One of the most common pushbacks I hear from SLPs related to data collection—is that contextualized language therapy makes it too hard to collect data. When I suggest working on narrative or discourse to simultaneously address goals like syntax, vocabulary, and inferencing, the response is often, ‘But how do you collect data on that? We have to track data in schools!’ The irony is, I’ve seen the kind of ‘data’ that often gets collected in isolated drills—and let’s just say, it’s not exactly meaningful or functional. This misconception can be common among any clinicians who haven’t yet seen how powerful and measurable contextualized therapy can be.
This type of interactions aren’t rare. But here’s the truth: the idea that contextualized, academically relevant therapy somehow undermines data collection is not just outdated—it’s counterproductive. You can track data meaningfully while targeting multiple goals, and doing so within rich, engaging contexts (like narrative, thematic units, or expository texts) is exactly what students with language and literacy deficits need.
Let’s walk through what that looks like in practice—citing both the research and real-life SOAP notes to prove it.
What the Research Actually Says
Contextualized language intervention (CLI) is evidence-based and data-collectable.
“Instead of focusing only on short-term, situation-specific isolated intervention session skills, contextual intervention aims to accomplish long-term, situation-independent, generalizable skills learning” – Kamhi, 2014
“Targeting language in meaningful contexts increases saliency… which in turn promotes deeper vs. shallow knowledge and greater retention of information” – Gillam, Gillam, & Reece, 2012
CLI promotes gains across multiple domains: vocabulary, syntax, inference, discourse cohesion, and more. The secret? Topic continuity and intentional goal mapping.
Tracking Multiple Goals Within Context: Real SOAP Note Example
Let’s look at this student’s session:
Goal Vehicle: CGI film Catch It (ESMA)
This is the actual session note, which can serve as a helpful teaching tool for graduate students and clinical fellows.
| Subjective | Initially slightly reluctant to attend the session but was easily redirectable. Completed the session without any issues. |
| Objective Materials: Catch It ESMA (Meerkat) | 1. Student will answer social inference questions regarding the presented CGI film in 7 out of 7 trials. 2. Student will define 5 target vocabulary words (e.g., “protect,” “determined,” “cooperation,” “share”) and use each word in a complete sentence by the end of the session. 3. Student will identify the thoughts of characters in the CGI film in 4 out of 4 trials, using language such as “I think the vulture is thinking…” 4. Student will identify the feelings of characters in the CGI film in 5 out of 5 trials. |
| Assessment | 1. Student answered social inference questions regarding the presented CGI film in 4 out of 7 trials resulting in 57% accuracy. When asked, “What are the meerkats going to do now that the vulture took their fruit?” Student stated “They want to all work together, they have a plan”. Additionally, Student stated they were “angry” because “they wanted to keep it”. Errors resulted due to a provision of vague, nonspecific unrelated and imprecise responses which directly did not answer the clinician’s questions. For example, when asked “Based on what you know, what problem are the meerkats going to face”, Student erroneously responded “it’s like precious to them”. When asked, “What are they going to protect it from?” Student confusingly responded, “eagles… sharks”. Student exhibited inconsistent performance on this task, frequently providing responses that did not directly address the question or included irrelevant words not related to the story. 2. Student defined 1 out of 5 target vocabulary words resulting in 20% accuracy but did not use any words in a complete sentence by the end of the session. Student correctly defined “protect” [to guard] but could not recall the definition of “determination”, “cooperation”, “vulture”, and “share”. 3. Student identified the thoughts of characters in the CGI film in 2 out of 4 trials, resulting in 50% accuracy. He was not able to use language such as “I think the vulture is thinking…”. When asked if he thought the animal really wanted the fruit, Student responded, “No.” When further asked, “Why do you think they [the meerkats] have a plan?” he stated, “Because they wanted to keep it“. While Student identified the thought of characters, his utterances lacked structure and were simplistic. 4. Student identified the feelings of characters in the CGI film in 3 out of 5 trials resulting in 60% accuracy. After observing the vulture taking the fruit, Student was asked, “How did that make them [the meerkats] feel?” He responded, “Angry” When asked, “How is the vulture feeling towards the meerkat?” Student replied, “Mad,” and explained, “Because he growled“. When asked, “What is this look?” referring to the character’s mouth hanging open, Student answered, “Shocked,” demonstrating awareness of how emotions are associated with specific facial expressions. Additionally, with verbal prompting and phonemic cues, Student was able to identify other emotions such as “accomplished” and “determination.” |
| Plan | Continue to work on reviewing vocabulary, answering social-inferencing questions, and identifying feelings of characters via CGI films, Social Squad Videos, Social Problem Solving Scenarios, YouTube Video Clips as well as CommonLit Fictional Stories at 2nd grade level. |
Goals Tracked:
- Defining target vocabulary (5 words)
- Identifying character feelings (5 trials)
- Identifying character thoughts (4 trials)
- Answering social inference questions (7 trials)
Data Collected:
✅ “Student identified the thoughts of characters… in 2 out of 4 trials (50%).”
