TOEIC Link Reading — Skimming and Scanning Techniques: The Eye-Movement Patterns That Decide Part 7 Time Management

TOEIC Link Part 7 — the reading-comprehension section with single, double, and triple passages — fails most test-takers on time rather than on comprehension. The total time budget for the reading section is roughly 35 minutes for 30 reading items; Part 7 typically consumes 22 to 25 of those minutes if read linearly, leaving 10 to 13 minutes for Parts 5 and 6, which is structurally insufficient. The candidates who finish on time are not faster readers in the absolute sense — they are candidates who deploy a disciplined two-pass reading strategy that separates the skim (a coarse pass that builds a passage map) from the scan (a targeted pass that locates the answer-bearing sentences). The two-pass strategy compresses Part 7 to 16 to 19 minutes without loss of accuracy and produces the time buffer that the section demands.

This guide describes the two-pass strategy in detail, identifies the six passage genres that appear in Part 7 and the genre-specific skim signatures, and provides the eye-movement patterns that operationalize the strategy. For related Part 7 topics, see the guides on paraphrase recognition techniques, reading strategies by question type, and pacing and time management.

Why linear reading fails on Part 7

Three properties of Part 7 make linear, top-to-bottom reading a structurally losing strategy.

Property 1 — most Part 7 items are localized. Roughly 70% of Part 7 questions ask about a single sentence or a single short paragraph in the passage. The remaining 30% are gist, inference, or cross-passage synthesis questions that require a broader view of the passage. A linear reader processes every sentence with equal attention, which means roughly 70% of the reader's attention is spent on sentences that are not directly question-targeted. The two-pass reader spends a deliberate eight to twelve seconds building a map of the passage, then targets attention to the question-bearing sentences during the scan pass — which is a more efficient allocation of attention given the localized question distribution.

Property 2 — passage structure is highly predictable. Part 7 passages fall into six genres — business correspondence, advertisements and notices, articles and reports, multi-message threads, application forms, and policy or procedure documents. Each genre has a predictable structure: subject lines and openings near the top, body in the middle, signatures and closing notes at the bottom. The skim pass exploits the predictability by inspecting the high-information positions (subject lines, openings, list markers, signature blocks) rather than reading the whole passage. The exploitation is reliable because the genres are stable across forms.

Property 3 — questions provide free passage cues. Each Part 7 question contains a set of lexical cues — proper nouns, dates, dollar amounts, product names, role titles — that anchor the answer to a specific position in the passage. The scan pass uses the cues to jump directly to the answer-bearing sentence rather than re-reading the passage in search of the answer. The cue-driven scan is roughly four times faster than a re-read and is more accurate because the scanner is looking for a specific lexical match rather than reconstructing the passage content from memory.

The two-pass strategy in detail

The two-pass strategy separates Part 7 into two distinct cognitive activities.

Pass 1 — the skim: build a passage map in eight to twelve seconds

The skim pass is not a reduced-speed read of the whole passage. It is a deliberate inspection of the high-information positions of the passage, with the goal of producing a four-element passage map.

Map element 1 — genre. Identify which of the six genres the passage belongs to. The genre determines which positions carry the most information and which positions can be safely ignored.

Map element 2 — sender and recipient. Identify who is writing to whom. For business correspondence, this is the From and To headers. For articles and reports, this is the byline and the implied audience. For multi-message threads, this is the participant list. Sender-recipient information is question-targeted on roughly 20% of Part 7 items, so the map element is high-yield.

Map element 3 — primary subject. Identify what the passage is about in five to eight words. The subject is usually carried by the subject line for correspondence, the headline for articles, the product name for advertisements, or the policy title for procedure documents.

Map element 4 — passage structure markers. Identify the structural markers that segment the passage — paragraph breaks, list markers, bolded headings, separator lines, signature blocks. The markers are the scan-pass anchors that allow the scanner to jump to the relevant section without re-reading.

The total skim time should be eight to twelve seconds for a single passage, fifteen to twenty seconds for a double passage, and twenty to twenty-five seconds for a triple passage. The skim does not require comprehension of every sentence — it requires recognition of the four map elements.

Pass 2 — the scan: locate the answer-bearing sentences in three to six seconds each

The scan pass reads each question, extracts the lexical cues from the question stem, and jumps to the position in the passage where the cues are most likely to appear. The scan pass relies on the passage map produced by the skim — without the map, the scanner has no efficient way to locate the answer-bearing sentence and reverts to linear reading.

Cue type 1 — proper nouns. The question stem contains a name, a company, a city, a product, or a brand. Scan the passage for the proper noun and read the surrounding sentence.

