Diagnose a first attempt, repair its cause and check transfer on a new sentence.
Learning path
- Task and scoring in practice
- Technique 1: reconstruct from evidence
- Technique 2: diagnose before drilling
- Complete demonstration: a lost ending
- Technique 3: measure transfer under time
- Complete demonstration: spelling and review
Examples are teaching illustrations, not officially scored candidate responses.
Task and scoring in practice
Write from Dictation in Practice
The first dictation lesson established the response routine. This lesson shows how to use that routine in practice when part of the sentence is uncertain. We will separate hearing, retention, spelling and checking errors, then choose a repair that addresses the cause. I will demonstrate the whole process twice with different sentences. The objective is a better first response on unfamiliar audio.
What earns credit in dictation
| Criterion / example | Meaning / explanation |
|---|---|
| Correct word, correctly spelled | 1 word point |
| Incorrect or misspelled word | 0 for that word |
| Practice record | Exact words, order and error cause |
Word credit is an item rule, not a reported-score prediction.
Explanation
Pearson awards one item point for each correct word with correct spelling. An incorrect or misspelled word earns no point for that word. Keep the source words in their original order. There is no separate grammar or punctuation band in this task, and a correct word point has no published fixed conversion into a reported score out of ninety. These distinctions decide what our practice record measures.
A useful practice result keeps the first attempt
Weak evidence
Only the final corrected sentence is saved.
Useful evidence
First attempt, source difference, repair and fresh check.
Explanation
A corrected sentence proves that you can copy or repair it after feedback. It does not show what you retained on the first playback. Keep the first attempt before replaying. Then mark the source difference and the action that caused it. We will use that record to decide the next exercise. The correction is valuable, but it answers a different question from the first attempt.
Four causes can produce the same wrong word
| Possible cause | Evidence to seek |
|---|---|
| Hearing | Could the ending be identified on diagnostic replay? |
| Retention | Was it heard, then lost before entry? |
| Typing or spelling | Was the intended form entered incorrectly? |
| Checking | Was a correct first form changed? |
Explanation
Suppose the source contains students and the response says student. The visible mismatch is clear, but the cause is not. Did the listener miss the ending, lose it from memory, mistype it or remove it during checking? I investigate the actual attempt instead of assigning every ending error to grammar. The right repair depends on what happened.
Technique 1: reconstruct from evidence
Technique 1: separate certainty from inference
- Keep secure phrases in their source order.
- Test one uncertain part against sound and context.
- Submit one coherent reconstruction.
Explanation
This reconstruction technique protects the wording you actually heard. First retain the secure phrases in order. Then identify a local uncertainty and test the most plausible reconstruction against sound, grammar and meaning. Grammar helps reject an impossible form, but it cannot recover an unheard noun by itself. I submit one response rather than append alternative words as a list.
A grammatical option can still be the wrong word
Overclaim
The ___ submitted the report.
Grammar proves the missing word is lecturers.
Evidence-based check
Both lecturers and students fit.
The recording determines the source word.
Explanation
The sentence frame allows both lecturers and students. Knowing that the blank is part of the subject does not identify the spoken noun. If I remember students from the audio, I keep it. If I do not, I cannot claim that grammar proves it. This is the limit of the technique: it checks a remembered candidate and does not create source evidence.
The word after an article is not always a noun
| Criterion / example | Meaning / explanation |
|---|---|
| the schedule | Determiner + noun |
| the revised schedule | Determiner + adjective + noun |
| the very detailed schedule | Determiner + adverb + adjective + noun |
Use the phrase structure, then verify the remembered wording.
Explanation
The old shortcut, article followed by noun, is too crude. In the revised schedule, revised modifies schedule. In the very detailed schedule, an adverb and adjective intervene. Look for the complete noun phrase and its head. This improves your ability to test a recovered phrase without inserting a word simply because it follows the.
A time phrase helps, but the sound still matters
| Evidence | Wording |
|---|---|
| Possible sentence A | They reviewed the report yesterday. |
| Possible sentence B | They submitted the report yesterday. |
Both sentences are grammatical. The source determines which wording to reproduce.
