In articulation therapy, a trial is one attempt at the target sound at the level you are practicing (a syllable, a word, a phrase or a sentence), scored correct or incorrect against a rule you set before the session. Progress monitoring means counting trials the same way every session, keeping independent productions apart from cued ones, and checking words you have not practiced (probes) to see whether the sound carries over. IDEA requires each IEP to describe how progress toward its goals will be measured, but it sets no accuracy level, and the familiar 80% across three sessions is a convention rather than a research finding.
Sources were checked on 1 November 2026.
What counts as a trial in articulation therapy?
IDEA does not define a trial, and the ASHA Practice Portal page on speech sound disorders gives no scoring rule, so the rule is yours. What matters is that you write it down and score every session the same way. These are the decisions that change the numbers most:
What happens | A common way to score it |
|---|---|
The student says the target correctly with no help | Correct, independent |
You model the word and the student imitates it correctly | Correct, with a cue |
You give a placement or gestural cue first, then the student gets it | Correct, with a cue |
The target sound is right but another sound in the word is wrong | Correct for a goal about the target sound; note the other error if it matters |
The target is distorted (for example a lateral /s/) | Incorrect, if the goal is accurate production |
The student self-corrects (“thun… sun”) | Score the first attempt; note self-corrections separately if you track them |
A sentence contains the target twice | Either each target is a trial or the sentence is one trial. Pick one and keep it. |
Score at the level named in the goal. Ten correct words do not show sentence-level accuracy, and the reverse is also true. If a session moves from words to phrases halfway through, keep two counts.
Why record independent and cued responses separately?
Because they measure different things. ASHA defines stimulability as “the child's ability to accurately imitate a model of the target speech sound that the child currently misarticulates.” A correct imitation shows the sound is within reach; it does not show the student uses it alone.
Research protocols treat supported and independent production as separate stages:
In a phonological treatment study by Gierut and Morrisette (2012), the imitation phase ran until 75% accuracy over two consecutive sessions, and the spontaneous naming phase until 90% over three consecutive sessions (each with a session limit). Imitated productions were held to a lower bar than independent ones.
Dynamic Temporal and Tactile Cueing, a treatment for childhood apraxia of speech, structures practice through four levels: simultaneous production, direct imitation, delayed imitation and spontaneous production (Grigos et al., 2024). The level of support is part of the data.
In practice, keep three counts: correct with no cue, correct with a cue, and incorrect. That gives you two percentages. In 20 trials with 14 independent correct and 4 more after a cue, independent accuracy is 70% and accuracy with cues is 90%. Reporting only the 90% would overstate what the student does alone; reporting only the 70% would hide how close the student is. If your district writes goals with “minimal” or “moderate” cues, define what each level means (for example, one verbal reminder versus a full model) so another clinician would score it the same way.
What is a probe, and how often should you probe?
Practice data tell you how the student does on words you are drilling. A probe checks words you have not practiced, usually with no cues, to see whether the sound generalizes. In the Gierut and Morrisette study, a full probe was given at baseline, at the move from imitation to spontaneous production, and again after treatment, with shorter subset probes on a schedule averaging every two sessions. Practice accuracy was used to decide when to move on; probe accuracy was used to measure generalization.
Probes also tell you when to stop drilling the same words. In a study of 19 children with phonological disorders, generalization occurred after teaching three word pairs in 59% of cases, five in 21% and ten in 14%, and did not occur after ten in 7% (Elbert, Powell & Swartzlander, 1991). Without a probe you cannot tell which of those groups a student is in.
A workable rhythm: a short probe of about 10 untrained words every second or third session, at the start of the session before any practice, scored with no cues. Use new words from the same target and position. Our lists for initial /s/, initial /k/ and initial /r/ are a source of untrained words that have been checked for the target sound.
Where does “80% accuracy across 3 sessions” come from?
Mostly from habit. In a 2026 survey of 125 school-based SLPs, most respondents used 80% accuracy across three sessions as their mastery criterion, and the most frequent reason was the belief that it would promote generalization and long-term retention (Schoepflin et al., 2026). The authors note “a reliance on familiar mastery thresholds, despite emerging evidence supporting higher accuracy thresholds to promote skill maintenance and generalization.”
The direct evidence on criterion level comes mostly from behavior analysis, with small samples:
Richling, Williams and Carr (2019) surveyed practitioners, who most often used 80% across three consecutive sessions, then tested it with 4 individuals with developmental disabilities. Their conclusion: 80% across three consecutive sessions “may be insufficient for producing maintenance in some cases.”
Fuller and Fienup (2018) compared 50%, 80% and 90% criteria with three children with autism; the 90% criterion produced higher maintenance.
In speech sound research, the Gierut and Morrisette protocol above required 90% over three sessions for spontaneous production.
ASHA's Practice Portal sets no number. It describes vertical treatment as practice “until the child reaches a specific criterion level (usually conversational level),” and for the Cycles approach says there is “no predetermined level of mastery.” Whatever criterion the IEP uses, apply it to the right number: a goal that says “independently” should be judged on independent accuracy.
How many trials do you need for a reliable percentage?
