Which part of phonics do you want to focus on? Here are a few common areas—tell me what you’re most curious about (you can pick one or more):

1. Basics: What phonics is vs. phonemic awareness (how they differ)
2. Letter–sound rules: short/long vowels, digraphs (sh, ch), blends (st, bl)
3. Syllables & decoding: breaking words into parts, sounding out longer words
4. Spelling patterns: CVC, CVCE, vowel teams (ai, oa), r-controlled (ar, er)
5. Tricky words: why “said,” “one,” “does” don’t follow common patterns
6. Teaching/learning methods: synthetic vs. analytic phonics, what works best
7. Reading difficulties: phonics support for dyslexia or struggling readers
8. Age/level: phonics for early readers vs. older learners or ESL/EFL students

Tell me:

• Are you learning phonics for yourself or to teach someone else?
• What age/reading level is the learner?
• What specifically feels confusing or interesting right now?

How can we predict the best time to stop explicitly teaching phonics for an individual child—using measurable indicators (e.g., decoding accuracy/fluency on pseudowords, oral reading rate, spelling of specific patterns)—so we don’t quit too early or keep drilling after it stops helping?

To investigate “when to stop explicitly teaching phonics for an individual child,” I’d study it as a decision-rule problem: What measurable indicators (and thresholds) predict that continued explicit phonics will yield little additional benefit compared with shifting time to other instruction?

1) Define the intervention decision you’re trying to optimize

Operationalize “stop explicit phonics” as something concrete, e.g.:

• Stop condition: reduce explicit phonics/word-level lessons to ≤5–10 min/day review; reallocate time to fluency + comprehension + vocabulary/morphology + writing
• Continue condition: keep explicit phonics/decoding lessons at the standard dosage (e.g., 20–30 min/day)

You’re not asking whether phonics matters; you’re asking when marginal returns drop for a given learner.

2) What you would measure (weekly/biweekly progress + periodic standardized tests)

A. Candidate “readiness to stop phonics” indicators (the predictors)

Collected weekly or biweekly:

1. Pseudoword decoding accuracy + fluency

   • e.g., DIBELS Nonsense Word Fluency (NWF) or an equivalent probe
   • Metrics: % accuracy, correct letter-sounds/whole words per minute, error patterns

2. Real-word reading accuracy + automaticity

   • timed word lists (regular + irregular)
   • Metrics: words correct per minute; proportion of slow-sounded-out vs automatic

3. Connected-text oral reading (decodable then controlled leveled text)

   • Metrics: WCPM, accuracy %, prosody rating, types of miscues

4. Spelling/encoding of taught patterns

   • brief dictated spelling aligned to the phonics scope and sequence
   • Metrics: pattern mastery %, error types (phoneme omission, vowel confusion, etc.)

5. Generalization checks

   • can the child decode untaught words containing taught correspondences?

B. Outcomes that determine whether stopping “worked” (the dependent variables)

Measured at baseline, decision point, and follow-ups (6–10 weeks later; end-of-term):

• Word reading efficiency (e.g., TOWRE-2 style outcomes: sight word + phonemic decoding efficiency)
• Multisyllabic decoding (controlled lists)
• Spelling transfer (untaught words, morphology-sensitive spelling if age-appropriate)
• Reading comprehension (brief passage comprehension, ideally controlling for listening comprehension)
• Optional but useful: motivation/avoidance, time-on-task, teacher-rated cognitive load

3) Data you’d need

For each student:

• Instructional dosage logs (minutes/week of explicit phonics; materials used; group size)
• Weekly/biweekly progress-monitoring scores (the predictors above)
• Periodic outcome tests (word reading, spelling, comprehension)
• Background moderators: age/grade, baseline phonemic awareness, language status, suspected dyslexia, etc.

4) The core experiment: a “threshold” trial to test stopping rules

Design: Randomized controlled trial at the “ready” moment

1. All students begin with the same systematic phonics program.
2. Define one or more stopping thresholds a priori, for example:
   • Pseudoword decoding ≥ 95% accuracy and
   • Oral reading accuracy ≥ 97% on controlled text and
   • Spelling of recently taught patterns ≥ 85–90% across 2 consecutive checks
3. When a student first meets the threshold, randomize them to:
   • STOP group: shift time away from explicit phonics (as defined above)
   • CONTINUE group: keep full phonics dosage for another 6–10 weeks
4. Compare outcomes after that period and at longer follow-up.

