Exact-detail listening is not simply “hearing carefully”. It is a precision problem inside a moving stream. The learner must recognise a name, number, date, time, price, unit, code or spelling sequence, bind it to the correct role, preserve the exact form when the task requires exactness, and then release it quickly enough to keep listening.
This guide is SETC-SEC-LISTEN-140, the exact-detail spoke inside the Secondary English Listening Examination Performance lane. It does not replace general spelling instruction, the action-state architecture in Listening Comprehension Announcements and Instructions, or the generic exam-writing owner Spelling Under Exam Pressure. Its narrower job is auditory exactness under listening conditions: how precise details enter, remain attached to meaning and become scorable responses without damaging the next part of the recording.
Maren protects answer form and role labels. Iona protects provenance, units and logical status. Leonie protects the stream when one difficult name or number threatens to consume too much attention.
IDENTIFY THE DETAIL TYPE → HEAR THE FORM → BIND IT TO ROLE → PRESERVE UNIT/STATUS → RECORD MINIMALLY → VERIFY ONLY IF NEEDED → RELEASE → RE-ENTER
Part I | Exact detail is relational
1. A number without a role is not a usable answer
The digits may be remembered perfectly and still be useless. “25” could be a price, room number, participant limit, date or duration. Exact-detail listening therefore requires two pieces: form and function. The best minimal note is often VALUE + ROLE.
2. A name without a source can also be wrong
Several people may be named. Knowing the spelling of one name does not answer a question about who made the decision. Bind identity to role or action.
3. Precision demand varies by question
Some tasks require exact spelling; others only require matching a printed name. Some ask for a number and already supply the unit. Others require the unit. The learner should spend exactness where marks depend on it, not uniformly.
4. Exact detail has a high attention cost
Letter sequences, digit strings and unfamiliar names can pull attention into rehearsal. The learner needs a stopping rule so one precise item does not damage later evidence.
5. Familiar details can still be misbound
The problem is not always recognition. A clear “nine thirty” can still be entered as start time when it was arrival time. Accuracy requires role.
6. Similar values compete strongly
15 and 50, 13 and 30, 18 June and 18 July, Room 14 and Group 14 can interfere. Labels reduce competition.
7. Corrections make exactness dynamic
The speaker may spell a name, correct one letter, or revise a time. Exact form is not final until the relevant thought closes.
8. Approximation can be part of exact meaning
“About twenty”, “just before six” and “roughly thirty per cent” should not become falsely exact if options distinguish approximation.
9. Units are semantic
Thirty seconds, thirty minutes and thirty dollars share the number and differ entirely in meaning. Unit loss is a comprehension error at the proposition level.
10. Exactness should serve the question, not dominate the recording
A learner must know when to stop perfecting a detail and protect the moving signal. Precision without recovery is fragile performance.
Part II | Numbers as sound
11. Number recognition can be trained separately
Short drills can isolate teen/ty contrasts, decimal points, years and grouped digits. Restore them to meaningful audio quickly so number skill remains contextual.
12. Thirteen and thirty
Stress pattern can help, but context and unit should also support recognition. Do not rely on one acoustic cue universally.
13. Fourteen and forty
Minimal contrasts are useful when a learner repeatedly confuses them. Test transfer with new speakers.
14. Fifteen and fifty
Practise at natural pace, not only exaggerated classroom pronunciation.
15. Sixteen and sixty
Record whether the error is hearing, memory or typing. Similar outputs can hide different causes.
16. Seventeen and seventy
Use units and roles to strengthen interpretation.
17. Eighteen and eighty
Do not let one uncertain digit string stop the next clause. Mark and continue.
18. Nineteen and ninety
Transfer across prices, ages, percentages and years.
19. Hundreds and thousands
Chunk by magnitude and attach unit. 1,500 participants differs from $1,500.
20. Large numbers should be semantically compressed
When exact digits are not required, the learner may only need “about 1.5 million”. Follow the question demand.
Part III | Decimal, fraction and percentage detail
21. Decimal points are low-duration, high-value signals
3.5 and 35 are radically different. Train the spoken patterns used in the target variety of English.
22. Zero can be pronounced in different ways
Codes, phone numbers and decimals may use different conventions. Exposure should be broad but task-specific.
23. Fractions can appear as quantities or proportions
One half, a third and three quarters require concept as well as sound recognition.
24. Percentages need a base
Twenty per cent of what? The number must remain attached to the measured group or change.
25. Percentage-point change differs from percentage change
Advanced tasks may distinguish them if the audio does. Preserve the speaker’s wording rather than performing unasked calculations.
26. Approximate percentages
Nearly half, just over a third and around twenty per cent carry degree. Do not erase it.
27. Comparative percentages
Higher than, lower than, double and half involve relation, not only value.
28. Percentages inside graphs
The visual may already provide units. Focus on correct series and trend.
29. Decimal prices
Currency notation can differ across systems. Follow the actual task’s required response format.
30. Scientific decimals
If used at secondary level, context and printed units can reduce acoustic burden.
Part IV | Dates
31. Date role comes first
Application date, deadline, event date and payment date can coexist. Label them.
32. Day and month can reverse across conventions
When written format varies internationally, use the examination’s official convention and printed context.
33. Ordinal endings are often acoustically light
First, second, third and later ordinals can matter in dates and ranking. Train recognition in phrases.
34. Month names can be reduced in connected speech
Practise natural sentences rather than isolated recitation only.
35. Similar month sounds need context
June and July differ clearly in careful speech but can still be misremembered under density. Role labels plus written month abbreviation can help.
36. Year recognition has multiple spoken forms
Different speakers may say the same year differently. Build flexible recognition.
37. Historical versus current date
A past date can be background; the question may ask the new deadline. Status matters.
38. Original versus revised date
Cross out obsolete values. Do not keep both as equal candidates.
