Long-term outcomes of Hanen’s Target Word program

How much of a long-term impact does parent-implemented speech–language therapy have on late talkers? That’s an important question. Most studies have focused on short- or medium-term outcomes and have not looked into long-term follow-up results.


These researchers wanted to know the medium- and long-term effects of a low-dosage parent program for late talkers. To do this, they provided the Hanen’s Target Word program to 30 parents of 24-month-old late talkers, and followed a different set of 30 parents–child pairs that did not receive the program. The program included five group sessions and two individual video-feedback sessions over the span of three months. They then tested parents’ use of the strategies at 36 months, and language skills at 36 and 48 months.

They found that the program accelerated the vocabulary growth for the late talkers, but by age 4, there were no longer any differences between the two groups’ expressive vocabularies and both had reached age-appropriate scores. Both groups, however, continued to fall behind their peers in syntax and grammar.

The parents who participated in the study followed their child’s lead more and put less communicative pressure on their child; however, the parents did not significantly improve in how often they responded to their child’s communication or in how much linguistic stimulation they provided their child. Through analysis, the researchers found the strategy of reducing communicative pressure to be specifically associated with children’s language growth.

Here are some important considerations and takeaways from the study:

  • Null long-term results don’t mean that the intervention wasn’t worthwhile. As we know, increasing two year olds’ vocabularies can help ease the frustration of not being able to communicate which is an important goal of early intervention.

  • Dosage and coaching style play an important role in treatment outcomes. The authors discuss how the parent results of this study differed sharply from the Roberts and Kaiser (2015) study in which the researchers taught parents Enhanced Milieu Teaching (EMT) strategies, and parents were able to learn all of them. One of the reasons why is that the intervention dosage was much higher for the EMT study (four workshops and 24 hour-long sessions, compared to two), giving parents more opportunity for practice and feedback. Adjusting the dosage or coaching style of the Target Word program may be one option for increasing its long-term effectiveness.

  • Reducing communicative pressure may be an especially important strategy for parents. This is a simple strategy to teach parents, and these results demonstrate how powerful it can be for supporting language growth in toddlers.

  • Ongoing monitoring is important for late-talkers. Nearly one third of the children who participated still had language scores below the normal range at four years of age, but fewer than half of those children received speech–language therapy after the study. Monitoring is important to ensure that these children don’t fall through the cracks when language demands as they grow older.

One final note: this isn’t a replication study of Hanen’s Target Word program, but rather it’s a study that tested its effectiveness in real-world contexts. That’s awesome for two reasons: (1) authors aren’t tied to Hanen, which helps eliminate bias, and (2) real-world contexts = more like what we SLPs face = more clinically applicable! 

Want more? These authors also published a study of Hanen’s Target Word program (so popular!) This retrospective study (looking back at client charts) similarly showed gains in communicative participation and vocabulary for children who participated in the program.


Kruythoff-Broekman, A., Wiefferink, C., Rieffe, C., Uilenburg, N. (2019). Parent-implemented early language intervention programme for late talkers: parental communicative behaviour change and child language outcomes at 3 and 4 years of age. International Journal of Language & Communication Disorders. doi: 10.1111/1460-6984.12451

Social disadvantage and language development—what matters, and how much?


This is one of the largest studies we’ve seen in a while on factors than impact infant and toddler language—over 1000 babies were followed to see what matters most for language development in the first two years of life.

First, one of the really beautiful things about this study is that they measure social disadvantage, but actually differentiate things parents can control (that is, what they do with their child) from things they can’t (e.g. family income, maternal education). So bravo for that! Then, note that this study is dense. We’re only skimming the surface, here, with the biggest takeaways for SLPs, which are:

 What parents do with their child matters:

  • e.g. “…reading to their child, telling stories, singing and taking the child on errands to public places… having toys… books available in the home… having a safe and supportive home environment.”