This directly tests: Once a child looks “ready,” does continuing phonics still cause measurable gains, or are gains equivalent when time is reallocated?

Analysis

• Primary: difference in growth on word reading efficiency + spelling transfer
• Secondary: comprehension and fluency
• Use growth models (mixed effects) to estimate marginal gains per additional week of phonics after threshold.

5) Add a second layer: optimize the indicator set (prediction study)

Run a parallel predictive modeling analysis:

• Use early indicators (pseudoword fluency, spelling mastery, error types, oral reading rate) to predict:
  1. who will maintain/improve after stopping, and
  2. who will regress or stall unless phonics continues.

Methods: logistic regression / decision trees to produce a practical rule (e.g., “if NWF fluency < X OR vowel-team spelling < Y, don’t stop yet”).

Key is to validate on a held-out sample so the rule isn’t just overfit to one cohort.

6) Smaller, practical alternative: single-case experimental designs (schools can do this)

If you can’t run an RCT, do multiple-baseline / ABAB-style studies:

• Baseline (A): phonics at standard dosage for several weeks (measure weekly probes)
• Shift (B): reduce phonics, increase fluency/morphology (continue weekly probes)
• Optionally return to A briefly if ethical/feasible, or stagger the shift across students (multiple baseline)

Look for:

• Immediate level/trend changes in decoding/spelling after dosage change
• Maintenance over time

7) What would count as evidence you’ve found a useful “stop” point?

A stopping rule is good if:

• STOP group outcomes are non-inferior to CONTINUE (decoding/spelling don’t drop; growth is similar)
• Students in STOP show better gains in whatever time was reallocated to (e.g., comprehension, fluency)
• The rule is accurate across subgroups (or you identify subgroup-specific thresholds)

If you tell me the age/grade range you care about (K–1 vs 2–3 vs older intervention), I can suggest realistic thresholds and a minimal measurement battery that a teacher/school could actually run.

If I got a clear, evidence-backed answer (a decision rule with measurable indicators) for when to stop explicit phonics for an individual child, it would change three main things:

Decisions that would be affected

• Instructional time allocation: I’d know when to reassign minutes from explicit phonics to fluency, vocabulary/morphology, comprehension, and writing—without guessing.
• Intervention intensity and staffing: Schools could decide who still needs small-group/1:1 decoding support versus who can move to a lighter “maintenance” plan, affecting schedules and resource deployment.
• Assessment cadence and exit criteria: Instead of exiting based on grade level or “finished the program,” we’d use specific mastery/transfer benchmarks (e.g., pseudoword decoding + spelling generalization + connected-text accuracy).

Assumptions that would be confirmed or overturned

• Confirmed/clarified: That “knowing the phonics rules” isn’t the same as being done—automaticity and transfer matter, and they can be measured.
• Potentially overturned:
  • The assumption that more phonics is always better once basic decoding looks solid. A clear answer might show a point of diminishing returns where continuing yields little added benefit compared with other instruction.
  • The opposite assumption—that you can stop as soon as a child can decode CVC/early patterns. The data might show many students need longer to reach stable fluency/orthographic mapping.
• Nuanced: It might reveal that the “stop point” differs by learner profile (e.g., dyslexia risk, language background), meaning a single universal cutoff is unrealistic.

Who else would care, and why

• Classroom teachers: because they constantly juggle limited time and need defensible criteria for moving students out of heavy phonics blocks.
• Reading specialists/interventionists: because “exit decisions” are high-stakes; a clear rule reduces both premature exit and endless remediation.
• School leaders/districts: because it affects scope-and-sequence pacing, intervention models (MTSS/RTI), and staffing budgets.
• Parents: because they want to know whether continued phonics tutoring is necessary or whether the child should shift to fluency/meaning-focused reading.
• Researchers and curriculum designers: because it informs how programs build review, mastery checks, and transition points (and could reduce over-teaching or under-teaching).

Net change: I’d move from an intuition-based transition (“seems fine, let’s move on”) to a transparent, testable exit criterion that better matches instruction to what each learner still needs.