39. Relative date language
The following Monday, two days earlier and the week after require relation before conversion.
40. Date ranges
From 12 to 18 May is an interval, not two unrelated dates.
Part V | Times
41. Arrival, start and finish need separate labels
These are among the most common role-binding errors in practical audio.
42. Twelve-hour and twenty-four-hour forms
Recognise the forms relevant to the target assessment. Do not assume one global convention.
43. Quarter past and quarter to
Relational clock expressions should be automatic enough not to consume calculation time.
44. Half past
Map phrase to time quickly, then release.
45. “Ten to” and “ten past”
Direction around the hour matters.
46. “By six” versus “at six”
Deadline and exact time differ.
47. “From six” versus “until six”
Start and endpoint differ.
48. Approximate time
Around six, just after six and before six preserve uncertainty or relation.
49. Corrected time
Initial and final values must be time-indexed.
50. Time plus duration
Do not confuse start time with length.
Part VI | Prices and money
51. Price needs category
Adult, student, member, early-booking and standard prices can all be true.
52. Deposit versus total
Label payment stage.
53. Refund versus fee
The amount paid and the amount returned can differ.
54. Discount percentage versus final price
Know which the question asks. Do not calculate unless required.
55. Currency matters
Printed symbols can externalise units; spoken currency may still distinguish options.
56. Per-person versus total price
Scope belongs to the amount.
57. Price range
From 20 to 30 is not a single exact value.
58. Approximate cost
About 50 should not become exactly 50 if precision matters.
59. Changed price
Temporary offer or corrected amount creates final-state demand.
60. Money details should not trigger unnecessary arithmetic
Listen for what is explicitly asked before calculating anything.
Part VII | Quantities and units
61. Count versus measurement
Three people, three kilograms and three metres require different conceptual categories.
62. Singular and plural can matter
One metre versus one metres is a language-form issue if written response is assessed.
63. Metric and imperial units
Use the units given by the source. Do not convert unless required.
64. Minutes versus hours
Duration unit must remain attached.
65. Metres versus kilometres
Magnitude can expose unit loss.
66. Grams versus kilograms
Same numerical form can represent very different quantity.
67. Litres versus millilitres
Capture unit where the answer field does not already supply it.
68. Temperature
Degree value may need scale if the task distinguishes it.
69. Capacity
Physical capacity can differ from allowed participant limit. Role and unit interact.
70. Quantity boundaries
At least, no more than, fewer than and more than must survive.
Part VIII | Names
71. Name recognition can be enough without spelling
If options are printed, identify the correct person and move on.
72. Exact spelling is a different task
When the answer requires a name typed or written, orthographic detail becomes part of the response chain.
73. Unfamiliar names should not be treated as language failure
A learner may not know the cultural origin of a name. The task is acoustic capture and role binding.
74. Role labels can support unfamiliar names
“speaker A / organiser” can preserve function while spelling remains uncertain.
75. Similar names need discriminators
If two printed names differ by one sound or letter, preview the contrast.
76. Names can be corrected
One letter or syllable may be restated. Update locally.
77. First name versus surname
Know which field is required.
78. Titles and roles
Dr, Professor, Ms or job role may help identify the correct source but may not belong in the answer.
79. Names inside reported speech
Do not confuse who is speaking with who is being mentioned.
80. Name precision should not destroy the next sentence
If exact capture is uncertain, mark cheaply and re-enter.
Part IX | Spelling from audio
81. Spelling is sequential memory
A letter string must be heard, held and output in order. Chunking reduces load.
82. Letter names can be acoustically similar
Minimal contrasts such as M/N or B/P can require context or repetition where the format provides it.
83. Capitalisation may or may not matter
Follow official marking rules. Do not assume from ordinary writing conventions.
84. Double letters
“Double l” must become two letters, not the word “double”.
85. Hyphenated forms
Exact punctuation matters only where the task’s rules require it.
86. Spelled words can be corrected mid-sequence
Keep the answer editable until the sequence closes.
87. Spelling by chunks
Group long strings into two- or three-letter chunks rather than rehearsing every letter independently.
88. Written vocabulary can support auditory spelling
If the learner recognises the word, orthographic knowledge reduces sequence burden.
89. Unknown words require pure capture
When lexical familiarity is absent, rely on letter sequence and any repeated confirmation.
90. General spelling study and listening spelling are not identical
The former builds orthographic knowledge; the latter adds real-time auditory capture and attention control.
Part X | Codes and identifiers
91. Codes mix letters and numbers
Separate character type as well as order.
92. Chunk code strings
AB-417 is easier to preserve as two units than five unrelated characters.
93. Repeated digits
Double or triple forms can be spoken compactly. Recognise target conventions.
94. Reference code versus phone number
Role determines formatting expectations.
95. Leading zero
Do not drop it if it is part of the identifier.
96. Mixed case
Where case matters, the task should provide a way to know it. Follow official design.
97. Code correction
Update only changed characters if possible.
98. Code is often low-semantic information
Because it carries little meaning, rehearsal cost is high. Write it quickly where permitted.
99. Do not let code capture erase subsequent instructions
One uncertain character should remain local.
100. Verify code only in safe time
Do not repeatedly reread while audio continues.
Part XI | Addresses and contact details
101. Address fields combine several detail types
Number, street name, unit, city or postal code can appear. Use field structure.
102. Street number versus unit number
Label role if both are spoken.
103. Street names can require spelling
Printed options reduce this demand; open fields increase it.
104. Postal codes are identifiers
Chunk and preserve leading zeros where relevant.
105. Phone numbers benefit from grouping
Use natural spoken groups.
106. Country or area code can be optional or required
Follow the field and question.
107. Email addresses contain symbol language
At, dot, hyphen and underscore require exact mapping where tested.
