TOEIC Link Speaking Number and Data Reporting Accuracy Under the Describe-the-Information Response: The Quantitative Precision Discipline That Keeps a Fluent Answer From Reporting the Wrong Figure

A TOEIC Link Speaking describe-the-information task hands the candidate a schedule, a table, or a chart and asks them to answer a caller's questions from it: when a session starts, how much a plan costs, which option is cheaper, how many seats remain. The scoring rewards a response that reads the data accurately and frames it in natural spoken English — and the single most damaging error is not hesitation but a confidently delivered wrong figure. A candidate who says "the workshop begins at two o'clock" when the table says three has not made a small slip; they have given the caller information that would make them late, and the response is penalized for the failure of its core job even when the surrounding English is fluent.

The reason wrong figures slip out is that the describe-the-information task splits attention between two demands that compete under time pressure: reading the right cell of a dense table and producing a grammatical spoken sentence around it. A candidate concentrating on fluency loses track of which row the question pointed to; a candidate concentrating on the table produces broken sentences. The accuracy failure is rarely ignorance of the number — it is a lookup error committed while the speaker's attention was on the sentence, the spoken equivalent of reading the line above the one you meant. The precise extraction this task demands is the productive twin of the receptive skill that the listening numerical data and comparison extraction discipline trains — both fail at the same point, when a number is grabbed from the wrong place while attention is elsewhere.

This article is the quantitative precision discipline for TOEIC Link Speaking describe-the-information responses. The guide covers the cell-locking habit that fixes the right data point before speaking, the comparative language that reports relationships between figures without inverting them, the unit and qualifier control that keeps a number meaningful, and the accuracy-first recovery that protects the figure when fluency falters.

The cell-locking habit that fixes the right data point before speaking

The defense against the lookup error is to separate finding the number from saying it, locking the correct cell before any sentence begins so the speaker is reading from memory of a verified value rather than scanning the table mid-utterance.

Match the question's two coordinates before reading any value. Most describe-the-information questions point to a cell defined by two coordinates — a row (which session, which plan, which date) and a column (time, price, location). The candidate confirms both coordinates first: "the caller asked about the afternoon session's room" locates row and column before the eye reads the value. Reading a value before both coordinates are fixed is how a candidate reports the morning room for an afternoon question.

Say the value silently once before saying it aloud. In the second between locating the cell and beginning the sentence, the candidate rehearses the bare value internally — "three o'clock," "forty dollars" — so the sentence is built around a number already held in mind rather than one being read while the mouth moves. This split protects the figure from being displaced by the grammatical effort of the sentence.

Re-anchor on the table for each new question, not each new clause. When a prompt asks several questions in sequence, the candidate returns to the table once per question and then speaks the full answer without glancing back. Re-scanning mid-sentence is what introduces the wrong-row error; one deliberate lookup per question, fully resolved before speaking, keeps each answer anchored to a single verified cell.

The comparative language that reports relationships without inverting them

Many describe-the-information questions ask not for a single value but for a relationship — which is cheaper, which is earlier, how much more — and comparative answers fail in a distinct way: the figures are read correctly but the relationship is stated backwards.

Fix the direction before choosing the comparative word. Before saying "cheaper" or "more expensive," the candidate confirms which item the comparison favors: "the standard plan is the cheaper one" decided first, then framed. Choosing the comparative word while still reading the two prices is how "A is cheaper than B" comes out inverted when A was actually the higher figure.

Report the difference as a computed value, not a re-read. When the question asks "how much more," the candidate states a single computed difference — "twenty dollars more" — rather than reading both prices and leaving the caller to subtract. Computing the difference once and reporting it removes the chance of stating two correct figures in a relationship that misleads.

Name the item, then the comparative, then the benchmark. A reliable comparative frame is "the [item] is [comparative] than the [benchmark]" — "the evening class is later than the morning class." Keeping that fixed word order prevents the mid-sentence scramble that produces inverted comparisons, the same ordering discipline that the speaking circumlocution and paraphrase fallback strategy relies on when a precise word is unavailable and structure has to carry the meaning.

The unit and qualifier control that keeps a number meaningful

A figure reported without its unit or its qualifier is incomplete information, and the describe-the-information task penalizes a bare number that leaves the caller unable to act on it as surely as it penalizes a wrong one.

Attach the unit every time, even when it feels redundant. "Forty" answering a price question is not an answer; "forty dollars" is. The candidate carries the unit as part of the value from the cell-locking step, so the number is never held in mind without it. Times take their period — "at nine in the morning" — because "at nine" leaves a caller guessing between two plausible hours.

Preserve the qualifier that bounds the figure. Tables often qualify a value — "from $40," "up to 20 seats," "starting at 9 a.m." — and dropping the qualifier changes the fact. "It costs forty dollars" misreports a cell that said "from $40"; the accurate report is "it starts at forty dollars." The qualifier is part of the data, and reporting the number without it is a quiet accuracy error that a careful caller would catch.

State the value at the table's precision, not rounded. If the schedule says 2:30, the answer is "two thirty," not "around two." Rounding a precise figure into an approximation discards information the task provided and the caller needs, and it reads as a candidate who could not read the cell rather than one who chose to simplify.

The accuracy-first recovery that protects the figure when fluency falters

Under time pressure the candidate will sometimes lose the thread of a sentence mid-delivery, and the recovery move determines whether the figure survives — accuracy-first recovery treats the number as the part of the sentence to protect when the grammar around it breaks.

When a sentence stalls, the candidate restarts the frame but keeps the verified figure fixed: "the session is — the afternoon session begins at three o'clock," rebuilding the sentence around the number rather than re-reading the table and risking a fresh lookup error. A clean self-correction that preserves the right figure costs a moment of fluency and loses nothing the task scores most; a smooth re-read that pulls a new, wrong number loses the point the task exists to award. The same self-monitoring that the listening disfluency marker and self-repair decoding discipline studies in others' speech is what the candidate applies to their own: a visible repair that lands on the correct value is a strength, not a flaw. The figure is the answer; the sentence is its delivery. When the two compete, the discipline protects the figure, because a describe-the-information response is scored first on whether the caller could act correctly on what they heard — and a fluent wrong number fails that test that a halting right one passes.