Phonological Awareness Assessment: Where the Construct Ends
You have a nine-year-old on your caseload who decodes slowly, guesses at unfamiliar words, and cannot hold a new word in memory from one session to the next. Everything you know says to look at phoneme awareness.
So you go to pick a phonological awareness assessment, and you discover that the four instruments on the shelf do not measure the same thing, do not agree on where the construct starts, and mostly stop before your student's age.
That is not a gap in your knowledge. It is a property of the field.
Phonological awareness is an umbrella term, and every instrument decides for itself how wide the umbrella is. One normed test scores sentence segmentation inside the same battery as phoneme deletion. Another scores three phoneme tasks and calls the result a phonological awareness composite.
A third changes which subtests make up that composite on the student's seventh birthday. A fourth is group-administered and samples one sound contrast per version. And the one measure that scores the dimension most relevant to an older struggling reader has no norms at all.
This piece compares what each instrument actually samples, in the publishers' own words, and names the places where the published record runs out.
The question is not which phonological awareness test is best. It is which grain size, which operation, and which dimension you need measured, and then which instrument happens to sample that combination at your student's age.
The three decisions buried inside one word
Every phonological awareness task makes three separate choices, and no two instruments make them the same way.
Grain size
Spoken language has units inside units: sentences contain words, words contain syllables, syllables split into onset and rime, and rimes break into phonemes. A task that asks a child to clap the syllables in butterfly and a task that asks a child to say slide without the /s/ are both phonological awareness tasks, and they sample opposite ends of that hierarchy.
The What Works Clearinghouse practice guide on foundational reading skills puts the hierarchy in its glossary plainly. Phonological awareness, it says, "is the ability to recognize that words are made up of individual sound units," and it is "an umbrella term that is used to refer to a student's sensitivity to any aspect of phonological structure in language."
What the umbrella covers is then spelled out. It "encompasses awareness of individual words in sentences, syllables, and onset-rime segments, as well as awareness of individual phonemes."
The National Center on Improving Literacy draws the same umbrella and places phonemic awareness inside it as the most advanced skill, running the sequence from rhyming and alliteration through sentence segmentation, syllables and onset-rime to phonemes.
So when a report says a student is "below average in phonological awareness," the sentence is incomplete until you know which rung of that ladder the instrument stood on.
Operation
Within any one grain size, a task can ask the student to match, to blend, to segment, to delete, to substitute, or to reverse. These are not interchangeable. Segmenting sled into four sounds and turning sled into led by removing one of them are different cognitive demands, and the research literature treats them differently, as the next section shows.
Dimension
A response can be scored right or wrong. It can also be scored for how fast it arrived. Almost every normed instrument in this area scores only the first. Two do something else, and the difference turns out to matter most for exactly the student described at the top of this article.
The CTOPP-2 changes its own composite at age seven
The Comprehensive Test of Phonological Processing, Second Edition is the instrument most practitioners name first. It covers ages 4-0 through 24-11, takes about forty minutes, is individually administered, and carries twelve subtests feeding five composites.
Its Phonological Awareness Composite is where the interesting thing happens. PRO-ED's own product page states that the composite "comprises the standard scores of three subtests-Elision, Blending Words, and Sound Matching-for 4 through 6 year olds, and Elision, Blending Words, and Phoneme Isolation for 7 through 24 year olds."
Read that twice. The composite keeps its name across the whole age range, and swaps one of its three constituent subtests at age seven. Sound Matching, which asks the child to identify words beginning or ending with the same sound, is replaced by Phoneme Isolation, which asks the child to pull a specified sound out of a word.
A phonological awareness composite score for a six-year-old and a phonological awareness composite score for a seven-year-old are not measurements of the same set of tasks, even though they appear in the same column of the same report.
That is a reasonable design decision. A sound-matching task floors out for older students, and the publisher says as much in describing what the second edition fixed. It is also a fact nobody mentions when two scores from different ages get laid side by side in a reevaluation.
The Alternate composite is where the segmenting lives
For ages 7 through 24 the CTOPP-2 also offers an Alternate Phonological Awareness Composite, built from Blending Nonwords and Segmenting Nonwords. The word "alternate" is doing real work: segmentation is not in the primary composite for older students at all. If your referral question is about segmenting, you are reading a supplementary score, not the headline one.