108. Domain names can be familiar or unfamiliar
Do not auto-correct an unfamiliar domain into a familiar one.
109. Contact detail correction
One digit or letter can be revised after the full string.
110. Contact-detail tasks need a recovery branch
Capture enough, mark uncertainty, protect the next field.
Part XII | Exact detail in conversations
111. Speaker source still matters
A price quoted by Speaker A may be rejected by Speaker B. Exact value plus source/status is required.
112. Names can identify speakers
Use introduction cues but do not over-invest spelling if the question only asks opinion.
113. Schedule negotiation creates multiple exact values
Track proposals and final agreement.
114. Correct value, wrong speaker
This remains a source-binding error, not number recognition.
115. Reported number
A speaker can quote another person’s estimate. Ownership matters.
116. Approximate number can become exact later
Update confidence and form.
117. Exact name can be irrelevant
Do not chase spelling if the question asks reason.
118. Speaker correction can be collaborative
One person supplies a corrected time that the other accepts. Final shared state governs.
119. Detail questions still require listening to relation
Numbers and names do not remove the need for discourse understanding.
120. Conversational exactness is dynamic
Values can be proposed, challenged and confirmed across turns.
Part XIII | Exact detail in monologues
121. Long audio increases delay
A date introduced early may be tested later. Selective notes become more useful.
122. Numbers can support claims rather than answer directly
Know whether the question asks data or conclusion.
123. Several statistics can interfere
Bind each to category or year.
124. Historical dates need timeline structure
Event order and mention order can differ.
125. Names can represent sources
A researcher name may matter only for attribution.
126. Definitions can include measurements
Unit and threshold may be central.
127. Approximation language often qualifies evidence
About, roughly and nearly should survive.
128. Repeated number can shift role
The same value might be cited, compared or corrected. Track function.
129. Note economy matters more over duration
Exact details should sit under section labels.
130. Global questions should not be hijacked by memorable numbers
Precision is only valuable when relevant.
Part XIV | Exact detail with visuals
131. Printed units reduce writing
Use them as external memory.
132. Table rows bind values
Write directly in the correct cell.
133. Forms define role
Field labels solve much of the binding problem if read before audio.
134. Maps can contain numbered locations
Do not confuse label numbers with spoken times or route steps.
135. Diagrams can include measurements
Attach values to components.
136. Graph axes provide units
Do not duplicate unless response requires it.
137. Visual exactness can hide auditory error
A learner may copy a nearby printed value instead of the spoken one. Source discipline matters.
138. Visual salience can distort number selection
Bold or central values are not automatically answer-bearing.
139. Cross-modal check should be brief
Confirm field and value, then return to the stream.
140. Digital entry adds another precision layer
The computer-based owner handles typing and interface execution.
Part XV | Failure modes
141. Naked-number failure
Correct digits, missing role.
142. Wrong-unit failure
Correct magnitude, wrong measurement.
143. Wrong-speaker failure
Correct detail, wrong source.
144. Old-state failure
Initial value chosen after correction.
145. Approximation loss
Rough value turned into exact claim.
146. Boundary loss
At least, under, before or by removed.
147. Digit-transposition failure
Correct digits, wrong order.
148. Teen/ty failure
Acoustic number contrast misheard.
149. Spelling-sequence failure
Letter heard but order lost.
150. One-letter correction failure
Old spelling remains.
151. Over-rehearsal failure
Exact detail held so long that next information disappears.
152. Under-recording failure
Detail understood but not preserved over delay.
153. Field-alignment failure
Correct detail entered in wrong box.
154. Typing failure
Correct auditory representation becomes wrong digital form.
155. World-knowledge correction failure
Learner “fixes” an unfamiliar name or value into something more familiar.
156. Exactness-over-scope failure
Time is spent perfecting a detail the question did not require.
Part XVI | Diagnostics
157. Number-only recognition test
Isolate teen/ty and digit sequences.
158. Number-plus-role test
Restore labels and meaning.
159. Unit-binding test
Use same numbers with different units.
160. Date-role test
Deadline, event and payment dates.
161. Time-role test
Arrival, start and finish.
162. Price-role test
Deposit, total, discount and refund.
163. Name-recognition test
Printed options remove spelling burden.
164. Name-spelling test
Open field adds orthographic sequence.
165. Code-chunking test
Compare ungrouped and grouped recall.
166. Correction test
Initial form changes one element.
167. Delayed-detail test
Increase time between hearing and response.
168. Note-benefit test
Compare no notes and minimal labelled notes.
169. Recovery test
Place difficult detail before easy answer.
170. Speaker-transfer test
Use new voices with stable content.
171. Rate-transfer test
Increase natural pace after accuracy is stable.
172. Digital-transfer test
Compare written and typed capture.
Part XVII | Practice progression
173. Stage 1: isolated numbers
Build basic auditory contrasts.
174. Stage 2: numbers in simple sentences
Add role labels.
175. Stage 3: two competing values
Train discrimination.
176. Stage 4: one correction
Track final state.
177. Stage 5: units and boundaries
Add semantic precision.
178. Stage 6: names and spelling
Introduce sequence capture.
179. Stage 7: codes
Mix letters and digits.
180. Stage 8: delayed recall
Increase memory distance.
181. Stage 9: multi-detail announcements
Dates, times and prices together.
182. Stage 10: multi-speaker detail
Add source binding.
183. Stage 11: unfamiliar voices
Transfer acoustic precision.
184. Stage 12: target timing
Remove practice-only control.
185. Stage 13: typed entry
Add interface layer where relevant.
186. Stage 14: late-paper detail
Test precision under fatigue.
187. Stage 15: mixed detail types
Task recognition becomes important.
188. Stage 16: unseen integrated section
Verify robust exact-detail performance.
Part XVIII | Drills
189. Label-every-value drill
No number may be written without role or unit.
