns that oral motor plan he doesn't have to learn it again for the rest of his life. We should strive for the same for our children when using an AAC device, particularly so when they have motor plan challenges.

Antti said...

It's possible to practice the small targets with apps like Scene Speak, Autismate or Scene & Heard. Make boards (for example with Photoshop Touch) with a single colour background. On that backround place an icon with a size and content you see fit, in different places on each board. Make the icon a hotspot with a reward (we use video clips) and a link to the next board using one of the apps mentioned. Now you have a fun game with improving pointing as a side effect. As the child gets better, you make a new game with smaller icons.The mitten doesn't work with our son and he likes to stim with the keyguard (or tears it off), but this works. (I may have mentioned this before, I'm not sure. Apologies, if I repeat myself!)