The norms are old, and the age range and the sample are described differently
PRO-ED states that "all new normative data were collected in 2008 and 2009" and that "the test was normed on 1,900 individuals ranging in age from 6 through 24 years." The same page lists the age range as 4-0 through 24-11.
Those two sentences sit on the same page. We are not in a position to reconcile them from public material, and the examiner's manual is where the normative sample's lower bound should be checked before anyone leans on a four-year-old's or five-year-old's standard score. If that score is load-bearing for a decision, open the manual rather than the product page.
Either way the data behind every CTOPP-2 standard score in use today were collected around seventeen years ago, which is the kind of thing we have written about before in what actually makes a set of test norms too old.
The PAT-2:NU is the one that samples the whole ladder
The Phonological Awareness Test-2: Normative Update runs ages 5-0 through 9-11 and takes forty to fifty minutes. PRO-ED describes it as "a standardized assessment of phonological awareness, phoneme-grapheme correspondence, and phonemic decoding skills," with six core subtests "that measure awareness of spoken syllables and phonemes."
Those six are Rhyming, Segmentation, Isolation, Deletion, Substitution and Blending. The Segmentation subtest alone covers "Sentences, Syllables, and Phonemes." The Deletion subtest covers "Compound Words, Syllables, and Phonemes."
So one test walks the entire grain-size hierarchy, from dividing a sentence into words down to deleting a phoneme, and reports a subtest profile across it. Two supplemental subtests, Phoneme-Grapheme Correspondence and Phonemic Decoding, then cross into print.
That breadth is the reason to reach for it, and its ceiling is the reason you often cannot. The test stops at nine years, eleven months.
For the fourth-grader who is the archetypal educational therapy referral, you are at the top of the range on the day you open the box, and past it a year later. PAR lists the same instrument at the same qualification level with the same age band, so there is no alternative edition to reach for.
The TOPA-2+ is named for the construct and samples the least of it
The Test of Phonological Awareness, Second Edition: PLUS is described by its publisher as "a group-administered, norm-referenced measure of phonological awareness." It comes in two versions: a Kindergarten version for five- and six-year-olds and an Early Elementary version for six- to eight-year-olds.
What it asks is narrower than its title suggests. The Kindergarten version asks whether two words start with the same sound, and adds a letter-sounds subtest. The Early Elementary version asks whether two words end with the same sound, and adds a pseudoword spelling task in which the score is the total number of words spelled correctly.
The instrument named most directly for the construct samples it at a single point: one sound position, one comparison, one version at a time.
That is not a criticism of the test, which was built as an efficient group screener and says so. It is a warning about inference. A standard score from an instrument called the Test of Phonological Awareness tells you about initial or final sound comparison in a group-administered format, and it does not tell you whether the student can delete a medial phoneme.
Its norms are older still: PRO-ED says "all new normative data were collected during the years 2002 and 2003," keyed to 2000 census data, from 2,085 students across 26 states.
The LAC-3 is measuring something adjacent, and says so
The Lindamood Auditory Conceptualization Test, Third Edition is often filed with the phonological awareness tests. Its publisher does not describe it that way. PRO-ED calls it "a norm-referenced assessment that measures an individual's ability to perceive and conceptualize speech sounds using a visual medium," adding that it "also measures the cognitive ability to distinguish and manipulate sounds."
The visual medium is the point. The LAC-3 has the student represent sounds with coloured blocks and then track changes to them, which makes the task a conceptualization-and-representation task rather than a purely oral one. Its categories include Tracking Phonemes (Monosyllables), Counting Syllables (Multisyllables), Tracking Syllables (Multisyllables), and Tracking Syllables and Phonemes (Multisyllables), so it reaches the multisyllabic level that most of this shelf does not.
Its age range runs 5-0 through 18-11, which is the widest of the dedicated instruments here, and it reports "scores, percentile ranks, and age and grade equivalents." On grade equivalents specifically, we have set out what the publishers themselves say about that metric, and the short version is to read the percentile.