1. Choose a working “stop-phonics” definition and timeframe

• Write a 1–2 sentence rule for what “stopping” means in your context (e.g., “reduce explicit phonics from 25 min/day to 5–10 min/day review; replace with 15 min fluency + 10 min morphology/vocab”).
• %%SGTERM:eb674967-0c51-465a-b47f-48c94ec32d41:Pick a short investigation window: 6 weeks%%, with the next 1 week used to set up measures and baseline.

2. Select 3 quick measures you can actually administer Use one from each bucket so you’re not relying on a single signal:

• Pseudoword decoding: DIBELS NWF (or any 1-minute nonsense-word probe).
• Connected-text oral reading: 1-minute oral reading (WCPM + accuracy) using a controlled passage/decodable.
• Spelling/encoding: a 10-word dictated spelling probe aligned to the phonics patterns you’ve taught recently.

Decide scoring rules now (accuracy %, WCPM, and error types).

3. Create a simple data sheet and baseline it this week

• Make a Google Sheet with columns: date, minutes of phonics instruction, NWF score, ORF WCPM, ORF accuracy, spelling %, notes on error patterns.
• Administer the 3 measures twice this week (e.g., Tue/Fri) to get a stable baseline.

4. Draft a provisional threshold to test Pick something you can revise later, but make it explicit. Example:

• NWF ≥ 95% accuracy (or above benchmark for grade) for 2 checks and
• ORF accuracy ≥ 97% and
• Spelling of taught patterns ≥ 85–90% Write it down as your “candidate stop point.”

5. Plan a tiny pilot comparison you can run next week (low-lift) If you have at least 2 similar students:

• Keep one student/group on full phonics dosage.
• For the other, reduce phonics and add fluency/morphology for the same total minutes. If you only have one student, do a mini A→B shift: keep phonics dosage stable this week (baseline), then reduce it next week and track whether decoding/spelling stays stable.

6. Pull 5–10 high-quality sources and extract “exit criteria” In the next week, do targeted reading (not broad browsing):

• Look for: “mastery criteria,” “exit criteria,” “nonword fluency benchmarks,” “response to intervention exit.”
• Make a 1-page note: what measures they use, any suggested cutoffs, and what outcomes they track after transition.

7. Schedule one expert conversation Email or message one person (reading specialist, dyslexia tutor, curriculum coach) with 3 precise questions:

• “What do you use as exit criteria from phonics intervention?”
• “Which measure do you trust most for transfer—pseudowords, oral reading, spelling?”
• “What failures do you see when kids exit too early?”

If you tell me the learner age/grade (and whether this is general ed vs intervention), I can suggest realistic cut scores/benchmarks and a one-page template for the data sheet + weekly routine.

Next-week action plan (practical, do-able)

1) Pick a single context + learner set (30 minutes)

• Decide: one student, one small group, or one class.
• Write it down: grade/age, current phonics program, daily minutes of phonics.

2) Define the “stop phonics” change you’ll test (15 minutes)

Make it operational, e.g.:

• Continue condition: 25 min/day explicit phonics.
• Stop/Shift condition: 10 min/day phonics review + 15 min fluency + 10 min morphology/vocab.

3) Set up 3 measures you can administer fast (60–90 minutes total)

Choose one from each category (don’t overcomplicate):

• Pseudoword decoding (1 minute): DIBELS NWF or similar.
• Oral reading fluency/accuracy (1 minute + brief scoring): a controlled passage (same difficulty level each time).
• Spelling/encoding (5–7 minutes): 10-word dictated list aligned to recently taught patterns.

Create:

• A scoring sheet (paper or Google Sheet).
• Clear scoring rules (e.g., ORF = WCPM + % accuracy; spelling = % correct + note error types).

4) Collect a quick baseline this week (2 short sessions)

• Session A (Day 1 or 2): administer all 3 measures (≤15 minutes per student).
• Session B (Day 4 or 5): repeat the same measures. Goal: see if scores are stable enough to interpret changes next week.

5) Draft your candidate stop criteria (20 minutes)

Write a provisional “exit threshold” you’ll test. Example:

• NWF: ≥95% accuracy across 2 checks
• ORF: ≥97% accuracy (rate can be “developing”)
• Spelling: ≥85–90% on taught patterns
  You can adjust later—the key is to make it explicit now.

6) Plan the smallest possible experiment starting next week (30 minutes)

Pick one design:

Option A: Two-group pilot (best if you have 2+ similar students)

• Student/group 1: continue full phonics minutes.
• Student/group 2: shift to reduced phonics + fluency/morphology.
• Measure weekly for 4–6 weeks.

