For parents, educators, and speech-language pathologists working with children with developmental language disorder
One of the most frustrating conversations I have with parents and professionals happens after a reevaluation. Everyone has seen the child improve. The child explains ideas more clearly and asks better questions. They need less support, write more effectively, and repair misunderstandings.
They may advocate for help instead of shutting down. Then the standardized scores come back and they are essentially unchanged. The immediate question for DLD therapy is: Did therapy actually work?
Sometimes yes. Sometimes no. A flat standard score does not answer that question by itself. Standardized scores and functional progress measures are answering different questions, and I think we create unnecessary confusion when we treat them as interchangeable.
Why the score can stay flat
A norm-referenced standard score primarily tells us where a child is performing relative to same-age peers. It is not designed primarily to tell us whether the child learned the specific skills I treated. This distinction is especially important in developmental language disorder (DLD), because the comparison group does not stand still. As children get older, vocabulary, syntax, discourse, inference, academic language, writing, and self-advocacy demands all become more complex.
Understanding DLD: Score Variability Over Time
If a child earns a standard score of 75 at age 8, that does not mean the child learned nothing.
It can mean the child acquired new skills but remained at approximately the same relative position.
The child improved, but the gap did not close.
Additionally, the reverse can also happen.
In DLD, a child can show gains, while the standard score drops because growth lags behind peers.
Kwok and colleagues (2022) demonstrated this measurement problem when they compared different score types across repeated standardized language testing. Standard scores were less sensitive to within-child change than raw and growth-scale scores. Their sample involved young autistic children rather than children with DLD, but the psychometric point, however, is directly relevant: relative standing and within-child growth are not the same construct.
A flat score is not a free pass
We should be equally cautious about the opposite mistake. We cannot tell families, “The score stayed low, but therapy is working,” unless we can actually demonstrate that. A persistently low standardized score still matters. It tells us that the child continues to have significant difficulty relative to age expectations. Functional improvement does not erase the disorder.
If treatment is effective, we should be able to show change in the skills we targeted. That means more than saying the child “seems better.” Progress should be visible in repeated data, comparable samples, decreasing cuing, novel tasks, and real-life communication. Hall-Mills (2018) and Hoffman and Ireland (2018) both emphasized that language progress monitoring requires measures beyond standardized norm-referenced testing.
Measure what you actually treated
This sounds obvious, but it is often where progress monitoring breaks down. If we treated inference, we need repeated novel inference probes. If We treated narrative organization, we need comparable narrative samples. If we treated written expression, we need writing samples scored with stable criteria. If we treated self-advocacy, we need to know whether the child can recognize a problem, identify what information is missing, formulate an appropriate request, and actually obtain the help needed.
Gallagher and Hoover (2020) made this point particularly well in children with DLD: measures focused directly on the treated grammatical target detected change more consistently than broad language-sample indices. The study was small, so I would not overstate it, but the clinical principle is sound. Measure the behavior you are trying to change.
What to look for as evidence of functional improvement
For me, the strongest evidence is convergence. No single probe proves that therapy caused the change. Children are also developing, receiving instruction at school, practicing at home, and being exposed to other supports. But when several independent measures move in the same direction, the case for meaningful progress becomes much stronger.
Repeated target-specific probes: Brief, comparable measures tied directly to treatment goals, preferably using unfamiliar items rather than rehearsed materials.
Language and writing samples: Narrative, expository, conversational, or written samples that allow me to compare the actual features I treated over time. Socially focused language sampling can capture pragmatic behaviors that standardized measures may miss (Mulvey, 2026).
Cuing and independence: I want to know not only whether the child got the answer, but whether the child moved from maximal support to minimal prompting and, eventually, independent use.
Generalization: A skill is much more convincing when it appears with unfamiliar passages, new conversational partners, classroom work, or novel social and problem-solving situations.
Participation: Ultimately, I want to know whether communication is working better in daily life. Can the child ask for clarification, explain an academic idea, repair a misunderstanding, contribute to a group, advocate for an accommodation, or complete a language-heavy task with less support? Participation outcomes are still measured far less often than impairment-level outcomes in intervention research (McCormack et al., 2025), even though they are central to whether treatment is functionally meaningful.
The same score can describe a very different child
Consider a child with DLD who initially responds to an unclear assignment by saying, “I don’t know,” waiting silently, or completing the wrong task. At baseline, the child identifies the source of confusion in 2 of 10 novel situations and can formulate a specific clarification question only with substantial prompting. Six months later, the child identifies the missing information in 8 of 10 unfamiliar situations, formulates an appropriate question in 7 of 10, and begins using the strategy at school with occasional reminders. The standardized language score remains at the 7th percentile.
I would not describe that as “no progress.” The child has demonstrated acquisition, generalization, and increased independence. I also would not say the language disorder has resolved. The low percentile still indicates substantial difficulty relative to peers. Both findings can be true at the same time.
What I tell parents
I usually explain it this way: the standardized score tells us that the language difficulty is still significant compared with other children the same age. But it is not our only measure of progress. We also have to look at what the child can now do independently that previously required substantial support, whether those skills generalize to new situations, and whether communication is becoming more successful in school, at home, and socially.
That is the distinction I want families and professionals to understand. A standardized score should never be dismissed, but it also should not be asked to do a job it was not designed to do. If the score stays the same, the next question should not simply be, “Did therapy fail?” The better questions are: What changed? Can we measure it? Is the child more independent? Does the skill generalize? And does the child still have meaningful areas of impairment that require continued intervention?
When we can answer those questions with data, we have a much more accurate picture of progress than any single score can provide.
References
- Gallagher, J. F., & Hoover, J. R. (2020). Measure what you treat: Using language sample analysis for grammatical outcome measures in children with developmental language disorder. Perspectives of the ASHA Special Interest Groups, 5(2), 350–363.
- Hall-Mills, S. (2018). Language progress monitoring for elementary students. Perspectives of the ASHA Special Interest Groups, 3(1), 170–179.
- Hoffman, L. M., & Ireland, M. (2018). A panoramic introduction to language intervention progress monitoring. Perspectives of the ASHA Special Interest Groups, 3(1), 161–169.
- Kwok, E., Feiner, H., Grauzer, J., Kaat, A., & Roberts, M. Y. (2022). Measuring change during intervention using norm-referenced, standardized measures: A comparison of raw scores, standard scores, age equivalents, and growth scale values from the Preschool Language Scales–Fifth Edition. Journal of Speech, Language, and Hearing Research, 65(11), 4268–4279.
- McCormack, J., Ttofari, K., Denman, D., Dixon, G., Crosbie, S., & Cronin, A. (2025). A scoping review of intervention outcomes for school students with communication difficulties. Language, Speech, and Hearing Services in Schools, 56(1), 142–157.
- Mulvey, N. (2026). Flexible use of language sampling to measure social communication skills: A tutorial. Perspectives of the ASHA Special Interest Groups, 11(5), 1856–1870.