TOEIC Link Reading — Skimming and Scanning as Separable Skills: How Deliberate Read-Mode Switching Recovers Two to Four Minutes Per Passage and Lifts the Reading Band
Skimming and scanning are the two most talked-about and least precisely trained reading techniques on the TOEIC Link reading module. Candidates are told to "skim the passage first" and "scan for the answer," but the two verbs are used interchangeably as if they named a single skill, when in fact they are distinct cognitive operations with different purposes, different eye-movement patterns, and different failure modes. A candidate who does not separate them ends up doing neither well — reading too slowly to be scanning and too shallowly to be comprehending. Internal timing data from practice sessions indicates that candidates who deliberately switch between the two modes complete the reading section with two to four minutes of margin, while candidates who read in a single undifferentiated mode routinely run out of time on the final passages and guess. The recoverable time is real, and it comes not from reading faster in general but from reading in the right mode for each sub-task.
The core distinction is one of goal. Skimming is reading for the gist — the overall structure, topic, and argument arc of a passage — by moving quickly across the whole text and sampling high-information positions while skipping low-information ones. Scanning is searching for a specific piece of information — a name, a number, a date, a keyword — by moving the eyes rapidly over the text while suppressing comprehension entirely until the target appears. Skimming builds a mental map; scanning locates a coordinate on that map. A strong reader skims once to build the map, then scans repeatedly to answer individual questions, and never confuses the two. For the comprehension operations that sit on top of the map, see the author purpose and tone identification guide and the answer choice comparison and elimination sequencing guide.
Skimming — the mechanics of building a map
Skimming is not reading fast; it is reading selectively. The eye does not touch every word. Instead it lands on the positions in a text that carry the most structural information and skips the rest.
Where to land when skimming
The highest-information positions are predictable. The first and last sentences of each paragraph carry the topic and the conclusion. The first paragraph carries the thesis or the purpose. The last paragraph carries the resolution. Within a paragraph, discourse markers — however, therefore, in contrast, for example — signal the logical turns that structure the argument. A skilled skimmer reads the opening, drops to the first sentence of each body paragraph, notes the discourse markers, and reads the closing, sampling perhaps a third of the words while capturing most of the structure. The output of skimming is not a summary in words; it is a spatial sense of where things are — where the problem is stated, where the evidence sits, where the conclusion lands.
What skimming is for
Skimming answers the global questions: main idea, author purpose, overall tone, and the logical relationship between paragraphs. These questions cannot be answered by scanning, because the answer is distributed across the whole passage rather than located at one point. Attempting to answer a main-idea question by scanning for a keyword is a category error, and it is one of the most common ways candidates pick a plausible-but-wrong answer that matches a local detail rather than the global point.
Scanning — the mechanics of locating a coordinate
Scanning is a search operation, and its defining feature is comprehension suppression. When scanning, the reader deliberately does not process meaning; the eyes sweep the text looking only for a visual or lexical target, and full reading resumes only when the target is found.
How to scan efficiently
Before scanning, the reader fixes the target in mind: the specific word, the type of number (a date, a percentage, a dollar figure), or a proper noun. Then the eyes move down the text in a fast vertical or zigzag sweep, matched against the target. Numbers and capitalized proper nouns are especially fast to scan for because they are visually distinct from surrounding lowercase prose. Once the target is located, the reader stops scanning, switches to careful reading of the sentence containing the target and the sentences immediately around it, and extracts the answer.
The paraphrase trap in scanning
The danger in scanning is that the correct answer is rarely stated in the exact words of the question. The question paraphrases the passage, so scanning for the question's literal keyword often fails or, worse, lands on a distractor that reuses the keyword in a wrong context. The scan target must therefore include synonyms and the underlying concept, not just the surface word. This is why scanning and paraphrase recognition are paired skills, covered in depth in the answer option paraphrase recognition without keyword matching guide.
The four time-wasting errors
Error 1 — Reading everything in one careful mode
The default failure is reading every passage word by word at comprehension speed, treating skimming and scanning as luxuries. This guarantees running out of time, because the reading section is engineered so that full careful reading of every word does not fit the clock. The fix is structural: adopt the two-phase skim-then-scan workflow as the default, and reserve careful reading only for target sentences.
Error 2 — Skimming when the question demands a scan
A candidate who skims for a detail question wastes time sampling the whole passage when the answer is at a single point. Detail questions — those asking for a specific fact, figure, or named entity — should trigger an immediate scan, not a re-skim. Matching the mode to the question type is the whole discipline.
Error 3 — Scanning when the question demands a skim
The mirror error: scanning for a keyword to answer a main-idea or author-purpose question. Because the answer is global, no single scanned location contains it, and the candidate ends up either guessing or, more dangerously, anchoring on a local detail that matches the keyword but contradicts the passage's overall point.
Error 4 — Failure to switch back
Some candidates skim well and scan well but never make the deliberate transition between modes; they drift from a scan back into slow careful reading of irrelevant sentences, losing the time the scan was supposed to save. The transition must be conscious: find the target, read the local sentences, extract the answer, and return to scan mode for the next question.
The four-week protocol
Week 1 — Isolate skimming
Practice skimming alone. Take a passage, give yourself sixty seconds to skim it, then write down the main idea, the author's purpose, and the function of each paragraph without looking back. This trains the map-building operation independently of question answering.
Week 2 — Isolate scanning
Practice scanning alone. Given a passage and a list of specific facts to find (a date, a name, a figure), time how fast you locate each without reading for meaning. The goal is to make comprehension suppression deliberate and the eye sweep fast.
Week 3 — Integrate the two phases
Combine skim-then-scan on full passages. Skim first to build the map, then answer each question by choosing the correct mode — skim-derived memory for global questions, a fresh scan for detail questions. Audit each miss for whether the wrong mode was used.
Week 4 — Full-section timing
Run full reading sections under time, tracking margin. The target is to finish with two to four minutes to spare, which is the signature of correct mode allocation. If time runs out, the diagnosis is almost always Error 1 — sliding back into single-mode careful reading — and the fix is to re-commit to the two-phase workflow.
What the reading band actually measures
The reading band measures comprehension accuracy under a time constraint, and the constraint is the whole point. A candidate who could answer every question correctly with unlimited time but runs out at question thirty is scored on the questions reached, not the questions they could have reached. Skimming and scanning are the two mechanisms that make the time constraint survivable, and treating them as one blurred skill is why so many capable readers plateau below their comprehension ceiling. Separated, trained independently, and then integrated into a deliberate mode-switching workflow, they convert the reading section from a race that ends in guessing into a controlled process that finishes with margin. The band gain is not from reading better; it is from reading in the right mode at the right time.