190. Teen/ty contrast drill
Use multiple speakers and natural sentences.
191. Date ladder drill
Build from single dates to competing deadline/event pairs.
192. Time ladder drill
Arrival/start/finish plus one correction.
193. Price ladder drill
Standard/discount/deposit/refund.
194. Unit-switch drill
Same number, different units.
195. Approximation drill
Exact, about, nearly, just over and under.
196. Boundary drill
At least, at most, before, after, by.
197. Name-recognition drill
Match audio to printed options first.
198. Name-spelling drill
Add open response later.
199. One-letter-correction drill
Update only changed character.
200. Code-chunk drill
Group mixed sequences.
201. Recovery-after-detail drill
Score the next question more heavily than the hard detail.
202. Delayed-recall drill
Ask after intervening material.
203. Sparse-note drill
Allow only labelled values.
204. Digital-entry drill
Type exact details without losing subsequent audio.
Part XIX | Resident casebook
205. Maren: writes every digit but no labels
Her page is full of accurate numbers that later become unusable. Role tagging repairs the system.
206. Iona: correct unit, wrong subgroup
She hears the student price but applies it to adults. Eligibility belongs to the value.
207. Leonie: one hard surname creates a cascade
She rehearses spelling while the speaker gives the reason for the event change. Marking uncertainty and releasing protects later marks.
208. Maren: over-precise answer
The audio says “about 50”; she writes exactly 50 when the option tests approximation. Precision should reflect source precision.
209. Iona: familiar-name substitution
An unfamiliar surname sounds close to a known one, and she “corrects” it from expectation. Auditory evidence outranks familiarity.
210. Leonie: teen/ty panic
One uncertain number triggers repeated mental comparison. A question mark plus re-entry keeps the issue local.
211. Maren: deposit becomes total
She records the amount but not payment stage. Role binding fixes it.
212. Iona: date scope error
She hears an extension for one group and applies it universally. Exact dates still require scope.
213. Leonie: typed digit transposition
Auditory memory is correct; keyboard execution reverses two digits. The computer layer is trained separately.
214. Maren: over-notes contact details
She rewrites printed field labels. Using the form as external memory saves attention.
215. Iona: number as evidence versus answer
A statistic supports the conclusion, but the question asks what the speaker concludes. Exact detail is not always the response.
216. Leonie: late-paper spelling errors
Recognition remains strong but output degrades under fatigue. Final-third exact-detail drills target execution.
Part XX | Transfer
217. Number to date
Role binding should survive surface change.
218. Date to price
Status tracking remains.
219. Price to quantity
Unit becomes the main discriminator.
220. Quantity to code
Semantic support decreases; chunking becomes more important.
221. Printed name to open spelling
Orthographic demand increases.
222. Familiar to unfamiliar names
Reduce lexical support.
223. One speaker to several speakers
Source binding increases.
224. Clear to new speaker
Acoustic adaptation becomes active.
225. Moderate to faster natural pace
Capture must become more efficient.
226. Immediate to delayed question
Memory support matters more.
227. Written to digital entry
Typing becomes downstream variable.
228. One detail to dense practical audio
Selection and recovery become central.
Part XXI | Stress tests
229. Number-density stress
Several values with different roles.
230. Similar-number stress
Teen/ty and near values.
231. Correction stress
Multiple revisions.
232. Unit stress
Same digits, different measurements.
233. Name-density stress
Several unfamiliar identities.
234. Spelling stress
Longer letter sequence.
235. Code stress
Mixed letters and digits.
236. Delay stress
Question comes later.
237. Rate stress
Natural pace variation.
238. Speaker stress
New voice.
239. Recovery stress
Hard detail before easy item.
240. Late-paper stress
Exact details after sustained listening.
241. Digital stress
Typed entry under live audio.
242. Full unseen stress
Mixed exact details under target conditions.
Part XXII | Measurement
243. Recognition accuracy
Was the form heard correctly?
244. Role-binding accuracy
Was it attached to the correct function?
245. Unit accuracy
Was measurement preserved?
246. Status accuracy
Was original/revised/final state preserved?
247. Scope accuracy
Was subgroup or condition retained?
248. Spelling-sequence accuracy
Were letters correct and ordered?
249. Digit-order accuracy
Were numbers transposed?
250. Note efficiency
Did minimal labels preserve detail without lag?
251. Recovery cost
How many later items were affected by one hard detail?
252. Delayed retention
Can exact detail survive intervening audio?
253. Transfer penalty
How much do new speakers or formats reduce precision?
254. Late-paper decay
Does exactness deteriorate under fatigue?
Part XXIII | Teacher protocol
255. Ask what the value represented
Do not stop at “you heard 25 correctly”.
256. Separate hearing from output
Have the learner repeat the detail before inspecting written form.
257. Use minimal contrasts
Teen/ty and one-letter differences should be trained clearly.
258. Restore natural context quickly
Exact-detail drills should not remain decontextualised.
259. Teach role labels before longer notes
Value + role often solves more than additional words.
260. Fade support
Printed role prompts should eventually disappear where the exam does not provide them.
261. Preserve unseen exact-detail material
Repeated codes and names cease to test first-contact precision.
262. Rotate speakers
Exact recognition should not depend on one voice.
263. Route generic spelling outward
Use the separate spelling owner for broader orthographic development.
264. Route practical state work outward
Use the announcements/instructions owner when the main problem is condition, exception or action state.
265. End with mixed details
The learner must recognise which type of precision the item requires.
266. Monitor recovery explicitly
Exactness is not successful if it destroys the next answer.
Part XXIV | Self-study protocol
267. Attempt audio-first
Do not reveal written detail before recognition is tested.
268. State the role
Say “arrival time”, “deposit” or “surname” before checking.
269. Replay only with a target
“Was it fifteen or fifty?” is better than replaying everything.