One sentence on that page deserves to be read closely. PRO-ED states that the characteristics of the normative sample "are the same as those reported for 2001 in the Statistical Abstract of the United States (U.S.
Bureau of the Census, 2001) and, therefore, are representative of the current U.S. population." The sample was matched to 2001 population figures. The claim that this makes it representative of "the current" population is a sentence written for a 2004 product page and still sitting on the live page today.
The achievement batteries added the dimension everyone else left out
Two of the big batteries have moved, and in opposite directions.
The WIAT-4 scores speed on phoneme manipulation, and calls it best practice
Pearson's own flyer for the Wechsler Individual Achievement Test, Fourth Edition states it without hedging: "The new Phonemic Proficiency subtest follows best practices by measuring speed AND accuracy on phoneme manipulation tasks!"
The same flyer puts Phonemic Proficiency inside a Phonological Processing composite alongside Pseudoword Decoding, with a grade range of 1 through 12 and above. That is a normed, timed phoneme manipulation measure that runs through high school, which is a genuinely different object from everything described above.
The WJ V splits the construct into two clusters
Riverside Insights' Woodcock-Johnson V brochure lists Phonological Awareness and Phonological Manipulation as two separate new clusters, fed by new tests including Sound Reversal and Rapid Phoneme Naming alongside the Sound Blending and Segmentation tests carried forward from the previous edition.
Having split the construct into two clusters, the same publisher's CHC ability definitions then define the underlying narrow ability, Phonetic Coding, as "the ability to distinctly hear phonemes; blend sounds into words; and segment words into parts, sounds, or phonemes," and add that "this ability is also frequently referred to as phonological processing, phonological awareness, and phonemic awareness."
One publisher, in two documents, both separates phonological awareness from phonological manipulation and states that phonological awareness, phonemic awareness and phonological processing are three names for one ability.
Both statements are defensible. Together they are the cleanest available evidence that the vocabulary in this area is not stable even inside a single test battery, which is the same structural finding we reached about the word "screener" in dyslexia screening and about what "aligned to the Knowledge and Practice Standards" claims.
The free test that scores a dimension the normed ones do not
The Phonological Awareness Screening Test, or PAST, carries a 2003, 2010 and 2016 copyright line and is adapted from the levels used by McInnis and by Rosner. Its initial report offers it "to educational professionals and researchers as a free, public domain test" and reports an administration time of five to eight minutes.
It is structured in four level bands: Basic Syllable out of twelve, Onset-Rime out of ten, Basic Phoneme out of ten and Advanced Phoneme out of twenty, for a total out of fifty-two. And every one of those is scored twice.
Correct, and automatic
The report describes the procedure in full. "When an item is given, the examiner mentally counts 'one thousand one, one thousand two.' If the student responds to that item before the examiner reaches the word 'two' in the silently counted phrase, the student receives an automatic score for that item."
And then the scoring rule: "All correct and automatic responses count toward the student's total correct score. However, only those items that were responded to within two seconds count toward the student's automatic score."
The record form reports a Highest Correct Level and a separate Highest Automatic Level, with a line for the levels not passed below each.
Why the second column exists
The argument for it is explicit and is about older students. Where a response takes three to five seconds or longer, the report states, "it is presumed that the student is strategizing, such as mentally spelling the word, manipulating the letter/sound as requested, and mentally 'reading back' the result. Such a strategy represents a mental application of phonics, and does not necessarily reflect developed phoneme awareness."
The developmental interaction is the part worth carrying into your practice. At first and second grade the timing element "had little impact on the interpretation of the test," because those students either answered correctly and instantly or not at all.
It is older students who "seem to be able to do this," and the report observes that older students with reading difficulties "are often able to get items correct on a phonemic awareness test, using non-instantaneous responding."
A student who answers every item correctly in four seconds each will look unimpaired on an accuracy-scored instrument. That is the student most likely to be sitting in front of you.
And the price of the second column
There are no norms. The record form says so in a footnote on its first page: the grade levels printed throughout "are estimates based on various research studies and clinical experience. There are no formalized norms."
So the PAST gives you a dimension nothing else on this shelf gives you, in five to eight minutes, for nothing, and it gives you no standard score, no percentile and no confidence band. Administration instructions are not on the form either; it refers the examiner to Chapter 11 of Equipped for Reading Success.