  • (In fact, here’s another recent study showing that responsive verbal behaviors in play interactions with a parent at 12 months are predictive of language outcomes at 36 months in a cohort of families experiencing adversity.)

Language at 15 months is highly predictive of language at 2 years.

  • Implication: Are we catching kids early enough?

Now, I think most SLPs pretty much already knew that. But what you may not have a good feel for is how much these things matter. To put it in perspective—the single strongest measured predictor of language was biological sex. Also, most of what predicts language development was not something the researchers were able to measure. Instead, it’s some other thing(s), likely, “… other developmental or genetic mechanisms.” So, basically, what parents do with their children is important not because it matters the most, but because we have control over it (well… we can try to).


Law, J., Clegg, J., Rush, R., Roulstone, S., Peters, T.J. (2018) Association of proximal elements of social disadvantage with children's language development at 2 years: an analysis of data from the Children in Focus (CiF) sample from the ALSPAC birth cohort. International Journal of Language and Communication Disorders. Advance online publication. doi: 10.1111/1460-6984.12442.

A (free) decontextualized vocabulary test for toddlers

If only we had a crystal ball to predict late talkers’ later language skills (ok, those would be nice in so many areas of our field!). As SLPs we are always trying to get an accurate picture of how many words a child uses and understands. But because we don’t have a dedicated assessment of toddlers’ early vocabulary, we often do this through parent report, either formally (like an MCDI) or informally. However, researchers have found that parents' report of their child's vocabulary doesn't do as good of a job when we try to use it to predict later language. It seems that parent report may not be giving us all of the information that we need about a child’s vocabulary in order to predict their later language abilities.

We know that there’s a continuum of what “knowing a word” entails for a toddler. When they first learn to say the word “milk,” they may only know it within the context of breakfast. The word “dog” however, they may use all day long whenever they see a dog or a picture of one. A parent would rightly conclude that a child knows both “milk” and “dog,” even though the child’s “level of knowing” varies between the two words. Contrast this with a picture ID vocabulary assessment, in which all of the words are presented out of context. For a toddler to correctly identify a word during the assessment, he would have to have a strong understanding of what that word really means without all of the support that context provides.


Because of this, Friend et al. (2018) set out to create a picture ID test for young children and test if it could better predict preschool language abilities than parent report. They developed the Computerized Comprehension Test (CCT) in which children are asked to identify decontextualized pictures from a field of two. They then tested it with 16-, 23-, and 36-month-old children. At 16 months, parent report continued to best predict preschool language abilities. However once the child reached 23 months, the CCT was a stronger predictor of later language abilities with great psychometric properties.

While the CCT is still in its early phases, it shows promise of becoming a useful instrument for EI SLPs to get a more accurate picture of toddlers’ vocabulary, especially after age two. In its current state, it could serve to complement other assessments already in your toolkit, such as an MCDI. And while we’ll never have a crystal ball to tell us what a child’s language will be like years down the road, assessments such as this may give clues to help us make more informed decisions about assessment and treatment.  

Note: The authors have published all of their CCT materials online, including the computerized assessment, training videos/instructions, and data sheets. I tested out the program myself, and while there was a learning curve, it didn’t prove to be too challenging! Let’s all take a moment and cheer for these scientists giving us access to the materials we need!


Friend, M., Smolak, E., Patrucco-Nanchen, T., Poulin-Dubois, D., & Zesiger, P. (2018). Language Status at Age 3: Group and Individual Prediction From Vocabulary Comprehension in theSecond Year. Developmental Psychology. Advance online publication. doi: 10.1037/dev0000617

Joint engagement predicts language scores, but which kind of joint engagement?

