My Year 1 reader can sound out every word, but slowly: how to build automaticity

If your Year 1 reader can sound out every word correctly, but every sound is a pause-and-think effort, nothing has gone wrong. Accurate-but-slow is a recognised stage of learning to read, and the fix is specific: short decodable texts the child finds easy, read three or four times each, with an adult listening and correcting. About six minutes a day. More accuracy drill will not move it, and neither will quiet reading time. Here is why, and what the routine looks like in a Term 3 classroom.

Accuracy is not the finish line

If you teach in a Victorian government school you are running the Year 1 Phonics Check right now: 40 words read aloud, 28 the expected level, and a separate piece on reading those results by sound if the score itself is your question this week.

Nowhere on that score sheet is there a box for how long each word took. The check measures decoding accuracy only, so a child can pass at 28, or 38, one effortful sound at a time. It certifies the code is in. It cannot tell you whether the code is automatic, and Australian research by Bell, Farrell-Whelan and Wheldall found that decoding speed predicted check outcomes anyway. Speed matters even at the accuracy stage.

The reader profile has a name. Kuhn and Stahl's developmental model puts accurate but slow, halting reading at stage two of four, after laboured decoding, before phrased reading. Each stage needs different teaching, and the right emphasis for stage two is automaticity building. The common mistake is to reach back for stage-one medicine, more accuracy drill on sounds the child already knows.

The curriculum agrees, quietly. The Year 1 content descriptions already name phrasing and fluency alongside phonic knowledge.

What slow decoding costs

Working memory is small and fast-fading. The phonological loop holds roughly two seconds of speech, and young children can hold as few as two or three novel items at once. When a word is long or its letter-sound links are unfamiliar, the first sounds decay before the blend completes.

That budget shortfall is behind a pattern you have seen: a child reads a page accurately, then cannot tell you what happened in it. LaBerge and Samuels called it automaticity theory back in 1974. While word recognition still consumes attention, comprehension starves. Only when recognition becomes automatic is working memory freed for meaning, the bridge Pikulski and Chard later described.

You can hear the stage in the pauses. Schwanenflugel and colleagues found children with faster decoding made shorter, less variable pauses when reading aloud. A follow-up longitudinal study by Miller and Schwanenflugel traced where the pauses lead: children who made fewer pause intrusions in their first year of instruction were more likely to read with adult-like intonation a year later, and both predicted stronger comprehension the year after that. The pause-and-think reading is measurable, and it has a known trajectory.

One more number that reframes the whole problem. The average reader needs somewhere between 4 and 14 encounters with a word before recognising it automatically, and struggling readers may need 40 or more. Re-reading is dosage.

The fix: guided re-reading

The evidence here is unusually clean. Therrien's meta-analysis of repeated reading found fluency gains of d = 0.83 on practised passages and d = 0.50 on new ones, with three to four re-readings an effective dose. In the studies where children read a passage seven times, comprehension gains stopped being significant after the third reading.

The active ingredient is you. Conditions with adult corrective feedback transferred to new texts far better than re-reading without support. The National Reading Panel reached the same place from 99 comparisons: guided repeated oral reading improved word recognition, fluency and comprehension (d = 0.41), while sustained silent reading showed no evidence of improving fluency at all. Sending the slow decoder off to read quietly on the mat is the one intervention we know does not work.

Which stings a little, given Allington's finding that a typical 90-minute reading block contains about 15 minutes of actual reading. The slow decoder needs those minutes more than anyone.

The routine, six minutes a day

Take a decodable text of 60 to 120 words, built only from graphemes the child already knows. Then:

1. Read one, accuracy. Child reads aloud, you correct on the spot.
2. Read two, pace. Same text, aiming for fewer pauses. Point out the improvement, they always improve.
3. Read three, expression. Read it like talking, then chat about what it said.

Three reads of a text that length total roughly six or seven minutes. It fits inside any literacy block, and it works one-on-one, in a small group taking turns, or as a partner routine. The NSW Department of Education's Effective Reading guidance recommends exactly this, students "reread familiar decodable texts independently or with a partner", focusing on rate and intonation.

Two free upgrades. Reading the passage aloud to the child first, before they touch it, reliably lifts words correct per minute and cuts errors, and a live teacher model beat a recording in one comparison. And for the parent who asks what to do at home: paired reading, where the adult reads along and the child takes over when ready, shows some of the largest effects in the fluency literature.

Pick texts they can nearly ace

Fluency practice runs on different fuel from instruction. A text for teaching a new GPC sits at 95 to 97 per cent accuracy. A text for building rate should sit at 98 per cent or better, the independent level, which means it will look too easy to you. That is correct. Below 95 per cent the child is back in effortful territory and off-task behaviour climbs.

So reach behind the teaching edge. If the class is on vowel digraphs, the fluency text is built from the short vowels and consonant digraphs they finished last term. Mesmer's research adds a useful caution here: children read tightly decodable text more accurately but levelled text faster, because very dense decodability invites grapheme-by-grapheme attention. Easy and familiar beats new and hard, every time, for this job.

And you do not need to grind the same six readers from the Term 2 box. Kuhn's Year 2 study found classrooms reading three different scaffolded texts a week matched or beat classrooms re-reading one. Variety at the same phonics level counts as practice. A blend, one text re-read across the week plus a fresh one at the same level, gives you both effects.

How slow is too slow?

Australia has no national oral reading fluency norms, so most schools lean on the US Hasbrouck and Tindal set as a guide:

A practical screening rule: a child 10 or more words per minute below the 50th percentile on two unpractised readings warrants a deliberate fluency routine. A Year 1 reader crawling through text in Term 3 is on the runway, not off it.

Hold the norms loosely, though. Studies have found adequate comprehension anywhere from 47 to 78 words per minute, so rate is a means, never the goal. No stopwatches waved at six-year-olds.

One caveat worth knowing. A small group of accurate-but-slow readers carry a naming-speed deficit alongside intact phonics, the double-deficit profile, and they stay slow despite good teaching and honest practice. If a term of daily re-reading moves accuracy but the rate needle barely twitches, that is a conversation with your learning support team, not more of the same.

The supply problem

The routine above needs one thing in steady supply: short decodable texts pitched a step behind each group, and fresh ones each week so the practice does not go stale. That supply is the part Phonicspal covers, decodable stories and passages matched to exactly the sounds your class has been taught, at whatever point in your sequence each group sits. See how fluency practice works, if Term 3 has left the reader box thin.

The child who sounds out everything correctly has done the hard part. The code is in. Six minutes a day is what turns it on.

---

Victorian Phonics Check details are published at vic.gov.au.