Functional speech assessment for children with CP


Although there’s a lot of information out there about children with cerebral palsy (CP) who use AAC, what about those who are verbal? The speech of children with CP presents uniquely, with at least half having dysarthria. Because of the myriad presentations of dysarthria (flashback to motor speech disorders in grad school!) it can be difficult to differentiate between dysarthria and other speech/sound disorders. Detecting motor speech disorders at the youngest age possible is vital to ensuring that we are using the most appropriate, evidence-based treatment.

Hustad et al. used measures of functional speech in an attempt to differentiate five-year-old children with CP who have motor speech involvement (i.e. dysarthria) and those who do not. Those functional measures of speech included intelligibility, speech rate, and intelligible words per minute (a measure of speech efficiency). Children’s speech was measured using delayed imitation, so that evaluators knew the target words. However, these measures could be used with just about any speech sample! Below is a little review for how to calculate these handy measurements:

Screen Shot 2019-08-13 at 9.00.00 PM.png

All three measures readily differentiated children with dysarthria from children without dysarthria (with both typical development and CP). Furthermore, they even differentiated children with CP but without dysarthria from typically developing children, showing that even kids with CP who appear to have typical speech may have borderline to mild speech difficulties. Intelligibility was the strongest differentiator, with 90% of typically developing five-year-olds falling at 87% intelligibility or greater. See Figure 1 in the article for the hard data, including cutoff scores for differential diagnosis of dysarthria in kids with CP.

Note: Although this study focused on children with CP, functional measures of speech can be useful for any speech evaluation. These measurements, along with other assessment tools, can help us both to identify speech disorders at the earliest possible age and to make decisions regarding intervention and the use of AAC.


Hustad, K.C., Sakash, A., Broman, A.T., & Rathouz, P.J. (2019). Differentiating typical from atypical speech production in 5-Year-Old children with cerebral palsy: A comparative analysis. American Journal of Speech–Language Pathology. doi: 10.1044/2018_AJSLP-MSC18-18-0108.

Diagnosing DLD when you don’t speak a child’s first language


We know that it’s best to assess children in their first languages. But, we simply don’t have access to measures or interpreters for all of the world’s languages. What’s a monolingual SLP to do?

New research supports what we’ve discussed previously: that by using parent questionnaires and measures of language processing, we can accurately diagnose language disorders in English language learners using only English measures. Li’el et al. recruited a sample of bilingual and monolingual Australian English-speaking 5- to 6-year-old children with and without developmental language disorder (DLD). “Bilingual” was defined as hearing English less than half the time at home. Parents completed a questionnaire and children completed the CTOPP nonword repetition and CELF-P2 recalling sentences subtests.

The researchers found that the parent questionnaire alone had the highest sensitivity and specificity (accuracy at ruling in and ruling out DLD). However, all of the assessments in combination still had good diagnostic accuracy, and it’s not a good idea to diagnose a child with only one test, so the authors recommend using more than one measure.

Overall, this study adds to evidence that by interviewing parents and using language processing tasks, we can do a pretty good job teasing apart a lack of English exposure from an underlying language disorder even if we can’t assess in a child’s first language.


Li’el, N., Williams, C. & Kane, R. (2018). Identifying developmental language disorder in bilingual children from diverse linguistic backgrounds. International Journal of Speech-Language Pathology. Advance online publication. doi: 10.1080/17549507.2018.1513073

Narrative skills of children who use AAC


Telling stories is an important social skill, and one that may be challenging for children who use AAC. This study looked at stories told by 8- to 15-year-old children who use AAC with the support of a familiar communication partner. Children watched short, wordless videos that featured some sort of problem (like a person slipping on a banana peel) and then explained the story to someone who hadn’t seen it. Overall, children who used AAC produced narratives that were shorter and contained fewer important elements than those of same-age speaking peers, although “[t]opic maintenance and setting or character descriptions” were relative areas of strength. Communication partners typically elaborated on what children were saying without taking over the interaction, and there were no significant differences in story quality between types of communication partners (peers, parents, and professionals).