Cue type 2 — numerals. The question stem contains a date, a price, a quantity, or a percentage. Scan the passage for the numeral and read the surrounding sentence.

Cue type 3 — distinctive lexical items. The question stem contains an unusual word or phrase that is unlikely to appear elsewhere in the passage. Scan the passage for the lexical item and read the surrounding sentence.

Cue type 4 — structural position. The question stem references a specific structural position (the closing of the email, the second paragraph, the postscript). Scan the passage for the structural marker and read the targeted position.

A well-executed scan takes three to six seconds per question, with the longer end reserved for inference questions that require reading two or three surrounding sentences rather than a single sentence.

The six passage genres and their skim signatures

Genre 1 — business correspondence (emails, letters, memos)

The high-information positions are the From, To, Subject, and Date headers at the top, and the closing signature block at the bottom. The body typically opens with a context-setting sentence and closes with an action-request sentence. The skim pass should produce: sender role, recipient role, subject in five to eight words, and the position of the action request if present. The action request is question-targeted on roughly 35% of correspondence items.

Genre 2 — advertisements and notices

The high-information positions are the headline, the product or service name, the price or offer details, and the contact information. Promotional copy in the middle is usually low-information. The skim pass should produce: product or service, primary offer, price or eligibility, and contact information position. The price-offer pair is question-targeted on roughly 30% of advertisement items.

Genre 3 — articles and reports

The high-information positions are the headline, the lead paragraph, paragraph-opening sentences (topic sentences), and the closing paragraph. The lead paragraph in journalistic prose follows the inverted-pyramid pattern and carries the who-what-when-where information; subsequent paragraphs elaborate. The skim pass should produce: headline gist, lead-paragraph facts, and the topic-sentence structure of the body.

Genre 4 — multi-message threads

The high-information positions are the message-by-message headers (From, Subject, Date) and the opening sentence of each message. The structure is sequential — earlier messages set context, later messages add information or request action. The skim pass should produce: participant list, message sequence, and the focal request or decision in the final message. Cross-message questions appear on roughly 40% of thread items.

Genre 5 — application forms

The high-information positions are the form title, the section headings, and the filled-in fields. Instructions and notes are usually low-information for question-answering purposes. The skim pass should produce: form purpose, applicant identity if filled, and the position of any conditional or eligibility fields. The eligibility fields are question-targeted on roughly 25% of form items.

Genre 6 — policy or procedure documents

The high-information positions are the document title, the section headings, the numbered or bulleted steps, and any deadline or eligibility statements. The opening paragraphs typically state the policy purpose; the body presents the procedure. The skim pass should produce: policy title, the structural outline of the procedure, and any explicit deadlines or eligibility conditions.

The eye-movement patterns that operationalize the strategy

The two-pass strategy is reliable only if the eye movements during each pass are disciplined. Three patterns matter most.

Pattern 1 — saccadic skipping during the skim. During the skim pass, the eyes should make large saccades — jumps of several words at a time — to the next high-information position rather than reading word-by-word. A trained skimmer makes roughly two to four saccades per line during the skim, compared to seven to ten saccades per line during a linear read. The saccadic-skipping pattern requires practice because the default eye-movement pattern is the linear-reading pattern; deliberate effort is required to suppress the default during the skim.

Pattern 2 — cue-driven jumps during the scan. During the scan pass, the eyes should jump to the position of the lexical cue and fixate on the surrounding context rather than reading sequentially from the top of the passage. The cue-driven jump is enabled by the passage map produced during the skim — without the map, the scanner has no target position and reverts to linear reading.

Pattern 3 — local reading at the answer-bearing sentence. Once the scanner has located the answer-bearing sentence, the eye-movement pattern reverts to careful linear reading for the two or three sentences surrounding the cue. This local-reading pattern is where comprehension actually occurs; the skim and scan passes were preparation for it. The local-reading window is short enough that careful comprehension is feasible within the time budget.

Implementing the strategy on a practice passage

The fastest way to internalize the two-pass strategy is to apply it to an official sample Part 7 passage with a stopwatch. Time the skim pass (target: eight to twelve seconds for a single passage), time each scan (target: three to six seconds per question), and verify that the local-reading windows landed on the correct answer-bearing sentences.

A candidate who runs ten practice passages with the stopwatch typically internalizes the eye-movement patterns within a single 90-minute session. The two-pass strategy then transfers to the live test with minimal additional practice, because the predictability of the passage genres means that the strategy generalizes across forms.

For complementary Part 7 guides, see TOEIC Link Reading — Paraphrase Recognition Techniques, TOEIC Link Reading — Strategies by Question Type, TOEIC Link Pacing and Time Management, and TOEIC Link Reading Module Overview.