Both of those facts belong in your report if you use it. Describing a PAST result as though it were normative is the error to avoid, and the honest write-up names the measure, names the absence of norms, and reports the pattern rather than a score.
Segmentation is the task the screeners use and the task that fades first
This is the finding that reorganises the whole area, and it has two independent halves.
The published argument
The PAST's initial report takes a direct run at curriculum-based measures of segmentation. It allows that they do use timing, calling it "a valid idea," and then states the objection: "the CBM assessments rely on segmentation, which is a type of task that is less strongly correlated with reading, and which appears to lose its discriminant validity reading after first grade."
It adds that "these CBM tasks are not used in their respective batteries past first grade, so they are not used at an age level in which the timing element would be most helpful."
In its second-grade study the report describes finding "no significant correlations between the DIBELS Phoneme Segmentation subtest and the other measures, except for a significant negative correlation with the PAST."
Read the author's own caveat alongside it. The report states plainly that "the DIBELS Phoneme Segmentation subtest is used only with first graders in the DIBELS battery, so our findings in no way reflect on its usage with first graders," and attributes the negative correlation to a specific behaviour: students who did well on the PAST tended, wanting to answer quickly, to default to onset-rime segmentation, which lowered their fluency score.
That is a finding about using a first-grade measure out of band, not a finding that the measure is poor.
Where the screeners actually stop
The published administration documents settle the range question exactly.
The DIBELS 8th Edition benchmark administration timeline, in its June 2026 format update, lists Phoneme Segmentation Fluency under Kindergarten and Grade 1 and under no later grade. Oral Reading Fluency runs to Grade 8 and Maze to Grade 8; the only phonological measure stops after first grade.
Amplify's comparison of DIBELS 8 to previous editions states that DIBELS 8 "does not include" Initial Sound Fluency or First Sound Fluency, so the eighth edition carries exactly one phonological measure.
Acadience Reading's benchmark goals publish Phoneme Segmentation Fluency cut points at three time points only: kindergarten middle of year, kindergarten end of year, and first grade beginning of year. First Sound Fluency has goals at two kindergarten time points. There are no phonological awareness benchmarks after the start of first grade.
So the universal screening apparatus that generates most phonological awareness data in schools is, by design, finished with the construct before the end of first grade. For a caseload of third-graders and up, it produces nothing, which is a different problem from the one covered in our comparison of progress monitoring tools outside a school district but lives next door to it.
What the strongest evidence document says, and what it never mentions
The WWC practice guide on foundational reading skills is the most authoritative instructional document in this area. Its Recommendation 2 is "develop awareness of the segments of sounds in speech and how they link to letters," and the panel assigned it Strong Evidence.
That is the guide's top tier, defined there as positive findings "demonstrated in multiple well-designed, well-executed studies, leaving little or no doubt that the positive effects are caused by the recommended practice."
Two things about that recommendation are worth a practitioner's attention.
First, the evidence base is kindergarten and first grade. The guide states that the seventeen qualifying studies "included diverse American students in the relevant grades-kindergarten and 1st grade." The strongest claim the field can make about teaching this skill rests on a body of research conducted with five-, six- and seven-year-olds.
Second, the guide is almost silent on measurement. Across the full document, the word "screening" appears once and the word "norms" appears once, and both occurrences are inside the same endnote, which defines performance below grade level as potentially indicated by "low scores on screening measures of reading (relative to standardized norms or classroom performance) or the instructor's perception of students' abilities."
The field's strongest instructional evidence for phonological awareness and the field's instruments for measuring it do not appear in the same document.
That is not a defect in the guide, which set out to answer an instructional question. It does mean that nobody should cite a practice guide as authority for an assessment decision.
We made the same structural point about the gap between the published oral reading fluency table and the technical report behind it, and the lesson transfers: the document that gets cited and the document that contains the finding are frequently not the same document.
The qualification letter is not the gate here
Every dedicated instrument described above carries the same purchasing level. The CTOPP-2, the PAT-2:NU, the TOPA-2+ and the LAC-3 are all listed at Test Level B by PRO-ED. PAR lists the CTOPP-2 at B, and so does Pearson's own store.