Some implications of the study:

  • Think about using narrative tasks to assess the ease and complexity of communication for children who use AAC—and remember to teach narrative structure, too!
  • Observe interactions with peers, parents, teachers, or other familiar communication partners to see how their assistance affects the child’s communication. You could also see what functions the partner is taking on—filling in details, compensating for lack of available vocabulary, etc.—and target those areas to increase independence.

Note that all children in this study had cerebral palsy and good receptive language and cognitive abilities per teacher report; it’s not clear how these findings would apply to children with different language and cognitive profiles.

Smith, M. M., Batorowicz, B., Dahlgren Sandberg, A., Murray, J., Stadskleiv, K., van Balkom, H., Neuvonen, K., & von Tetzchner, S. (2018). Constructing narratives to describe video events using aided communication. Augmentative and Alternative Communication, 34(1), 40–53. doi: 10.1080/07434618.2017.1422018.

Modeling AAC is an evidence-based practice (officially)

Before you say anything — we know. This is an obvious fact to anyone who’s been paying attention in the AAC world for quite some time. However, there are still times when it’s helpful to have clear, scientific evidence to share with those who aren’t, shall we say, on board. Evidence like a recent systematic review, for instance? Voilà!

This new study synthesized the available research on direct teaching strategies** that support symbol learning and aided language expression for AAC users, and found four strategies they identify as “potentially effective.” They looked at 15 studies that met their criteria for participants, intervention type, outcome measures, and study quality.

Important note: They specifically did not look at AAC users with autism.  The studies in the review included children up to age 18, with better receptive than expressive language, and “no more than a moderate intellectual disability.” The most frequent diagnoses represented were cerebral palsy and Down Syndrome.

These four strategies were supported by the evidence:

  1. Aided AAC modeling (8 studies)
  2. Narrative-based interventions (4 studies, limited by small number of participants and variation among studies)
  3. An “eclectic” approach, including increased communication opportunities, modeling, and least-to-most prompting (1 study)
  4. Mand-Model (2 studies) E.g., Child is interested in a ball. Clinician models, “That’s a BALL.” Prompts child to say “Ball,” possibly using a cueing hierarchy. Child gets the ball.

A few things to keep in mind: First, these strategies are not mutually exclusive, and overlap to some degree; they all include some aspect of modeling. Second, we can’t say anything at this point about whether one of these methods is more effective than another, and it could be that certain strategies might work better at specific stages (when a device is first introduced, for example, to teach a specific language skill, or with a very young child). Finally, it’s pretty tricky to make comparisons between studies, since so many different terms can be used for the same, or similar, procedures (modeling, aided language input, aided language stimulation, etc.). Different people define these procedures in different ways—a fixed ratio of statements to questions, for example, or a minimum expectation of how many utterances are also modeled on the device.

Fun Fact! Did you know that there’s a rule of thumb to determine if an intervention can be *officially* considered evidence-based, on the basis of single-subject or small studies? Horner et al. (2005) say you want to find:

  • At least five peer-reviewed studies (well-designed ones!)
  • By at least three different people
  • In at least three geographic locations
  • With at least 20 subjects total

Based on these criteria, and the findings of the review, aided AAC modeling is an evidence-based practice. There wasn’t enough evidence available for the narrative-based, eclectic, or mand-model strategies to meet this threshold.

For systematic review articles like this one, the gold for clinicians is often the big ol’ summary table (Table 1). It’s a quick way to see which of the included studies are directly relevant to YOU, by breaking out details like who the participants were, the AAC systems they used, the type and dosage of the intervention, and the outcome measures. Plus, the authors have already screened these studies for quality, which can be reassuring if you aren’t super confident about assessing this yourself. If you want to learn more about specific intervention procedures for specific clients, this paper has done your searching for you.

**So—not things like communication partner training, which is also good.

Lynch, Y., McCleary, M., & Smith, M. (2018). Instructional strategies used in direct AAC interventions with children to support graphic symbol learning: A systematic review. Child Language Teaching and Therapy, 34(1), 23–36. doi: 10.1177/0265659018755524.