270. Use transcript after reconstruction
Identify whether the failure was form or role.
271. Write one repair rule
Examples: “label every value”, “preserve the unit”, “mark old value obsolete”.
272. Test a new detail
Transfer immediately.
273. Return after delay
Use different material later.
274. Add speaker variation
Build acoustic robustness.
275. Add rate gradually
Do not turn speed into random difficulty.
276. Add digital entry if relevant
Separate typing errors.
277. Track recovery
One hard detail should remain local.
278. Finish with target conditions
Exactness must survive real timing and playback rules.
Part XXV | Parent-supported exact listening
279. Use short practical details
Read a time, price and place in a meaningful sentence.
280. Ask “what was the number for?”
This reinforces role binding.
281. Use one correction
Train original versus final state.
282. Spell one unfamiliar but reasonable name
Keep it brief.
283. Ask for unit
Value plus measurement becomes habitual.
284. Avoid constant quizzing
Home language should remain communication.
Part XXVI | Exam-day operating model
285. Before audio
Identify whether the task needs name, number, date, time, price, unit, code or spelling.
286. When the detail arrives
Capture form plus role.
287. When two values compete
Wait for status or question scope to decide.
288. When a unit appears
Keep it unless the answer field already supplies it.
289. When a correction appears
Update only the changed field.
290. When exact spelling is required
Chunk the sequence.
291. When one character is uncertain
Mark cheaply and protect the next evidence.
292. When a value is approximate
Preserve approximation.
293. When the question asks relation, not value
Do not output the memorable number.
294. When notes are crowded
Return to VALUE + ROLE.
295. During review
Check digits, units, field and status before cosmetic details.
296. Before moving on
Release the exact detail mentally. The next recording needs clean attention.
Part XXVII | Technical appendix: exact-detail matrix
TYPE → FORM → ROLE → UNIT → STATUS → SCOPE → OUTPUT. TYPE identifies name/number/date/time/price/code. FORM is the exact heard representation. ROLE is what the detail refers to. UNIT is measurement or currency where needed. STATUS captures original/final/approximate. SCOPE captures subgroup or condition. OUTPUT converts the detail into the required answer form.
Part XXVIII | Technical appendix: exact-detail error ledger
Useful categories include teen/ty recognition, digit order, decimal point, date role, time role, price role, unit loss, approximation loss, boundary loss, name recognition, spelling sequence, code chunking, source confusion, correction failure, delayed retention, note overload, recovery cascade and typed transfer. Keep only recurring or expensive patterns active.
Part XXIX | Final checklist
- I know what kind of exact detail the question needs.
- I label every number by role or unit.
- I distinguish arrival, start, finish and deadline.
- I distinguish standard price, deposit, discount and refund.
- I preserve approximation and boundaries.
- I keep names attached to the correct speaker or role.
- I spell exactly only when the task requires it.
- I chunk long letter and digit strings.
- I update corrected details locally.
- I do not let one uncertain character or digit steal the next sentence.
- I can carry important exact details across a delay.
- I can transfer exact auditory detail into the required scorable form.
Closing synthesis | Precision is exact meaning, not isolated accuracy
Exact-detail listening is strongest when the learner stops treating precision as a collection of digits and letters. A correct number must still belong to the right event. A correct time must still be arrival rather than start. A correctly spelled name must still belong to the right speaker. A price must keep its currency, category and status. A code must be exact enough to score without consuming the next thirty seconds of attention.
Maren protects response form. Iona protects role, unit and source. Leonie protects continuity. Their shared rule is economical precision: capture exactly what matters, attach it to meaning, then release it quickly enough to keep the listening system alive.
Previous: SETC-SEC-LISTEN-130 | Computer-Based Listening Tests · Hub: Secondary English Listening Examination Performance · Next: SETC-SEC-LISTEN-150 | Listening Answer Transfer
Advanced Exact-Detail Casebook | Precision Without Losing the Stream
The main article separates exact form from role, unit, source, status and scope. This casebook stress-tests that distinction. Each case contains a detail that is easy to hear and easy to misuse, or difficult to hear and not worth sacrificing later evidence for. The training goal is economical precision: preserve the detail exactly enough to answer, attach it to the correct proposition, and stop processing it when the next part of the recording becomes more valuable.
Case 1 | Fifteen or fifty when the unit is dollars
The learner hears an uncertain teen/ty contrast in a price. The rest of the sentence says the student rate is much cheaper than a standard fee of sixty. Context makes fifteen plausible and fifty less plausible, but context should support rather than replace acoustic evidence. In training, replay the contrast to diagnose sound recognition, then restore the full proposition. The exam-day rule is different: if certainty remains incomplete, use the best-supported value and release the item rather than mentally repeating the sound while the next answer arrives.
Case 2 | Fifteen or fifty when both are plausible
A room can hold either number and the context provides no useful semantic constraint. Acoustic recognition carries more weight. This is the proper place for targeted number drills. The contrast should be practised across several clear speakers and natural sentence positions so the learner does not depend on one teacher’s exaggerated pronunciation.
Case 3 | A clear number attached to the wrong object
Iona hears 40 accurately. The recording has just mentioned Room 40 and then says the event allows 40 participants. Her notes contain one “40”. The answer later asks for capacity. Form recognition is perfect; binding is weak. A minimal role label—Rm40 / cap40—solves the problem more efficiently than another number lesson.
Case 4 | Two identical numbers with different roles
The event starts on the 12th in Room 12. A learner who stores “12” once may later be unable to distinguish date from location. Exact-detail memory needs semantic slots, not only strings. This case is useful because sound offers no discriminator; only role does.
Case 5 | A corrected digit inside a long code
The speaker gives AB4729, then corrects the third digit from seven to six. Rewriting the entire code can cost the next clause. The mature response is local editing: AB4629. Correction tracking should operate at field level, not restart the whole memory sequence.