Explanation
Yesterday supports a past interpretation. It does not tell me whether the speaker said reviewed or submitted. In this written comparison, the whole verb differs. I use yesterday to reject an incompatible reconstruction, then return to the sound I retained to select the actual verb. Treating the time marker as an answer generator would overstate what grammar can do.
An uncertain ending deserves a local check
Retained phrase
The students are ready.
Typed response
The student are ready.
Source-based repair
The students are ready.
A grammar check locates the mismatch. The source supplies the repair.
Explanation
My note says students, and the retained phrase ends with are ready. When typing, I accidentally enter student. The agreement mismatch tells me where to inspect the response. I recover students because that is also the form retained from the recording. Changing are to is would produce a grammatical sentence but alter another source word. The technique uses grammar to locate the problem while keeping the recording in control.
Locate an addition in the response history
Saved note
Applications close Friday.
First typed entry
Applications will close on Friday.
Replay check
Applications close Friday.
The saved note already preserved the wording. Inspect what changed during entry.
Explanation
This practice record gives us more than a final wrong sentence. The saved note contains Applications close Friday. The first typed entry adds will and on. After the attempt, replay confirms the shorter wording. Because the note already preserved the source, the addition appears during entry or editing in this record. I practise transferring the saved phrase without expanding it into a preferred style, then inspect any words added during the final check.
A fresh recording is needed to test whether that change transfers. The note narrows the cause here, but it cannot prove how every similar error began. The added words are not additional correct source words. Pearson does not publish a universal fixed reported-score penalty for each added word.
Technique 2: diagnose before drilling
Technique 2: use replay to identify the cause
- Save the first response.
- Replay to investigate the specific mismatch.
- Check a fresh sentence under one-play conditions.
Explanation
After saving the one-play attempt, replay becomes a diagnostic tool. First hear the whole sentence again without the transcript. Then inspect the source and, if necessary, a short phrase. Finally test a fresh sentence once. Repeated listening to the same sentence can explain the error, but the fresh response is what shows whether the repair transferred.
The same ending error can need different practice
| Observed evidence | First practice choice |
|---|---|
| Ending unclear even on replay | Short phrase listening and source comparison |
| Ending heard but lost during notes | Retain a complete noun phrase |
| Correct in notes, wrong in typed response | Keyboard entry and final word-form check |
Explanation
These are three possible diagnostic records, not claims about every learner. In the first record, the ending remains unclear until the transcript is shown. I practise recognising that ending in phrases. In the second record, the learner identifies it on replay and reports having heard it originally, but loses it while writing notes. I reduce note load and keep the whole phrase together.
In the third record, the correct ending is already in the notes but disappears in the typed response. More listening would not directly repair that observed change. I practise transferring the complete word into the answer and checking its ending before submission. The same visible error now leads to three different drills, each aimed at preserving a correctly spelled source word.
Inspect the spelling before retrieval
First entry
availible
Correct spelling
available
Explanation
Available is familiar in speech, but this first entry has the letter I in the sixth position. The correct spelling has the letter A there. I inspect the whole word and locate that change. This is the observation stage. On the next page both spellings disappear, so the following attempt will depend on memory rather than the model staying in view.
Retrieval with the model off screen
Listen to the word. The teacher writes from memory before the check.
Explanation
Both spellings have now left the screen. I will retrieve the word from this spoken cue.
Source wording
Available.
I will write it from memory before bringing the model back.
The entry is complete. Now I can compare it with the model.
Check the retrieved spelling
Retrieved entry
available
Reference spelling
available
Next use
Seats are available tomorrow.
Compare the whole word. Later, check it again in a fresh sentence.
Explanation
Here is the entry from the model-hidden interval: available. I now uncover the reference and compare the complete word, including the letter A in the sixth position. This entry matches. That demonstrates the retrieval-and-check routine, but one immediate success does not establish lasting recall. Next I use the word in a different sentence: Seats are available tomorrow. In a later practice session, a fresh sentence without the spelling preview can check whether the repair holds.
The point is to keep observation, retrieval and checking separate.
A variant reveals whether the phrase was heard
| Evidence | Wording |
|---|---|
| Variant A | The teacher explains the homework to the students. |
| Variant B | The teacher explained the homework to the class. |
The changing words are the lesson. A familiar version is not an answer key.