Joint engagement predicts language scores. Got it. But which kind? You lost me. There are different kinds? As it turns out, yes, and this paper offered some pretty good information on them. You have your good old, traditional joint engagement, which is when at least two people actively attend to the same thing at the same time, and each person knows they are attending to the same thing at the same time. But you can break that down even further, into supported and coordinated joint engagement. Supported joint engagement (SJE) is when the child focuses on a shared item, but the parent carries the communicative load, so to speak. The parent supports the rest of the interaction while the child only has to spend cognitive energy on that item. Coordinated joint engagement (CJE), on the other hand, is when the child splits focus between the object at hand and a communication partner. For example, consider a Jack-in-the-Box toy. In SJE, the parent might turn the handle, narrate what is happening, and react to the doll popping out of the box while the child watches, listens, and learns about the weird musical box. In CJE, the child might turn the handle as she otherwise divides her attention between the toy and her parent, commenting and sharing in the suspense and surprise. 

Conway, et al. recorded mothers playing with their two-year olds and administered language assessments at 2, 3, (PLS-4) and 4 years of age (CELF-P2). Mean language scores for the group fell within low average range. The idea was to rate the videos and compare time spent in supported and coordinated joint engagement with maternal responses and language scores. Check out Table 2 in the study for details on the coding and descriptions of joint engagement, and Table 3 for descriptions of maternal responses measured.


It turned out that the more time a mother and child spent in supported joint engagement, the higher the receptive and expressive language scores at 2- and 3-years, but the association was not significant at the 4-year testing. Coordinated joint attention was not associated with language scores, and the authors suggest that coordinated joint attention could be associated with other language skills, like pragmatics. Looking at maternal responses, the authors found that expansion and imitation were associated with language scores when the children spent less time in supported joint attention. The authors suggested that “where SJE is less frequent or of lower quality, expanding or imitating a child’s utterance may be especially important.”

While the authors caution that more studies are needed before generalizing these results, encouraging supported joint engagement and using imitation and expansion in interactions may be a good choice for supporting language development in both typical and late-talking 2-year-olds.


Conway, L. J., Levickis, P. A., Mensah, F., Smith, J. A., Wake, M., & Reilly, S. (2018). The role of joint engagement in the development of language in a community-derived sample of slow-to-talk children. Journal of Child Language. Advance online publication. doi: 10.1017/S030500091800017X

Throwback Pub (2012): Training parents to support children with developmental language disorder

As SLPs working in EI, we know that developmental language disorder presents risk for later academic skills. We also know that parent-implemented language interventions can be effective (see Roberts & Kaiser, 2011, for a meta-analysis of parent-implemented language interventions). This study examined whether parent-implemented Enhanced Milieu Teaching (EMT) would impact receptive and expressive language growth in children with language disorder (with cognitive standard score > 80 and no other primary diagnoses; see article for full inclusion/exclusion criteria).

The researchers examined three groups of 24–42-month old children:

  • Typical language

  • Language disorder w/ no treatment, or a “wait-and-see” approach

  • Language disorder w/ parent-implemented Enhanced Milieu Teaching


Families treatment group participated in parent training in 28 sessions (workshops, clinic, and home visits) over a three-month period. EMT strategies were trained in four phases: 1) setting a communicative context, 2) modeling and expanding communication, 3) time delay strategies, and 4) prompting strategies (see Table 5 for examples and description). During training at the clinic, therapists and parents practiced strategies using a specific set of toys. Home visits also included integrating strategies into functional family routines such as snack. All children were assessed with monthly language samples (MLU, total words, different words) plus the Preschool Language Scale, Fourth Edition (PLS-4), at the beginning and end of the study.

So what happened? Parent-implemented EMT was effective for improving language outcomes for children with LI, and parents in the treatment group used significantly more strategies than either other group! Parent use of strategies for typically-developing children and those with language disorder undergoing “wait and see” were about the same.

