Making Our Interventions Count or What’s Research Got To Do With It?

Two years ago I wrote a blog post entitled: “What’s Memes Got To Do With It?” which summarized key points of Dr. Alan G. Kamhi’s 2004 article: “A Meme’s Eye View of Speech-Language Pathology“. It delved into answering the following question: “Why do some terms, labels, ideas, and constructs [in our field] prevail whereas others fail to gain acceptance?”.

Today I would like to reference another article by Dr. Kamhi written in 2014, entitled “Improving Clinical Practices for Children With Language and Learning Disorders“.

This article was written to address the gaps between research and clinical practice with respect to the implementation of EBP for intervention purposes.

Dr. Kamhi begins the article by posing 10 True or False questions for his readers:

1. Learning is easier than generalization.
2. Instruction that is constant and predictable is more effective than instruction that varies the conditions of learning and practice.
3. Focused stimulation (massed practice) is a more effective teaching strategy than varied stimulation (distributed practice).
4. The more feedback, the better.
5. Repeated reading of passages is the best way to learn text information.
6. More therapy is always better.
7. The most effective language and literacy interventions target processing limitations rather than knowledge deficits.
8. Telegraphic utterances (e.g., push ball, mommy sock) should not be provided as input for children with limited language.
9. Appropriate language goals include increasing levels of mean length of utterance (MLU) and targeting Brown’s (1973) 14 grammatical morphemes.
10. Sequencing is an important skill for narrative competence.

Guess what? Only statement 8 of the above quiz is True! Every other statement from the above is FALSE!

Now, let’s talk about why that is!

First up is the concept of learning vs. generalization. Here Dr. Kamhi discusses that some clinicians still possess an “outdated behavioral view of learning” in our field, which is not theoretically and clinically useful. He explains that when we are talking about generalization – what children truly have a difficulty with is “transferring narrow limited rules to new situations“. “Children with language and learning problems will have difficulty acquiring broad-based rules and modifying these rules once acquired, and they also will be more vulnerable to performance demands on speech production and comprehension (Kamhi, 1988)” (93). After all, it is not “reasonable to expect children to use language targets consistently after a brief period of intervention” and while we hope that “language intervention [is] designed to lead children with language disorders to acquire broad-based language rules” it is a hugely difficult task to undertake and execute.

Next, Dr. Kamhi addresses the issue of instructional factors, specifically the importance of “varying conditions of instruction and practice“.  Here, he addresses the fact that while contextualized instruction is highly beneficial to learners unless we inject variability and modify various aspects of instruction including context, composition, duration, etc., we ran the risk of limiting our students’ long-term outcomes.

After that, Dr. Kamhi addresses the concept of distributed practice (spacing of intervention) and how important it is for teaching children with language disorders. He points out that a number of recent studies have found that “spacing and distribution of teaching episodes have more of an impact on treatment outcomes than treatment intensity” (94).

He also advocates reducing evaluative feedback to learners to “enhance long-term retention and generalization of motor skills“. While he cites research from studies pertaining to speech production, he adds that language learning could also benefit from this practice as it would reduce conversational disruptions and tunning out on the part of the student.

From there he addresses the limitations of repetition for specific tasks (e.g., text rereading). He emphasizes how important it is for students to recall and retrieve text rather than repeatedly reread it (even without correction), as the latter results in a lack of comprehension/retention of read information.

After that, he discusses treatment intensity. Here he emphasizes the fact that higher dose of instruction will not necessarily result in better therapy outcomes due to the research on the effects of “learning plateaus and threshold effects in language and literacy” (95). We have seen research on this with respect to joint book reading, vocabulary words exposure, etc. As such, at a certain point in time increased intensity may actually result in decreased treatment benefits.

His next point against processing interventions is very near and dear to my heart. Those of you familiar with my blog know that I have devoted a substantial number of posts pertaining to the lack of validity of CAPD diagnosis (as a standalone entity) and urged clinicians to provide language based vs. specific auditory interventions which lack treatment utility. Here, Dr. Kamhi makes a great point that: “Interventions that target processing skills are particularly appealing because they offer the promise of improving language and learning deficits without having to directly target the specific knowledge and skills required to be a proficient speaker, listener, reader, and writer.” (95) The problem is that we have numerous studies on the topic of improvement of isolated skills (e.g., auditory skills, working memory, slow processing, etc.) which clearly indicate lack of effectiveness of these interventions.  As such, “practitioners should be highly skeptical of interventions that promise quick fixes for language and learning disabilities” (96).

