Distractors in TOEFL Reading follow a small number of recipes. The nine traps that cost the most marks, how to spot each one, and the habits that help.
12 min readMonthly & Premium
Quick answer
Most Reading losses come from a short list of repeating traps: word matching, plausible but unstated options, dropped qualifiers, answering from memory, taking the first mention, promoting a detail to a thesis, and the two big time traps of double reading and refusing to guess. Each has a specific counter.
Distractors are manufactured, not random
The wrong options on this exam are written, deliberately, by someone whose job is to make a wrong answer look right to a student who has understood the passage almost well enough. They are not filler. They are targeted.
Which is very good news, because manufactured things have patterns. Once you can name the recipe a wrong option was built from, you stop being caught by it, and the recipes are few enough to learn in an afternoon.
This lesson is the catalogue. Nine traps, roughly in order of how many marks they cost the students whose results we look at. Six of them are content traps, dealing with how options are constructed. Three are behaviour traps, dealing with what you do with your time and your nerve. The behaviour traps cost more marks than the content traps, which surprises people.
Read this lesson after you have sat at least one timed Reading section, because the traps are much more memorable when you can recognise yourself in them.
The nine traps
Trap
Type
What it costs you
The counter
Word matching
Content
Factual and vocabulary questions
Paraphrase the answer in your own words before reading the options.
Plausible but unstated
Content
Inference questions, mostly
Point at the sentence. No sentence, no answer.
Dropped qualifier
Content
Factual and inference questions
Read most, some, believe, appears, does not as though they were bold.
First mention
Content
Factual questions
Check the sentence contains the information, not just the search word.
Detail promoted to thesis
Content
Important idea questions
Run the coverage test against your paragraph labels.
Invented ranking
Content
Important idea and inference questions
A contrast is not a ranking. Look for the comparative word in the passage.
Answering from memory
Behaviour
Every question type
Put your eyes back on the line. Every time, no exceptions.
The double read
Behaviour
The last five to eight items of the section
Skim once for structure, then let questions navigate.
Refusing to guess
Behaviour
Two or three items per section
Hard cap of 50 seconds, then choose and move.
Content traps come from how options are written. Behaviour traps come from how you spend the 30 minutes. Both are fixable, but the behaviour ones move your score faster.
The six content traps in detail
Word matching. The option that shares the most vocabulary with the passage is usually not the answer. Distractors are built from passage language because that is what makes them look plausible, so heavy lexical overlap is evidence of manufacture rather than of correctness. The correct option is more often a paraphrase. Counter: decide your answer in your own words first, then shop.
Plausible but unstated. An option that is entirely reasonable, that you would agree with in conversation, and that the passage never says. This is the dominant failure on inference questions and it is worst for well-informed students, who supply the missing link without noticing they supplied it. Counter: name the sentence. If you cannot, discard.
Dropped qualifier. The passage says "most vivid dreaming occurs during this phase" and the option says dreaming occurs only during REM. The passage says a degraded field "does not fail" and the option says the farmer abandons it. One word removed, one option made wrong, and the removal is invisible unless you go back to the line. Counter: read the hedges as load-bearing, because they usually are.
First mention. Passages name a term in passing before treating it properly. A question about where most evaporation occurs will find "evaporation" in the general introduction and again in the paragraph that answers the question. Students take the first hit and answer from the wrong sentence. Counter: check that the sentence you landed on contains the information the stem asked for, not just the word.
Detail promoted to thesis. The vivid fact from paragraph 2 offered as the main idea. Billions of microorganisms in a gram of soil. Sixteen percent rising to fifty-four percent. These stick in memory precisely because they are striking, and being striking is the opposite of being a thesis. Counter: coverage test.
Invented ranking. The passage compares two things without ordering them, and an option says one is more important. NREM and REM sleep are described with different functions and no hierarchy. Traditional and mRNA vaccines are described with trade-offs and no winner. Counter: look for the comparative in the passage. If the passage never wrote "more", neither should your answer.
The three behaviour traps, which cost more
Answering from memory. You read the passage four minutes ago and the question feels familiar. You answer from your impression rather than from the text, and your impression does not store the detail the distractor was built from. This is the single most expensive habit in the section, and it is invisible to the student doing it, because answering from memory feels exactly like knowing the answer.
The counter is uncomfortable and simple: every question, eyes back on the line. Students resist this because it feels slow, and it is slower, by about eight seconds a question. Over 25 multiple choice items that is three minutes. It is also worth roughly three to five items on a typical test, which is a great deal more than three minutes buys you anywhere else.
The double read. Reading each academic passage carefully before the questions, then reading it again while answering them. It feels like the responsible approach and it duplicates two and a half minutes per passage. Across a section that is most of the reason people do not finish. Counter: skim for structure, produce four-word function labels, let the questions navigate.