Children whose parents used EMT

  • significantly improved their PLS-4 Total and Expressive Communication scores and gained an average of 50 more words than untreated children with language disorder

  • gained an average of 15 more words per month compared to untreated children with language disorder

  • grew at about the same rate as children with typical language

Fantastic, right? But what about the kids with language disorder undergoing a “wait and see” approach? They showed significantly slower growth. Not only did those “wait and see” kids start out with lower language abilities, it looks like “waiting and seeing” might just hold them at a disadvantage when compared to those who have access to intervention. The authors stated, “Children in the LI-control group did not catch up but fell farther behind their peers with TL.”

How about a little extra info? The authors also looked at which child characteristics predicted language growth and outcomes. They examined risk at birth (e.g., a NICU stay), cognitive skills, and receptive skills (based on the Bayley-III scores). Risk at birth and cognitive skills were not predictive of expressive language analyses performed. However, “Receptive language at the start of the study predicted growth in language for all three groups of children after controlling for differences in IQ.”

The authors note that more research is needed to determine long-term outcomes. For example, how would these children progress over a period of 12 months? What kind of gains would come from 28 hours of traditional therapy in 3 months? Overall, parent-implemented EMT looks to be a promising model to add to our intervention options for young children with developmental language disorder.

Note: For our bilingual SLPs, see Peredo, Zelaya, & Kaiser, 2017 for a study on adapting parent-implemented EMT for Spanish-Speaking families.


Roberts, M. Y., & Kaiser, A. P. (2012). Assessing the effects of a parent-implemented language intervention for children with language impairments using empirical benchmarks: A pilot study. Journal of Speech, Language, and Hearing Research55(6), 1655-1670.

Throwback Pub (2014): Telegraphic or grammatically complete prompts—which is best?

“Dog sitting.”  or “The dog is sitting.”

Which would you choose to use as a model in therapy for a language delayed toddler?

If you said the first sentence, you’re not alone. The common assumption among many pediatric SLPs and parents of young children is that short phrases with the grammar removed—aka: “telegraphic utterances”—are a better choice for young kids because they make it easier for them to understand and imitate. And, popular, research-based treatment programs like Enhanced Milieu Teaching (Hancock & Kaiser, 2006) include telegraphic prompts, so they have to be good, right?

But, here’s the problem: previous research has actually shown just the opposite (e.g., van Kleeck et al., 2010). And, in fact, some studies have shown that when young kids don’t hear grammatically complete models, they begin to assume that those telegraphic utterances are the rule (e.g., Leonard & Deevy, 2011), and then have trouble using them in spontaneous productions (e.g., Theakston, Lieven, & Tomasello, 2003).

So, what’s an SLP to do?

That’s where this study comes in. Because we know that hearing correct syntax and morphology is important, and particularly for young kids with language impairments, Bredin-Oja and Fey wanted to find out what happens when models for imitation are grammatically complete. Can young kids with expressive delays still imitate them? And, how does that compare to their ability to imitate telegraphic models?


Five 2-, 3-, and 4-year-olds with expressive language delays participated in 14 play-based therapy sessions. Seven of the sessions involved grammatically complete prompts (“The boy is jumping”), and seven included telegraphic models (“Boy jumping”). Results show that all five of the kids responded just as reliably to grammatically complete prompts to imitate as they did to telegraphic. And, three* of the five kids included morphemes in their imitated utterances only following a grammatically correct model.

This small study has big implications when it comes to the models that we provide in therapy, and also how we teach parents to talk to their young kids. Put simply, the message has to be simplified, but not at the expense of accuracy when it comes to grammar. The authors provide some helpful suggestions for how clinicians and parents can achieve this at the end of this article.

*Two of the kids didn’t produce the morphological markers at all, regardless of whether they were presented in a telegraphic or a grammatically complete utterance. The authors hypothesized that they were probably just not developmentally ready to produce those language forms, and that makes a lot of sense and aligns with previous research (e.g., Fey & Loeb, 2002). 


Bredin-Oja, S. L. & Fey, M. (2014). Children’s responses to telegraphic and grammatically complete prompts to imitate. American Journal of Speech-Language Pathology, 23, 15 – 26.