Case 6 | A code correction arrives after another sentence
The speaker gives the code, explains what it is for, then says, “One correction—the final digit is eight.” Final-state monitoring must remain open even after the learner has mentally closed the string. A light visual representation makes the code editable without continuing to rehearse it.
Case 7 | A proper name resembles a familiar word
An unfamiliar surname sounds like an ordinary English word. Maren “corrects” the name into the familiar spelling. The source did not provide that evidence. Name capture is one area where lexical familiarity can mislead. If the name is spelled aloud, the letters control. If not, use the required response format and avoid inventing certainty.
Case 8 | Two printed names differ by one sound
The options are Marston and Marsden. Because spelling is visible, the task is auditory discrimination rather than orthographic generation. Preview the difference, listen for the decisive consonant corridor, select, and move on. Do not waste attention spelling the whole name mentally.
Case 9 | Exact spelling is required but the word is familiar
The learner hears “environment” and knows the meaning but misspells it under time. Listening comprehension is not the first weak link. Generic spelling development can repair the orthographic form; this listening page coordinates the timing—capture the known word, enter the best form, and avoid prolonged editing during live audio.
Case 10 | Exact spelling is required and the word is unfamiliar
The speaker spells the item letter by letter. The task becomes sequential auditory memory. Chunking two or three letters at a time can reduce load. If one character remains uncertain, mark the position rather than replaying the entire sequence mentally.
Case 11 | Double letter is lost
The recording says “double r”. The learner writes one r because she stores letter identity but not multiplicity. Train “double” as an operator on the following letter rather than an extra spoken word to remember.
Case 12 | “O” and zero collide in a code
Alphanumeric identifiers can contain visually similar characters. The question or platform may constrain what forms are possible. Use character type, spoken convention and printed pattern if provided. Never substitute a likely character solely from appearance.
Case 13 | Leading zero disappears
A phone or reference code begins with zero. Because numerical quantity is not the point, the zero is part of identity and cannot be dropped. Identifier listening differs from arithmetic number listening.
Case 14 | Decimal point disappears in note-taking
3.5 is written as 35. The learner heard “point” but her shorthand omitted it. The repair is a stable symbol and an automatic rule: decimal marker is part of the number, not decorative language.
Case 15 | Percentage is remembered without its base
The speaker says 30 per cent of younger participants and 45 per cent overall. Notes contain 30 and 45. The later question asks about the younger group. Percentages need group labels just as prices need categories.
Case 16 | “Nearly half” becomes 50 per cent
A multiple-choice option says exactly half. The learner treats approximate language as permission to round. That can strengthen the claim beyond the source. Preserve the speaker’s precision level. Exact-detail skill includes knowing when not to become more exact.
Case 17 | “Just over a third” loses direction
The learner remembers one third but not whether the value is above or below it. Approximation words carry directional information. They belong to the proposition.
Case 18 | A date is correct but belongs to the old schedule
The event was planned for 14 May and moved to 21 May. Both dates are heard clearly. A question asks the new date. The failure is status, not date recognition. Cross out or downgrade the old value when the change is confirmed.
Case 19 | Deadline versus event date
Applications close on 5 June; the event is on 18 June. The learner chooses 18 because it is mentioned more recently. Role labels defeat recency bias: apply 5 Jun / event 18 Jun.
Case 20 | Payment date versus booking deadline
The booking must be made Friday, payment by Monday. Both are deadlines for different actions. “Deadline” as a generic label is insufficient. Use action-specific labels.
Case 21 | Day/month convention creates ambiguity
A printed field shows 03/04 while the spoken audio says “the third of April”. The examination’s stated convention and spoken form should align the answer. Learners should not import a regional date convention without checking the target system.
Case 22 | Relative date requires calculation
The speaker says “two days after the 14th”. If the task asks for the final calendar date, calculation may be required. Preserve the relation first, then compute during a safe window if possible. Do not perform arithmetic while losing subsequent audio unless the format forces immediate entry.
Case 23 | “Next Friday” depends on reference date
Relative expressions require a temporal anchor. If the reference date is printed, use it. If not, the recording should provide enough context. Avoid importing the learner’s real-world current date into an examination scenario.
Case 24 | Time is clear but twelve-hour label is lost
Seven in the morning and seven in the evening are acoustically identical at the numeral level. AM/PM context is part of the time proposition.
Case 25 | “Quarter to eight” becomes 8:15
The learner maps “quarter” to 15 but loses direction. Clock language combines quantity with relation. Short focused drills can automate it before returning to full listening.
Case 26 | “Half past eight” triggers calculation delay
A learner silently converts the phrase every time and misses the following word. The goal is chunk-level recognition: half past eight should map quickly to the time concept rather than require repeated arithmetic.
Case 27 | “By eight” becomes “at eight”
Boundary language is lost. If a question asks latest arrival time, “by eight” matters. Exact-detail listening includes prepositions that define the number’s role.
Case 28 | “From eight” becomes “until eight”
Start and end are reversed. The number is identical; relationship carries the meaning. Train time expressions as propositions, not digit retrieval.
Case 29 | Arrival time versus departure time
A timetable contains both. The learner hears them accurately but swaps fields. Printed headings can externalise the role if previewed.
Case 30 | Duration becomes finish time
The speaker says the session starts at two and lasts ninety minutes. The learner writes 3:30 when the field asks duration. Correct calculation answers the wrong job. Question control remains necessary even in numerical tasks.
Case 31 | Price versus deposit
The full fee is 90 and the deposit is 20. A booking question asks what must be paid now. “90” is a true price and a wrong answer. Stage belongs to money.
Case 32 | Deposit is part of the total, not additional
A learner hears both numbers and adds them, assuming the deposit is extra. The source says “a 20 deposit towards the 90 fee”. Exact detail can require relationship, not arithmetic reflex.