That uniformity is unusual. We documented the opposite pattern in detail in what the qualification letters actually mean, where the same instrument carries different letters at different storefronts and the letters themselves mean different things from one publisher to the next.
Here, PRO-ED's qualifications policy sets Level B at a master's degree in psychology, education, occupational therapy, social work, counseling or a related field with formal training in ethical assessment administration, scoring and interpretation, or an equivalent route through professional certification, licensure or supervised training specific to child assessment.
Level A, by contrast, states that "there are no special qualifications to purchase these products," and Level C tests "require a high level of expertise in test interpretation."
So for most qualified educational therapists, purchasing is not the constraint. Competence on the specific instrument is. The AET Code of Ethics, Revision #9, puts three obligations in the same section: select and use appropriate assessment instruments "recognizing their limitations with respect to reliability, validity, and bias"; use only those instruments "for which they have been adequately trained"; and seek interpretation of assessment data from professionals in related fields.
The dotted clause references practitioners use for these, including the ones this site uses, are a pointing convention rather than the document's own printed numbering.
Applied to this shelf, the first of those three is the live one. A test whose norms were gathered in 2002 or 2008, whose composite changes composition at age seven, or whose stated construct is narrower than its title, has limitations a report should name.
Can any of this identify dyslexia or a learning disability?
No. None of the instruments described here identifies a disorder, and none of them should be described in a report as having done so.
A phonological awareness measure tells you how a student performed on a set of sound-manipulation tasks relative to whatever sample the publisher gathered, or, in the PAST's case, relative to no sample at all.
Identification is a determination made by qualified professionals in a defined process using a full picture, and the educational therapist's role in reading and using those reports is the subject of its own discussion in what you may do with a psychoeducational evaluation.
What these instruments are good for is instructional decisions: which grain size to work at, which operation is failing, whether accuracy or automaticity is the problem, and whether any of it changes over six weeks of teaching.
A decision sequence, not a decision
The sequence below is the order of questions, not a recommendation of an instrument. The right instrument falls out of the answers, and often the answer is that no normed instrument covers the case.
How old is the student? Above nine years eleven months, the PAT-2:NU is out. Above eight, the TOPA-2+ is out. Above eighteen eleven, the LAC-3 is out. The CTOPP-2 and the WIAT-4's Phonological Processing composite are the normed options that reach adolescence.
Which grain size is the referral question about? If it is syllables and sentences, the PAT-2:NU is the only dedicated test here that samples them within a normed profile. If it is phonemes only, the CTOPP-2's primary composite is built entirely from phoneme tasks for ages seven and up.
Which operation? Matching, isolation, blending, deletion, substitution and reversal are sampled unevenly. Check the subtest list, not the test title.
Accuracy, or accuracy and speed? If the student answers correctly but slowly, an accuracy-scored instrument will under-identify the difficulty. The WIAT-4's Phonemic Proficiency and the PAST are the two measures named here that score the dimension, one normed and one not.
What will you re-administer in eight weeks? Most of these have no alternate form and were not built for repeated use. Plan the progress measure before the diagnostic one, which is the same sequencing argument that runs through our piece on intervention session frequency and duration.
What will your report say the score does not mean? Name the norm vintage, the construct boundary, and the age-band caveat in the report itself rather than in a conversation nobody records.
Where this sits next to the programs
The instruments are only half the picture. Nearly every structured literacy program claims to build phonological awareness, and they bound it as differently as the tests do. Our comparison of reading intervention programs found that these are not even the same kind of object, and the piece on what "Orton-Gillingham based" actually means found that no organisation adjudicates the label at all.
The practical consequence is that a program's claim to address phonological awareness and a test's claim to measure it can point at different rungs of the ladder without either being wrong.
If you are choosing a training route rather than a program, the comparison of structured literacy training programs sets out what each route costs in hours and what credential it yields, and the broader question of what the evidence base for this work actually supports is covered in is educational therapy evidence-based.
For the wider question of how this role sits beside the others who assess reading, the scope map in what educational therapy is and the role comparison in educational therapist versus learning specialist versus executive function coach are the places to start.