Explanation
These two written variants differ in tense and the final noun phrase. A memorised version can be fluent and still mismatch a new recording. Before audio practice, I avoid previewing the expected text. In review, I compare the response with the actual recording for that attempt. Sentence lists are useful language material, but familiarity cannot verify the wording that played.
Contractions require more than a final t rule
Unsafe shortcut
No clear final t means the speaker said can.
Stronger listening routine
Use the whole phrase, stress and context. Verify against the actual recording.
Explanation
Can and cannot change the meaning, but there is no safe rule that a clearly released final t must always be audible in cannot or can’t. Accent, stress and connected speech affect the signal. In practice, compare complete phrases and review the actual recording. If a contrast remains unclear, mark it as unresolved and seek a better audio model instead of turning one sound cue into an absolute rule.
Choose the material for the skill
| Material | Useful for | Evidence it cannot supply |
|---|---|---|
| Text-only sentence list | Spelling and phrase recall | First-play hearing |
| Written variant pair | Noticing changed wording | Recognition in audio |
| New continuous audio | Hearing, retention and entry | Cause without follow-up review |
Explanation
Reading a sentence and then covering it can train recall and spelling. It cannot test whether you identified words in speech. A variant comparison trains attention to changes, but it also does not replace a one-play dictation attempt. I choose the material according to the diagnosed cause, then finish with audio when the target is listening-to-writing transfer.
Complete demonstration: a lost ending
Demonstration 1: a fresh recording
Teacher demonstration. The source appears after reconstruction.
Explanation
I will play one new sentence and keep the words hidden. Watch the reconstruction on the next page.
Source wording
The updated schedules are available near the reception desk.
I have retained the main phrases. Next I will show the notes and the first entry before checking the source.
Notes become a first response
Notes
updated schedules / are avail / near reception desk
First response
The updated schedule are available near the reception desk.
Local uncertainty
schedule or schedules?
The source has not been revealed yet. Inspect the note-to-entry change.
Explanation
My notes retain updated schedules, are available and near the reception desk. I type the sentence, but the first entry loses the plural ending from schedules. Notice that the note already contains the correct form. In this teaching case, the record points toward an entry or checking error, so more listening alone would not be the most direct response.
The source identifies one changed word
| Evidence | Wording |
|---|---|
| First response | The updated schedule are available near the reception desk. |
| Source | The updated schedules are available near the reception desk. |
One word form changed. The first response remains in the review record.
Explanation
Now the original appears. The response has schedule where the source has schedules. The other words are unchanged. This illustration contains nine source words, so eight words match and one word is wrong. That is a word-level practice comparison, not an eight-out-of-nine prediction for a reported skill score. The diagnostic detail is that the plural survived in the note and disappeared during entry.
The repair follows the note-to-entry error
| Criterion / example | Meaning / explanation |
|---|---|
| Evidence | The note had schedules; the entry had schedule. |
| Repair | Keep the whole plural phrase during entry. |
| Fresh check | A new plural phrase in a different sentence. |
Explanation
I retain the complete plural phrase, type it once and compare the noun with the following are before submitting. The agreement check points me toward the form I already heard. It does not license changing are to is. For the next attempt I choose a new noun and a different topic. That fresh example will check whether I preserve the ending during entry.
A different sentence tests the repair
Teacher demonstration. The source appears after reconstruction.
Explanation
I will play one new sentence and keep the words hidden. Watch the reconstruction on the next page.
Source wording
The revised prices are displayed beside the entrance.
This time I will carry revised prices as a single phrase into the typed response.
The fresh response preserves the full phrase
| Criterion / example | Meaning / explanation |
|---|---|
| The revised prices | Plural noun phrase retained |
| are displayed | Complete verb phrase |
| beside the entrance. | Location and article retained |
Original: The revised prices are displayed beside the entrance.
Explanation
I reconstruct the sentence before showing the original. First I keep The revised prices together, including the plural ending. Then I enter are displayed as a complete verb phrase. Finally I add beside the entrance, preserving the article in the location phrase. Now compare that reconstruction with the original: The revised prices are displayed beside the entrance. All eight source words match in this teacher demonstration.
Keeping the whole noun phrase protected the ending of prices; the earlier response lost the ending of schedules. That is useful evidence for this particular transfer case. It is still only one case, so a study decision should use repeated unfamiliar attempts rather than declare the problem solved from one success.