Option B: Single-student A→B shift (works with just one learner)

• This week = A (current phonics dosage baseline).
• Next week = B (reduced phonics + replacement instruction).
• Keep weekly probes to watch for drop/maintenance.

7) Start a “decision log” (10 minutes)

Each day next week, jot:

• Minutes of phonics delivered
• What replaced it (fluency, morphology, etc.)
• Any notable errors (e.g., vowel confusions, guessing, slow blending)

This matters because “stop phonics” only means something if you track what instruction actually happened.

8) Do one targeted literature pass (45 minutes)

Take 2–3 of the sources you already surfaced and extract just this:

• What measures they recommend (pseudowords? spelling? ORF?)
• Any suggested benchmarks/cut scores
• Any warnings about stopping too early

Put it in a 1-page note so you can refine your threshold.

If you tell me whether you’re doing this with K–1, 2–3, or older intervention, I can help you pick measures/passages that match that level and suggest realistic benchmark targets for your draft “stop criteria.”

Answer to Question 1 (dependence on which kind of phonics is being stopped / replaced)

You’re right: my earlier plan implicitly treated “explicit phonics” as a single, uniform dose, and “replacement” as categorically non-phonics. If the optimal stop point depends on what kind of phonics you stop—and whether the replacement is actually another form of code instruction—then the methodology has to shift from a simple dose reduction study to a component-and-match study.

How I’d change the investigation

1) Build an “instruction taxonomy” and measure fidelity Instead of logging only minutes/day, I’d code instruction along dimensions such as:

• Type: synthetic (grapheme→phoneme blending), analytic (pattern discovery), analogy/onset-rime, word sorts, etc.
• Unit of instruction: phoneme-grapheme, rime units, syllable types, morphology (prefix/suffix/root), etymology/orthographic conventions.
• Teaching moves: explicit modeling, guided practice, error correction style, cumulative review, mixed practice.
• Content match: whether instruction targets the learner’s current error patterns (e.g., vowel confusions vs multisyllable segmentation).

Data needed: structured observation rubrics + audio/video samples for a subset to quantify quality and consistency (inter-rater reliability).

2) Redefine “stopping” as switching modes, not just reallocating time Rather than STOP vs CONTINUE, the decision becomes: WHEN to transition from Mode A → Mode B (and which B). Example modes:

• A: basic GPC blending + controlled decodables
• B1: advanced code (vowel teams, multisyllable routines)
• B2: morphology-focused decoding/spelling (still code-based, but different grain size)
• B3: fluency/comprehension emphasis with minimal code review

So the question becomes: Which transition, at which mastery point, yields the best outcomes for which learners?

3) Use a multi-arm design (or a SMART) rather than a single threshold RCT At the “ready” moment, randomize among different transitions, not just stop vs continue. For example:

• Arm 1: continue same phonics approach (synthetic)
• Arm 2: switch to word sorts / variability training emphasizing flexibility
• Arm 3: switch to morphology + syllable division routines
• Arm 4: reduce code instruction and increase fluency/comprehension

A SMART design (Sequential Multiple Assignment Randomized Trial) fits well because kids who don’t respond after switching can be re-randomized to a different approach.

4) Treat “match” as a hypothesis: learner profile × instruction type I’d pre-specify moderators to test interaction effects, e.g.:

• baseline phonemic awareness
• error profile (vowel errors, guessing, slow blending, multisyllable breakdown)
• working memory/attention indicators (if available)
• dyslexia risk markers
• language background

Outcome comparisons would then answer the more realistic question: who should stop which kind of phonics, and what should replace it?

5) Expand outcomes beyond “reading got better” to include mechanism-level changes If the replacement is “morphology,” I’d track:

• multisyllabic decoding accuracy/latency
• morphological spelling/meaning transfer
• decoding of derived forms (e.g., heal→health, sign→signal)

That prevents falsely concluding “phonics can stop” when the student is actually benefiting from code instruction at a different grain size.

Answer to Question 2 (measuring the “self-teaching mechanism” / orthographic mapping efficiency)

Share’s Self-Teaching Hypothesis implies that exit criteria shouldn’t just reflect current decoding accuracy, but whether the child can use decoding to teach themselves new words through independent encounters. So I’d add direct measures of orthographic learning rate and retention from context.