Case 33 | Discount percentage versus final price
“Twenty per cent off” and “the final price is 64” appear in one clip. If the field asks price paid, 20 per cent is the wrong type. Role before value.
Case 34 | Refund amount versus refundable condition
A fee is refundable before Friday. The learner records the amount but misses the time condition. Exact money detail still sits inside conditional logic.
Case 35 | Per-person price versus group total
Both values may be stated or one derived. The task asks per person. Scope belongs to currency.
Case 36 | Minimum quantity versus recommended quantity
The speaker says participants need at least two copies but recommends three. A question asks minimum. Modality and boundary words decide between true numbers.
Case 37 | Capacity versus allowed attendance
The room holds 80, but registration is capped at 60. Physical capacity and operational limit differ. Label both roles.
Case 38 | Measurement unit changes after correction
The speaker first says metres, then corrects to centimetres. The numeric value may remain the same while the quantity changes dramatically. Units must be editable status fields.
Case 39 | A number is quoted by one speaker and rejected by another
Speaker A estimates 200; Speaker B says the verified count is 160. The question asks the confirmed figure. Source plus status determines the answer.
Case 40 | A statistic belongs to a study, not the current event
A monologue cites 45 per cent from previous research, then gives 38 per cent for the current sample. Both are relevant and exact. Time/source labels prevent substitution.
Case 41 | A year is part of a study name
The speaker references a “2024 survey” and later says data were collected in 2023. A question about collection year needs the second value. Noun association can mislead exact-detail selection.
Case 42 | A sequence number is mistaken for chronology
“Model 3” is a product identifier, not the third event. Labels protect semantic category.
Case 43 | Room 204 becomes two hundred and four participants
Again, the string is correct and role is wrong. This case is useful in training because it proves that exact-detail errors are often relational, not auditory.
Case 44 | Telephone digits are grouped differently from the learner’s habit
The speaker uses an unfamiliar grouping pattern. The learner should follow the audio chunks while preserving order. Re-chunking can be done later if necessary; first capture should minimise transformation.
Case 45 | Email address contains a hyphen
The learner hears the words correctly but omits the symbol because ordinary vocabulary dominates attention. Email-style tasks require symbol language to become meaningful units.
Case 46 | Underscore versus hyphen
These are different characters. Exact digital identifiers need exact symbols where assessed. Practise them explicitly if part of the target format.
Case 47 | “Dot” is repeated inside a domain
A long address can overload sequence memory. Chunk around natural structural boundaries: local part / at / domain / dot / suffix.
Case 48 | The learner autocorrects an unfamiliar domain mentally
A known website pattern replaces the actual spelling. Familiarity should not override source evidence in identifiers.
Case 49 | A street number and postal code share digits
Similar strings compete. Form fields are valuable external structure; use them instead of holding all contact information in one memory cloud.
Case 50 | A name is unnecessary but steals attention
The question asks why a schedule changed. The speaker introduces an unfamiliar organiser’s surname immediately before the reason. Leonie tries to capture the name and misses the answer. Exactness must be task-weighted. Not every proper noun deserves precision.
Case 51 | A number is unnecessary but memorable
A talk mentions a striking statistic as background. The question asks the speaker’s recommendation. Maren writes the number and misses the transition to the recommendation. Precision is valuable only when relevant to the job.
Case 52 | The learner hears “about seventy” and notes “70?”
The question mark can be interpreted later as uncertainty about the digits rather than a marker of approximation. Notation should distinguish source approximation from listener uncertainty. For example, “~70” versus “70?”.
Case 53 | Two uncertainties collide
The learner is unsure whether the value is 17 or 70 and also whether it belongs to group A or B. Mark both dimensions if replay is available: value? source? This gives the second pass a precise target. A generic question mark does not.
Case 54 | One exact detail is remembered because it is emotionally salient
A dramatic price or age sticks in memory and displaces a less interesting but tested value. Examination relevance should control attention more than emotional salience.
Case 55 | One exact detail is remembered because it matches prior knowledge
The learner expects a familiar date and retains it despite the audio giving another. Prediction proposes; source decides. This rule is especially important for historically familiar topics.
Case 56 | A repeated number changes meaning
The same “20” appears first as a proposed fee and later as a participant limit. The learner assumes repetition confirms the price. Role can change even when form repeats.
Case 57 | Speaker says “no more than 20”
The learner writes 20 and later answers that exactly 20 people must attend. The boundary is a maximum, not an exact target. Exact detail includes logical operator.
Case 58 | Speaker says “at least 20”
The same number now marks a minimum. A naked number loses the direction of the constraint.
Case 59 | Speaker says “between 20 and 30”
The learner selects 20 because it was heard first. A range is the proposition. Endpoints are not interchangeable with the interval.
Case 60 | The learner becomes accurate after she stops chasing perfection
This is the central outcome. She captures value plus role, marks one uncertain character cheaply, and continues. Paradoxically, accepting bounded uncertainty improves whole-paper exactness because fewer later details are sacrificed.
Advanced Exact-Detail Labs
Lab 1 | Form-versus-role split
Give the learner a clip with several numbers. First score whether each form was heard correctly. Then score whether each was attached to the correct role. This simple two-score method reveals whether training should target acoustics or binding.
Lab 2 | Unit swap
Use the same numerical values with minutes, dollars, metres and participants. The learner must record value + unit. This prevents a digit-only listening habit.
Lab 3 | Status chain
Present an original time, revised time and final confirmed time. Ask for each state and the final answer. Transfer the same architecture to dates, prices and locations.
Lab 4 | Role collision
Use the same number for two roles inside one recording: Room 12 and 12 participants, for example. This makes semantic labels necessary.