Technique 3: measure transfer under time
Technique 3: measure the first response on new material
- Use unfamiliar sentences with one initial playback.
- Record word accuracy and complete-sentence accuracy separately.
- Track the recurring cause alongside the result.
Explanation
A useful measure keeps the conditions comparable: first playback, unfamiliar wording and the same response routine. Count correct words against the source, record complete-sentence accuracy separately, and keep a short cause note. Neither practice percentage has a guaranteed conversion to a reported PTE score. The benefit is that the record can show whether the specific repair survives new wording.
Two progress measures can move differently
| Illustrative set | Correct words | Whole sentences correct |
|---|---|---|
| Earlier: 10 sentences | 83 / 100 | 2 / 10 |
| Fresh: 10 sentences | 93 / 100 | 5 / 10 |
Practice evidence only. These counts do not predict an official band.
Explanation
A learner can improve word accuracy while still missing one word in many sentences. That is why I keep both measures. These fictional sets contain the same number of sentences and source words. The second set has fewer wrong words and more completely correct sentences. I still inspect which words changed and whether the sets were reasonably comparable before calling the routine more reliable.
A shared Listening clock needs a rehearsal record
| Rehearsal evidence | Useful interpretation |
|---|---|
| Dictation reached with little time left | Inspect earlier decision delays |
| Time left but words still missed | Inspect hearing and reconstruction |
| Responses unfinished after long checking | Practise a focused final check |
Explanation
Dictation comes late in Listening, so earlier decision time can affect the time left to type and check. I record the time remaining when dictation begins during a full rehearsal. A fixed reserve such as five minutes is not a universal official requirement. The useful question is whether my actual routine completes the remaining items without unnecessary delay or unfinished responses.
A local time target is something to test
Rigid rule
Always abandon an earlier item at the same second.
Evidence-led adjustment
Shorten unproductive delay and compare accuracy plus completion.
Explanation
Suppose one rehearsal leaves two dictation responses unfinished after long multiple-choice deliberation. I reduce an unproductive reread in the next full rehearsal and compare both the choice accuracy and dictation completion. I do not sacrifice every earlier question automatically. The method weighs the observed gain and cost across the section.
Complete demonstration: spelling and review
Demonstration 2: the wording is heard clearly
Teacher demonstration. The source appears after reconstruction.
Explanation
I will play one new sentence and keep the words hidden. Watch the reconstruction on the next page.
Source wording
Accommodation is available within walking distance of the station.
I have kept the complete wording. I will type it and then inspect one spelling uncertainty.
Reconstruction comes before the spelling check
Retained note cues
accom / is available / within walking distance / of the station
First spelling
accomodation
Checked response
Accommodation is available within walking distance of the station.
Keep the source word. Repair its spelling locally.
Explanation
The retained phrase begins with accommodation is available, then within walking distance and of the station. My abbreviated note cue avoids committing to the uncertain spelling yet. I enter the whole sequence first. The first spelling on screen has only one letter M. I inspect that word without replacing the source phrase with a simpler synonym. Now I check the spelling model: accommodation has two of the letter C and two of the letter M.
The complete response now keeps the heard word and its correct spelling. That protects its word credit; the misspelled version earns no credit for that word. Replacing it with housing would be fluent English but would alter the dictation word.
The review chooses the next fresh check
| Observation | Action | Transfer evidence |
|---|---|---|
| Correct heard word, wrong spelling | Hide model and retrieve accommodation | New sentence, model hidden |
| Full phrase retained | Keep the phrase routine | No added or omitted words |
Explanation
In this case the source was heard and retained correctly, while the written entry lost one letter M. I choose spelling retrieval for accommodation, then use a new sentence containing it. I do not replay the same audio until it merely feels familiar. The result that matters is whether the word can be entered correctly in a new response without seeing the model.
A practice session ends with a decision
- First attempt and diagnosis
- A repair matched to the cause
- A fresh one-play check
Explanation
The practice loop now has a purpose. Save the first response, identify the source difference and investigate its cause. Choose a repair that matches the evidence, then test an unfamiliar sentence under one-play conditions. The first dictation lesson supplied the response routine. This lesson turns the routine into targeted practice without mistaking memorised corrections for transferable performance.