Which reading test is the screener, which is the diagnostic?
A reading screener tells you who is off track. A diagnostic tells you what that child is missing. Most schools own two or three instruments and no clear answer about which of those jobs each one does, which is why the same question keeps coming up: we run the mandated check and something else, so which one is the diagnostic?
Usually neither. The Year 1 Phonics Check is a screener, and it says so in its own name in New South Wales and South Australia, where it is officially the Phonics Screening Check. It sorts a cohort into fluent, developing and struggling decoders. The score answers who. Read item by item, the papers can also point at likely gaps, which is why the official guide walks you through grouping misses by code type. That is useful follow-up, but the check was built to screen, so treat what it suggests as a lead for a closer look rather than a full profile.
Three jobs, not two
The Australian Education Research Organisation's decision tree for a multi-tiered system of supports covers three types of assessment: universal screening, diagnostic, and progress monitoring. Its guide to choosing reading assessments defines the first two and hands progress monitoring to a separate AERO guide on interventions. One thing before you open the link: that guide is written for secondary schools. The three jobs are the same at any year level. What changes is which instruments you would reach for.
| Job | Who sits it | When | What the result lets you decide |
|---|---|---|---|
| Screen | Every student in the year level | Start of the year or on enrolment, then again across the year | Who needs a closer look |
| Diagnose | Only the students the screen flagged | Once, on referral | Which skills the teaching should target |
| Progress-monitor | Only students already on extra support | Weekly or fortnightly | Whether the support is working |
Screening looks at "the reading skills of a student population" and is "usually administered at the beginning of the school year or upon entry to a school as a new student". Where a student lands below the benchmark, the result tells you they need a diagnostic. That is the whole output: a name on a list.
A diagnostic is a different animal. It looks for which reading-related skills sit behind the difficulty so the intervention can target them, and it is "typically administered in a one-to-one or small group setting by someone trained in standardised assessments". That last clause is why nobody runs one on 24 children. One useful side effect once you have it, though: the pattern across the flagged students usually points at something to fix in the whole-class lesson too.
Three questions that sort any assessment
You do not need the manual for this. Ask three things about anything already in the staffroom cupboard.
- Who sits it? Everybody in the year level, or only the students something already flagged?
- When, and how often? Once or twice a year, once on referral, or every week or fortnight?
- What does the result let you decide? A category or a cut score, a named skill gap, or a change in a trend line?
Run the Year 1 Phonics Check through it. Whole cohort. One annual window. Three achievement bands. Screener on all three counts, with item-level results you can mine afterwards, and the official guide is explicit that the expected level is a guide rather than a pass mark. Two students on the same score can need completely different lessons, which is the whole argument of reading the papers by sound instead of by score.
Why so many schools end up with two screeners
Because many batteries are built to do two jobs at once.
The DIBELS 8th Edition Australasian version, relabelled for Australian year levels, describes itself as covering benchmark assessment and progress monitoring. Two jobs, one battery, and neither of them is diagnosis. Plenty of other tools sit in the same spot, so check the publisher's own description of what each one is for before you file it.
There is a fourth job hiding in there too. The placement tests bundled with a phonics programme, the one-minute word lists and the entry checks, are tied to that programme's teaching sequence. A result maps to "start at lesson 43". That is placement, not a skill profile, and it is a perfectly good use of ten minutes as long as nobody files it as a diagnostic.
One more source of the tangle. What a teacher wants in November is often a second screening form, not a diagnostic at all. The official guidance backs using a phonics check with Year 2 and older struggling readers, but each jurisdiction hands you one fixed set of items, so re-checking means finding another form. That is a screening problem wearing a diagnostic costume.
What makes a question diagnostic
A quiz item carries one bit of information: right or wrong. That is all a screener needs, since its job is counting correct answers against a benchmark. A diagnostic item is built so that each plausible wrong answer points at a different cause.
Pseudo-words are the clearest classroom example. A made-up word has no entry in the child's vocabulary, so it cannot be retrieved from memory or recognised as a heart word. There is no picture and no surrounding sentence to guess from. Correct reading is only possible by applying letter-sound knowledge and blending it, so a miss narrows sharply to the mechanics.
The official scoring rules make the contrast sharp. On a real word the correct phoneme is required, so reading blow to rhyme with cow is marked wrong. On a pseudo-word, every plausible pronunciation the child can justify is accepted. A real-word miss could be decoding, vocabulary or plain memory. A pseudo-word miss is much harder to explain any way other than blending.
One caution before you build a whole assessment out of them. Tim Shanahan's position is that nonsense-word tests earn their place early, Prep through the first half of Year 1, and that real-word identification becomes the clearer index of growth after that. Teaching to the nonsense-word test also ruins it as a measure.
The same logic runs through other modalities:
- A developmental spelling inventory scores which features of a word the child represented, the initial sound, the short vowel, the blend, the digraph, the inflected ending. Not how many words were wrong. The error pattern is the output.
- Accuracy and rate get tracked separately for the same reason. A child who reads accurately but slowly and a child who is both slow and inaccurate need different teaching, and one number covering both hides which you have. That distinction has its own six-minute routine.
The sequence, in order
Screen everyone, more than once a year. The International Dyslexia Association recommends screening all children at least three times a year through to Year 2, layering the measures by year: phonological awareness, rapid naming and letter-sound knowledge early, then oral vocabulary and word-recognition fluency in Year 1, then real and nonsense word identification, oral reading fluency and comprehension in Year 2. The mandated Year 1 check is one annual snapshot inside that, not the whole cadence.
Diagnose only the students the screen flagged. It is the expensive step, it is one-to-one with someone trained to run it, and spending it on a whole cohort is how schools end up assessing instead of teaching.
Progress-monitor only the students getting extra support. Weekly or fortnightly, brief enough that you will actually repeat it, and sensitive enough to show growth in a single skill over a few weeks. If the trend lines stay flat, there are three things to check before you change the programme. Home-made tools are tempting here and hard to hold steady across a year, which is most of the argument for buying something.
Two limits worth stating. A classroom diagnostic is not diagnostic of a condition: identifying dyslexia or a specific learning disorder needs referral to a trained specialist, and no phonics assessment gets you there. And the informal version most of us actually do, sitting with the marked papers and grouping missed words by grapheme, is lightweight error analysis rather than a formal diagnostic. Usually enough to pick a reteach target. Not the same thing as a battery.
Where this leaves your data folder
Take the instruments you already own, ask the three questions, and write the answer on the front of each folder. If that leaves you holding two screeners, a placement test and nothing that actually diagnoses, that is a fixable gap now it has a name.
The step after the screen is the slow one: turning a list of flagged sounds into lessons that only use words those students can actually read. Phonicspal tags every word by grapheme, so a group's shared gap becomes the next lesson without you sorting word lists by hand. You can see how the lesson planning works or start free.
The screening and diagnostic definitions quoted above are from the Australian Education Research Organisation's practice guide on choosing reading assessments, published at edresearch.edu.au.