Now let us move on to language and particularly the models we provide to our clients to encourage greater verbal output. Research indicates that when clinicians are attempting to expand children’s utterances, they need to provide well-formed language models. Studies show that children select strong input when its surrounded by weaker input (the surrounding weaker syllables make stronger syllables stand out).  As such, clinicians should expand upon/comment on what clients are saying with grammatically complete models vs. telegraphic productions.

From there lets us take a look at Dr. Kamhi’s recommendations for grammar and syntax. Grammatical development goes much further than addressing Brown’s morphemes in therapy and calling it a day. As such, it is important to understand that children with developmental language disorders (DLD) (#DevLang) do not have difficulty acquiring all morphemes. Rather studies have shown that they have difficulty learning grammatical morphemes that reflect tense and agreement  (e.g., third-person singular, past tense, auxiliaries, copulas, etc.). As such, use of measures developed by Hadley & Holt, 2006; Hadley & Short, 2005 (e.g., Tense Marker Total & Productivity Score) can yield helpful information regarding which grammatical structures to target in therapy.

With respect to syntax, Dr. Kamhi notes that many clinicians erroneously believe that complex syntax should be targeted when children are much older. The Common Core State Standards do not help this cause further, since according to the CCSS complex syntax should be targeted 2-3 grades, which is far too late. Typically developing children begin developing complex syntax around 2 years of age and begin readily producing it around 3 years of age. As such, clinicians should begin targeting complex syntax in preschool years and not wait until the children have mastered all morphemes and clauses (97)

Finally, Dr. Kamhi wraps up his article by offering suggestions regarding prioritizing intervention goals. Here, he explains that goal prioritization is affected by

• clinician experience and competencies
• the degree of collaboration with other professionals
• type of service delivery model
• client/student factors

He provides a hypothetical case scenario in which the teaching responsibilities are divvied up between three professionals, with SLP in charge of targeting narrative discourse. Here, he explains that targeting narratives does not involve targeting sequencing abilities. “The ability to understand and recall events in a story or script depends on conceptual understanding of the topic and attentional/memory abilities, not sequencing ability.”  He emphasizes that sequencing is not a distinct cognitive process that requires isolated treatment. Yet many SLPs “continue to believe that  sequencing is a distinct processing skill that needs to be assessed and treated.” (99)

Dr. Kamhi supports the above point by providing an example of two passages. One, which describes a random order of events, and another which follows a logical order of events. He then points out that the randomly ordered story relies exclusively on attention and memory in terms of “sequencing”, while the second story reduces demands on memory due to its logical flow of events. As such, he points out that retelling deficits seemingly related to sequencing, tend to be actually due to “limitations in attention, working memory, and/or conceptual knowledge“. Hence, instead of targeting sequencing abilities in therapy, SLPs should instead use contextualized language intervention to target aspects of narrative development (macro and microstructural elements).

Furthermore, here it is also important to note that the “sequencing fallacy” affects more than just narratives. It is very prevalent in the intervention process in the form of the ubiquitous “following directions” goal/s. Many clinicians readily create this goal for their clients due to their belief that it will result in functional therapeutic language gains. However, when one really begins to deconstruct this goal, one will realize that it involves a number of discrete abilities including: memory, attention, concept knowledge, inferencing, etc.  Consequently, targeting the above goal will not result in any functional gains for the students (their memory abilities will not magically improve as a result of it). Instead, targeting specific language and conceptual goals  (e.g., answering questions, producing complex sentences, etc.) and increasing the students’ overall listening comprehension and verbal expression will result in improvements in the areas of attention, memory, and processing, including their ability to follow complex directions.

There you have it! Ten practical suggestions from Dr. Kamhi ready for immediate implementation! And for more information, I highly recommend reading the other articles in the same clinical forum, all of which possess highly practical and relevant ideas for therapeutic implementation. They include:

• Clinical Scientists Improving Clinical Practices: In Thoughts and Actions
• Approaching Early Grammatical Intervention From a Sentence-Focused Framework
• What Works in Therapy: Further Thoughts on Improving Clinical Practice for Children With Language Disorders
• Improving Clinical Practice: A School-Age and School-Based Perspective
• Improving Clinical Services: Be Aware of Fuzzy Connections Between Principles and Strategies
• One Size Does Not Fit All: Improving Clinical Practice in Older Children and Adolescents With Language and Learning Disorders
• Language Intervention at the Middle School: Complex Talk Reflects Complex Thought
• Using Our Knowledge of Typical Language Development

References:

Kamhi, A. (2014). Improving clinical practices for children with language and learning disorders.  Language, Speech, and Hearing Services in Schools, 45(2), 92-103

Helpful Social Media Resources:

SLPs for Evidence-Based Practice