Articulation Screener

Tick errored sounds on a 30-word single-word probe covering every English consonant. Enter the child age and the tool flags each errored sound against the McLeod & Crowe (2018) 90% ages of mastery — headline pass / at-cusp / refer decision in one keystroke.

30-Word ProbeEvery English ConsonantMcLeod & Crowe 2018Client-Side
Tool details, related tools, and citation

Tick each errored sound

Elicit each target word, then tick the box beside every sound position where the child produced an error (substitution, omission, or distortion). Enter the child age to get an automatic pass / refer decision against the McLeod & Crowe (2018) 90% ages of mastery.

0 / 63 targets ticked
1
pig
animal
2
ball
toy
3
mom
family
4
no
everyday
5
hat
clothing
6
water
drink
7
dog
animal
8
top
toy
9
cat
animal
10
go
action
11
ring
object
12
fish
animal
13
van
vehicle
14
yes
everyday
15
leaf
nature
16
shoe
clothing
17
chip
food
18
jump
action
19
sun
nature
20
zoo
place
21
red
color
22
thumb
body
23
bathtub
object
24
feather
object
25
measure
action
26
banana
food
27
kitten
animal
28
pencil
school
29
five
number
30
spoon
object

Legend — I initial, M medial, F final. Tap a sound-position to mark it errored. Dialect features (th-stopping, /r/-vocalisation, /t/-glottalisation) should NOT be ticked.

Tick every sound-position where the child produced an error, then enter the child age to see the pass / refer decision.

Automate this workflow

Skip the manual count with ConductSpeech

ConductSpeech transcribes the audio, runs the analysis, and writes the clinical report — all in minutes instead of hours.

Automate this with ConductSpeech
  • SLP caseload intake — decide whether a new referral needs a full articulation evaluation
  • Well-child paediatric visits — flag children who need a speech referral
  • Classroom speech/language screening — quick 3-minute probe per child
  • IEP re-evaluation — confirm which sounds are still in error
  • Graduate SLP student training — practice sound-by-position error transcription
  • Parent/teacher intake forms — tick errored sounds on a printable probe sheet before the SLP visit

Don't use for

  • As a substitute for a norm-referenced articulation test (GFTA-3, Arizona-4, CAAP-2)
  • For children with diagnosed hearing loss, cleft palate, or apraxia of speech — these follow different developmental trajectories
  • For dialect speakers without dialect-specific norms — many "errors" are dialect features, not speech sound disorder
  • For bilingual children without cross-language transfer considerations
  • As the only source for IEP articulation goals — pair with a full evaluation and a connected-speech sample
  • For vowel errors or prosody — the screener is consonant-focused

What Is an Articulation Screener?

An articulation screener is a short single-word probe used to decide whether a child needs a full articulation evaluation. Unlike a norm-referenced test like the Goldman-Fristoe Test of Articulation 3 (53 target words, ~30-minute administration, standardised scoring) or the Arizona Articulation and Phonology Scale 4 (72 target words, composite severity score), a screener samples only a handful of targets — usually 20 to 50 — and gives a fast pass / refer answer. Screeners are the standard intake tool for preschool well-child visits, classroom speech screenings, SLP caseload triage, and IEP re-evaluations.

Two questions drive every screener. (1) Did the child produce any errors? (2) If yes, is each errored sound past its typical age of mastery? The second question is the whole point of a screener — a three-year-old who says "wabbit" for "rabbit" has a developmentally normal error and should NOT be flagged, while a six-year-old producing the same error has a delayed sound that warrants a full evaluation. The ConductScience screener asks both questions automatically once the child age is entered.
How this screener works. The tool presents 30 single-word targets picked to cover every English consonant in at least one position (initial, medial, or final). For each word the clinician ticks the error box beside every sound the child produced incorrectly. The tool then lists every errored consonant, compares each against the McLeod & Crowe (2018) 90% age of mastery, and flags the child pass, at-cusp, or refer. The full word list with target positions is visible inside the tool so you can use it as a printable probe sheet in the clinic room or on the iPad.