How I’d modify the investigation

1) Add a “novel word self-teaching” task (experimental orthographic learning paradigm) Design:

• Create short passages/stories containing novel words (pseudowords or rare real words) that are decodable but not previously taught.
• Ensure the meaning is inferable from context (so it mimics real reading).
• Control exposures: each target appears, say, 2, 4, 6, 8 times across texts.

Measures (immediate + delayed, e.g., 24 hours / 1 week):

• Reading accuracy + speed for those novel words in isolation
• Spelling of the novel words (strong test of orthographic mapping)
• Orthographic choice (pick correct spelling among foils: dain/dane/dein)
• Recognition/lexical decision (is this the word you saw?)
• Optional: meaning recall (to ensure it’s not rote visual memory only)

Key outcome: exposures-to-criterion (how many encounters needed to become accurate/automatic and retain).

2) Measure transfer to new complexity, not just mastered patterns Include targets that require:

• multi-syllable decoding routines
• variable vowel pronunciations / “flexibility” demands
• morphologically complex forms (e.g., a novel root + familiar affix)

This tests whether the self-teaching mechanism works when the spelling system is less one-to-one.

3) Track “independent orthographic learning” in a semi-naturalistic way For a week or two:

• Give the child short independent reading with embedded novel targets.
• Do no pre-teaching of those targets.
• Then probe whether the words were learned (reading/spelling/recognition).

This is closer to the real exit question: Will the child keep building word knowledge without explicit code lessons?

4) Integrate these measures into the stop/transition experiment Instead of only asking, “Did decoding scores stay high after stopping?”, I’d ask:

• Does the STOP/SHIFT condition preserve or improve orthographic learning rate compared to CONTINUE?
• Do some transitions (e.g., morphology-focused) increase self-teaching efficiency for complex words?

5) Use self-teaching efficiency as part of the exit rule An evidence-based exit criterion might become something like:

• adequate decoding accuracy/fluency plus
• adequate orthographic learning rate (e.g., learns and retains novel decodable words after ≤X exposures)

That directly operationalizes Share’s mechanism.

What this does to the original question

These changes would likely shift the “answer” away from a single stop point and toward a profiled transition rule:

• Stop/reduce basic phoneme-level phonics when decoding is accurate/automatic and the learner shows efficient self-teaching on novel words;
• but continue/switch to a different code manifestation (multisyllable routines, morphology, flexibility training) if self-teaching is weak or breaks at higher complexity.

If you tell me the age/grade band (K–1 vs 2–3 vs older intervention), I can outline a concrete novel-word task that’s developmentally appropriate and doesn’t take more than ~10–15 minutes to administer.

1. Instruction-type dependence: How much does the “right time to stop phonics” vary depending on which phonics approach is used (synthetic vs analytic vs word-sorts/flexibility training), and can we build different exit criteria for each?

2. Transition vs stop: Is the real decision not “stop phonics,” but “when to transition from phoneme-level decoding to other grain sizes” (syllable-level and morphology/etymology), and what indicators best signal readiness for that shift?

3. Self-teaching as an exit criterion: Can we measure a child’s orthographic learning rate (how many exposures they need to map/retain new words) and use that as a better “done with phonics” marker than pseudoword/ORF benchmarks alone?

4. Mismatch hypothesis: Are apparent “diminishing returns” from continued phonics actually caused by a mismatch between instruction content and the child’s current error profile—meaning the child still needs code instruction, just different code instruction?

5. Replacement matters: When phonics time is reduced, which replacement (fluency practice, morphology, comprehension, writing) produces the best outcomes for which learners, and does that differ by decoding profile?

6. Flexibility vs rigidity: How can we quantify “decoder flexibility” (handling alternative pronunciations and exceptions) and test whether certain instructional moves (e.g., contrasting word sets/oddballs) prevent rigid over-application of rules?

7. Quality/fidelity effects: What minimum level of instructional fidelity/quality must be met before any “stop rule” is meaningful—i.e., how do we avoid concluding a child is “done” when the issue was inconsistent or low-quality phonics delivery?

8. Long-term maintenance: After reducing explicit phonics, what is the best low-dose maintenance schedule (frequency and content) to prevent backsliding while still freeing time for higher-level literacy?