More than most data sheets allow. With 10 trials, one production moves the score by 10 percentage points, so 70% and 80% are a single word apart. With 20 trials each production is worth 5 points, and with 50 trials, 2 points. For an 80% criterion you need 8 of 10, 16 of 20 or 40 of 50.
Practice itself needs far more productions than a 10-trial sample: the most cited guideline is at least 50 production trials in a 30-minute session. Our guide to trials per session covers that research. You can score every practice trial or a fixed sample, but do it the same way every session so the graph compares like with like.
A simple data routine for a 30-minute session
Probe first (every second or third session). About 10 untrained words, no cues, before any practice.
Score every practice trial. Correct, correct with a cue, or incorrect, at the level the goal names.
Write both numbers in the note. For example: “/s/ initial, word level: 14/20 independent (70%), 18/20 with cues (90%). Probe: 6/10 untrained words (60%).” Our guide to SOAP notes for speech therapy has full examples.
Review the trend every two to three weeks. Move up a level when independent accuracy meets the criterion across sessions; if practice accuracy is high but probes stay low, change the words or the activity rather than adding more of the same.
Our free speech therapy accuracy calculator does the counting during the session: tap correct, with cue or incorrect, and it shows independent and with-cues accuracy and checks a criterion such as 80% across three sessions. Nothing is saved.
How Ga-loo tracks articulation progress
Disclosure: we build Ga-loo. On its Take data screen you tap every trial live (correct, correct with a cue, incorrect, undo) for one student or a group of up to 6, at word, phrase or sentence level, with a separate probe mode for untrained words that has no cue button. Each goal shows a graph against an 80% line, the baseline, the last three sessions, the latest probe, and independent and cued accuracy side by side. Session notes, plain or SOAP, and a progress-note draft are built from those numbers without AI.
Every weekly practice pack also includes a data sheet with untrained probe words. Students are kept under a code, not a name. The free plan covers 3 students with no card; Pro is $79 a year. Start free.
Frequently asked questions
What is the difference between a trial and a probe?
A trial is any scored attempt at the target. A probe is a set of trials on words you have not practiced, usually given without cues, to check whether the sound generalizes beyond the practice words.
Should cued responses count as correct?
Count them as correct with a cue, and report them separately from independent productions. Which number the goal is judged on depends on its wording: “independently” means independent accuracy.
Is 80% accuracy enough to call a sound mastered?
It is the most common criterion, not a proven one. The small studies that exist, mostly outside speech therapy, found better maintenance with 90% criteria, and speech sound research has used 90% over three sessions. The IEP team sets the criterion.
What is percent consonants correct (PCC)?
A measure of overall speech accuracy computed from a 5- to 10-minute conversational sample (Shriberg et al., 1997). The original severity ranges are mild 85–100%, mild to moderate 65–85%, moderate to severe 50–65% and severe below 50% (Shriberg & Kwiatkowski, 1982, as reported by Chenausky, 2022). It describes the whole speech system, while trial data track one goal.
Keep reading
Sources
IDEA regulations, 34 CFR 300.320 (content of the IEP): eCFR
ASHA Practice Portal, Speech Sound Disorders: Articulation and Phonology: asha.org
Gierut, J. A., & Morrisette, M. L. (2012). Age of word acquisition effects in treatment of children with phonological delays. Applied Psycholinguistics, 33(1), 121–144. Open access (PMC)
Grigos, M. I., Case, J., Lu, Y., & Lyu, Z. (2024). Dynamic Temporal and Tactile Cueing in young children with childhood apraxia of speech: A multiple single-case design. Journal of Speech, Language, and Hearing Research. Open access (PMC)
Elbert, M., Powell, T. W., & Swartzlander, P. (1991). Toward a technology of generalization: How many exemplars are sufficient? Journal of Speech and Hearing Research, 34(1), 81–87. doi:10.1044/jshr.3401.81
Schoepflin, E. K., Khamis-Dakwar, R., Gorman, B. S., & Cox, S. R. (2026). A survey of speech-language pathologists' current clinical practices in setting mastery criteria. Language, Speech, and Hearing Services in Schools, 57(4), 1193–1211. doi:10.1044/2026_LSHSS-25-00139
Richling, S. M., Williams, W. L., & Carr, J. E. (2019). The effects of different mastery criteria on the skill maintenance of children with developmental disabilities. Journal of Applied Behavior Analysis, 52(3), 701–717. doi:10.1002/jaba.580
Fuller, J. L., & Fienup, D. M. (2018). A preliminary analysis of mastery criterion level: Effects on response maintenance. Behavior Analysis in Practice, 11(1), 1–8. Open access (PMC)
Shriberg, L. D., Austin, D., Lewis, B. A., McSweeny, J. L., & Wilson, D. L. (1997). The percentage of consonants correct (PCC) metric: Extensions and reliability data. Journal of Speech, Language, and Hearing Research, 40(4), 708–722. doi:10.1044/jslhr.4004.708
Chenausky, K. V. (2022). The relationship between single-word speech severity and intelligibility in childhood apraxia of speech. Journal of Speech, Language, and Hearing Research, 65(3), 843–857. Open access (PMC)