The 30-Word Target List

The 30-word list is modelled on the Goldman-Fristoe Test of Articulation 3 sounds-in-words structure and the Iowa-Nebraska Articulation Norms (Smit et al. 1990 / 1990 IA-NE), plus /r/, /l/, and /θ/ coverage drawn from the Arizona Articulation and Phonology Scale 4. The criteria for inclusion were:

Coverage. Every English consonant (/p, b, m, t, d, n, k, g, ŋ, f, v, s, z, ʃ, ʒ, θ\theta, ð, tʃ, dʒ, h, w, j, l, ɹ/) is sampled in at least one position, and the 20 most clinically targeted sounds are sampled in two or three positions.
Picturability. Every target is a concrete noun or a short picturable phrase so the word list can be elicited from a 3-year-old without a picture book or presented as a reading probe to a school-age child.
Developmental appropriateness. Every target is in the typical receptive vocabulary of a 3-year-old (MacArthur-Bates CDI / Dale & Fenson 1996).
Position sampling. Initial and final positions are sampled for every consonant where English phonotactics allow. Medial positions are sampled when they are clinically informative (e.g. intervocalic /t/ for flapping, medial /ŋ/ for nasal place).

The full list is inside the screener. Use the list as a printable probe sheet for a three-minute in-clinic screening, or as a reading probe for an SLP intake visit.

How to Score the Screener

For each target word, listen to the child's production and tick the error box beside every sound position where the child produced a consonant in a way that would be flagged on a single-word articulation test. Count the following as errors:

Substitutions. The child produced a different consonant than the target (e.g. "tat" for "cat" — /k/ → /t/ substitution).
Omissions. The child dropped the target consonant entirely (e.g. "ca" for "cat" — /t/ omission in final position).
Distortions. The child produced an unusual variant of the target consonant that would not be transcribed with a different IPA symbol but does not sound age-appropriate (e.g. a lateral /s/ or a frictionless /r/).
Additions. The child added a consonant that is not in the target (e.g. "stop" → "stops" in a reading probe). Additions are rare in single-word probes and are usually a sign of a separate morphological or literacy error rather than an articulation error.
  • Dialect features (th-stopping, /r/-vocalisation, rounding of /ɹ/ — these are African American English, Southern English, and other dialect variants, not errors)
  • Vowel errors (the screener is consonant-focused; vowel errors go in a separate probe)
  • Prosody, rate, or fluency issues (those go in the voice or fluency screeners)

The tool totals the errored sounds automatically, groups them by word position, and flags each errored sound against the McLeod & Crowe (2018) age of mastery.

The Pass / Refer Rule

The screener uses a single-sound age-of-mastery rule for the pass / refer decision, not a standardised total score. The rule:

Pass. Every errored sound is age-expected — the child is younger than the McLeod & Crowe (2018) 90% age of mastery for every errored sound. The screener returns "pass" and the clinician re-screens at the next visit.
At cusp. At least one errored sound is within ±3 months of the McLeod & Crowe (2018) 90% age of mastery, and no sound is clearly past its age of mastery. The screener returns "at cusp" and the clinician re-screens in 3 months or adds a stimulability check in the current session.
Refer. One or more errored sounds are past the McLeod & Crowe (2018) 90% age of mastery. The screener returns "refer" and the clinician follows up with a full articulation evaluation (GFTA-3, Arizona-4, CAAP-2) before writing IEP goals.
Why the single-sound rule. The screener deliberately avoids a composite score because composite scores on brief probes are noisy (ICC typically below 0.70 for word lists under 30 targets). A single clearly delayed sound is a more reliable refer signal than a borderline composite. For children with many errors, pair the screener with a full GFTA-3 or Arizona-4; the screener is not meant to substitute for a normative test. For suspected phonological disorder (pattern across many targets), add a phonological process analysis — the ConductScience Phonological Process Identifier is the companion tool.

Frequently Asked Questions