The same structural problem, seven times over
The finding in this piece is not unusual on this site, because it is not unusual in this field. Each of the following reached the same shape from a different direction:
Dyslexia screening tools found seven different objects sold under one unregulated word.
Achievement test qualification levels found that the letter is a purchasing rule for one storefront, not an industry scale.
Grade equivalent scores found publishers cautioning against the metric they continue to print.
Outdated test norms found that the ten-year rule appears in none of the documents cited for it.
Reading program evidence ratings found five public systems rating something other than what a practitioner is choosing between.
Reading comprehension assessment tools found two instruments disagreeing about the same child because they define the construct differently.
Written expression assessment tools found the discipline boundary running straight through a single construct.
The common thread is that a shared word hides an unshared definition, and the practitioner is the person who finds out.
Where the published record runs out
Several things in this area cannot be established from publicly readable sources, and it is more useful to say so than to fill the gap.
The CTOPP-2's normative sample floor. PRO-ED's page describes the sample as ages 6 through 24 while listing the test for ages 4-0 through 24-11. The examiner's manual is the place to resolve it, and we have not read it.
Internal consistency and test-retest figures by subtest. These live in the manuals for every instrument here. Product pages report that evidence exists; they do not print the coefficients.
Standard errors of measurement for any of these composites. None of the public product pages carries one, so no confidence band can be reported from publicly available material, and that absence is exactly the problem described in our piece on outdated test norms.
Whether the PAST's automaticity criterion has been independently replicated. The reliability and validity material we read is the test author's own initial report.
How publishers' telepractice terms apply to these specific subtests. Remote administration is governed test by test, as set out in what publishers permit for remote assessment.
Frequently asked questions
Is phonemic awareness the same thing as phonological awareness?
No. Phonemic awareness is one part of phonological awareness, specifically the part concerning individual phonemes. The WWC practice guide glossary states that "phonemic awareness is a type of phonological awareness" and that phonological awareness is "an umbrella term." In marketing copy the two are often used interchangeably, and in test manuals they are not.
Which phonological awareness test covers the widest age range?
Among the dedicated instruments described here, the CTOPP-2 is listed for ages 4-0 through 24-11 and the LAC-3 for 5-0 through 18-11. Within the achievement batteries, Pearson lists the WIAT-4's Phonological Processing composite, which contains Phonemic Proficiency, at grades 1 through 12 and above.
Why do universal screeners stop measuring phonological awareness after first grade?
The published administration documents show the pattern without explaining it: DIBELS 8 lists Phoneme Segmentation Fluency for kindergarten and first grade only, and Acadience Reading publishes PSF benchmark goals at three time points ending at the start of first grade. The PAST's initial report offers one account, arguing that segmentation tasks lose discriminant validity with reading after first grade, with manipulation tasks continuing to index phonological skill later.
Can a phonological awareness test tell me whether a student has dyslexia?
No. These instruments describe performance on sound-manipulation tasks. They do not identify a disorder, and a report should not present a low composite as an identification.
Is a free measure like the PAST acceptable to use in private practice?
It is offered by its author as a free, public domain test, and it scores a dimension that normed instruments mostly do not. It also has no formalized norms, which its own record form states. The defensible use is as a criterion-referenced instructional measure described accurately in the report, not as a substitute for a normed score where a normed score is what the question requires.
How old is too old for a set of phonological awareness norms?
There is no published rule, and the instruments here were normed across a wide span, from 2002 and 2003 for the TOPA-2+ to 2008 and 2009 for the CTOPP-2. What matters is whether anything about the population or the construct has shifted in a way that affects the comparison, and whether your report says which year the comparison sample belongs to.
One more thing worth doing
If you take one practice change from this piece, make it the habit of writing the instrument's boundary into the report. Not "phonological awareness was in the low average range," but which grain size was sampled, which operations, whether speed was scored, and when the comparison sample was collected. It costs two sentences and it is the difference between a score and a finding.
Illuminate maintains a list for educational therapists, learning specialists and executive function coaches who want this kind of practice-level analysis as it is published, alongside referrals from our student-matching network. Join the educator list here.
Comments