✅ “Student defined 1 out of 5 target vocabulary words, 20% accuracy.”
✅ “Student answered social inference questions in 4 out of 7 trials (57% accuracy).”
This SOAP note provides:
- Quantified performance per goal
- Qualitative feedback explaining why errors occurred
- Embedded linguistic targets within a naturalistic, academic task
Was the session focused on one single skill? No. Did that prevent clear data collection? Also no.
Why This Works: Explicit Data Meets Implicit Learning
According to Seidenberg & MacDonald (2018), language learning requires both explicit instruction and statistical learning (i.e., learning patterns through repeated meaningful exposure). By embedding structured tasks within narrative or expository texts, you maximize both.
Even better, functional language data can be collected simultaneously on:
- Expressive syntax (Student’s sentence structure)
- Vocabulary acquisition (Student’s word definitions)
- Pragmatics (inferences, thoughts, feelings)
- Reading fluency and accuracy (in connected expository text)
- Written language (via contextual sentence writing)
Take this example from a spelling session:
“Student wrote 10 sentences with CVC words in 10/10 trials spontaneously… applied capitalization but frequently omitted punctuation.”
This tells us about:
- Spelling accuracy
- Grammar application
- Sentence formulation
- Mechanics (capitalization/punctuation)
All tracked, all functional, all in one session.
Cognitive Load and Transfer: More Goals ≠ More Confusion
Kamhi (2014) reminds us: the problem isn’t generalization—it’s misunderstanding learning. Children don’t need less; they need meaningful, repeated, integrated input.
Contextualized tasks actually lighten cognitive load by:
- Using shared background knowledge
- Creating logical linguistic expectations (e.g., emotions follow actions, characters have motivations)
- Promoting schema-building through continuity
Compare this to decontextualized drills like:
- “Define ‘protect.’ Now define ‘cooperate.’”
- “Complete this sentence using a conjunction.”
Sure, you can gather data—but will it stick? Probably not.
Pro Tips for Practical, Data-Driven CLI
Here’s how you make this seamless and school-friendly:
1. Start With Language Rich Materials
Use CGI videos, YouTube clips, picture books, grade-level passages (e.g., ReadWorks, CommonLit). Select your context, then layer the goals.
2. Prewrite Data Targets
Your SOAP note goal grid can look like this:
| Goal | Trials | Accuracy | Notes |
|---|---|---|---|
| Vocabulary | 5 | 20% | Needed visuals, phonemic cues |
| Syntax | 10 | 80% | Omitted punctuation, inconsistent capitalization |
| Inference | 7 | 57% | Responses vague, off-topic |
3. Be Transparent With Teachers/Admins
Show them your data tables. Explain how CLI allows you to collect real data on the skills that transfer—like inference, sentence cohesion, vocabulary depth—not just “following 3-step directions.”
Bottom Line
You can do contextualized intervention and collect data. In fact, you must, if you want to achieve meaningful, generalizable learning gains. And when someone says, “But how do you collect data that way?”—show them your SOAP notes. Or better yet, invite them to sit in on a session.
Let’s retire the myth that real therapy can’t be real therapy unless it fits on a checkbox.
References:
- Gillam, S. L., Gillam, R. B., & Reece, K. (2012). Language outcomes of contextualized and decontextualized language intervention: Results of an early efficacy study. Language, Speech, and Hearing Services in Schools, 43(3), 276–291.
- Kamhi, A. G. (2014). Improving clinical practices for children with language and learning disorders. Language, Speech, and Hearing Services in Schools, 45(2), 92–103.
- Seidenberg, M. S., & MacDonald, M. C. (2018). The impact of language experience on language and reading: A statistical learning approach. Topics in Language Disorders, 38(1), 66–83.