Lab 5 | Teen/ty speaker rotation
Practise common contrasts across several clear speakers rather than one repeated model. Record first-contact accuracy and adaptation. The goal is robust recognition, not memorising one voice.
Lab 6 | Approximation ladder
Compare exactly 50, about 50, nearly 50, just over 50 and no more than 50. Ask the learner to state the mathematical or semantic relationship, then restore the phrases to audio.
Lab 7 | Date-role matrix
Use event date, booking deadline, payment date and arrival date across several practical recordings. The learner fills a four-column matrix. Later remove the visible labels and require mental role binding.
Lab 8 | Time-expression automaticity
Use quarter past, quarter to, half past, by, from and until in short natural phrases. The purpose is to reduce conversion latency so the learner can return to the next clause quickly.
Lab 9 | Money-state matrix
Standard price, discounted price, deposit, balance and refund are presented across a small set. The learner must preserve both amount and payment state.
Lab 10 | Name recognition before spelling
First match spoken names to printed options. Once auditory identity is stable, remove the printed form and add spelling. This progression separates sound recognition from orthographic generation.
Lab 11 | One-character uncertainty
Use a spelled identifier with one deliberately difficult letter. The learner marks the uncertain position and continues. On replay, only that character is targeted. This trains precise uncertainty.
Lab 12 | Code chunking
Compare recall of six ungrouped characters with the same code spoken in meaningful chunks. Then ask the learner to create her own stable chunking notation. The purpose is reducing working-memory load, not inventing elaborate codes.
Lab 13 | Contact-detail form
Use a name, street number, phone, email and appointment time. The form itself supplies role structure. Score exact form and field placement separately.
Lab 14 | Exact-detail recovery stress
Place one difficult spelling item before two easy questions. The score of the drill is downstream preservation. The learner should learn that losing one character does not justify losing the next proposition.
Lab 15 | No-note versus labelled-note comparison
For a short set, compare internal memory with minimal VALUE+ROLE notes. If notes improve delayed accuracy without causing later misses, they are useful. If accuracy falls, note cost is too high.
Lab 16 | Typed versus handwritten detail
Use matched sets to determine whether digital input creates transpositions or spelling errors. The result routes practice to the correct downstream modality.
Lab 17 | Delayed-detail corridor
Introduce an exact value early and ask about it after a minute of other material. This tests whether role labels and notes create durable retrieval paths.
Lab 18 | Salience resistance
Include one dramatic but irrelevant statistic and one quieter answer-bearing value. The learner must follow the question job rather than memory salience.
Lab 19 | Mixed precision section
Combine name, date, price, percentage and code in one realistic sequence. The learner must recognise which detail type is currently active and change output precision accordingly.
Lab 20 | Full unseen exact-detail test
Use unfamiliar speakers and contexts under target playback rules. Review every error by form, role, unit, status, scope, output or recovery. A single total score hides too much.
Advanced measurement framework
Metric 1 | Form recognition
Was the exact sound string correctly perceived?
Metric 2 | Role binding
Was the string attached to the correct event, person or field?
Metric 3 | Unit retention
Did the learner preserve measurement or currency?
Metric 4 | Operator retention
Did at least, by, before, about and related qualifiers survive?
Metric 5 | Status accuracy
Was the current or final value distinguished from obsolete states?
Metric 6 | Output accuracy
Did the internal exact detail become the intended written or selected form?
Metric 7 | Recovery cost
How many later items were damaged by one difficult detail?
Metric 8 | Delayed retention
Can labelled detail survive intervening material?
Metric 9 | Transfer penalty
How much does accuracy fall with new speakers, topics or response modes?
Metric 10 | Late-paper precision decay
Does exact-detail output become less reliable under fatigue?
Advanced teacher protocol
When a detail is wrong, ask the learner what she heard before revealing the answer. If the form was correct, inspect role, unit and status. If the form was wrong, inspect acoustic access. If the oral detail is right but the written entry is wrong, inspect spelling or input. This sequence prevents one “careless” label from hiding several distinct mechanisms. Use minimal contrasts early, restore natural context quickly, and always test whether a repaired detail survives a new speaker and new task.
Advanced self-study protocol
After a practice set, create a tiny table: heard form / role / final answer / first weak link. Keep only recurring errors active. A learner who repeatedly misses teen/ty needs different work from one who hears every number but swaps deadline and event date. Use replay diagnostically, transcripts only after an unaided attempt, and finish every repair with unseen transfer.
Advanced parent protocol
Parents can help with short, low-pressure activities: give two times with different roles, one corrected price, or one name spelled aloud. Ask “what was that number for?” rather than only “what number did you hear?” This keeps the exercise semantic. Avoid turning ordinary family communication into constant precision testing; broad language use remains more important than endless digit drills.
Advanced final checklist
- I preserve exact form and meaning role together.
- I know when the task actually requires exact spelling.
- I can distinguish teen/ty numbers across unfamiliar clear speakers.
- I preserve decimal points, units and percentage bases.
- I separate deadline, event date, arrival time and start time.
- I separate standard price, discount, deposit and refund.
- I preserve approximation, boundaries and conditions.
- I can chunk long identifiers without rehearsing them indefinitely.
- I update corrected characters or values locally.
- I do not let an irrelevant proper noun steal answer-bearing audio.
- I can carry important exact detail across delay.
- I can lose one difficult detail without losing the next question.
Advanced closing synthesis | Precision becomes useful only when it remains attached
The mature listener does not collect numbers, dates and names as trophies. She stores precise information inside a working model. A time has an event. A price has a category and payment stage. A percentage has a base. A name has a source. A code has an identity function. A correction has a final state. Once the detail is safely bound, the listener can stop rehearsing it and return to the moving stream.
That is the real discipline of exact-detail listening: enough precision to score, enough context to mean the right thing, and enough recovery control to ensure that one difficult character never becomes five lost answers.