Refusing to guess. One question absorbs two minutes because leaving it unresolved feels like failure. Meanwhile three items at the end of the section go unanswered, and unanswered items score zero with certainty. Counter: 50 second cap, then pick and move. The emotional difficulty of this is real and it is worth rehearsing in practice so that it is not a new experience on test day.
Notice that all three behaviour traps share a root: they are all versions of trying to be more certain than the section has time for. Reading has 50 items in 30 minutes. It is not asking you to be sure. It is asking you to be right more often than not, quickly, fifty times.
A review routine that finds your traps
1
Sit a full timed section without pausing
Traps only show up under real conditions. An untimed test tells you what you understand, which is not what you need to know.
2
For every wrong answer, name the trap before reading the explanation
Look at the option you chose and the correct one, and decide which recipe you fell for. Word matching? Dropped qualifier? Detail promoted to thesis? Do this before the explanation tells you, because the naming is the skill.
3
Count the traps, not the questions
You will find you have two or three recurring ones, not nine. Everybody does. The distribution is what matters, and one bad habit usually accounts for half your losses.
4
Fix one trap per week
Pick the largest cluster and give it a specific counter you can execute under pressure. "Read more carefully" is not a counter. "Paraphrase before reading the options" is.
5
Re-sit a fresh section and recount
If the cluster has shrunk, the counter works and you move to the next one. If it has not, the counter was too vague, and you need a more mechanical version of it.
Practice: name the trap
Each of these is a question a student got wrong. Identify what the wrong option was doing. This is the diagnostic skill that makes your own review sessions useful.
Most of the plastic entering the ocean each year comes from a small number of large rivers, and the majority of that load is delivered during a few weeks of peak flow. Fishing gear, by contrast, enters directly at sea and accounts for a substantial share of the material found in offshore accumulation zones.
The passage says: "Most of the plastic entering the ocean each year comes from a small number of large rivers." A student selects the option "All ocean plastic originates in rivers." What trap is this?
The passage says "most", the option says "all", and the second sentence explicitly names a source that does not come from rivers. Removing a hedge is the cheapest way to manufacture a wrong option, and it works because hedges are the words readers skip.The option does share vocabulary with the passage, but that is incidental. What makes it wrong is the change from most to all, which is a logical alteration rather than a lexical one.Nothing has been promoted from detail to thesis. The option is about the same claim the passage makes, just an absolute version of it.No ranking has been invented. Rivers and fishing gear are both mentioned, and the option does not order them, it simply excludes one.
The defences were designed for a storm surge of a given height and they were not overtopped. They failed because the ground behind them was saturated after three weeks of rain, and water pressure from the landward side lifted sections of the wall off their foundations. Post-incident reviews found the drainage system behind the wall had been decommissioned eight years earlier.
A question asks what the passage identifies as the reason coastal defences failed. A student picks an option about rising sea levels, which are never mentioned. What trap is this?
Sea level rise is a real phenomenon, an obvious candidate for coastal defence failure, and completely absent from this passage, which attributes the failure to saturated ground and a decommissioned drainage system. The option is supplied by the reader knowledge rather than by the text.First mention requires the term to appear at least twice, with the student taking the wrong occurrence. Sea level does not appear at all.There is no lexical overlap to match. The option uses vocabulary the passage never employs, which is precisely the tell.No qualifier has been dropped, because the passage makes no claim about sea level to qualify.
Paragraph 1: Forest ecology has moved away from treating trees as individual competitors.
Paragraph 2: Fungal networks connect root systems across a stand, and isotope tracing shows carbon moving between trees, including between different species.
Paragraph 3: Whether this constitutes cooperation is contested. The fungi benefit from the transfer and may be driving it for their own reasons.
Paragraph 4: What is agreed is that the stand behaves as a connected system, and management that treats trees as independent units mismeasures what it is managing.
A main idea question. The student picks "Fungal networks can transfer carbon between trees of different species." What trap is this?
The option is a true statement from paragraph 2 and it accounts for one paragraph out of four. It misses the contested interpretation in paragraph 3 and the management consequence in paragraph 4, which is where the passage argument actually lands. Coverage test, failed.The overlap with paragraph 2 is real but incidental. A main idea option can share wording with the passage and still be correct if it covers the argument, so overlap is not the diagnosis here.No two things are ranked. The option makes a single factual claim without ordering anything.Answering from memory is a behaviour, and this student has clearly retained the passage accurately. The error is in what they did with an accurate memory, not in the memory.
Paragraph 1: The initial survey and funding round took three years, a delay the trustees later described as the least avoidable part of the project.
Paragraph 3: Physical restoration began in the fourth year and continued for a further eight, so that the building reopened eleven years after the campaign started.
A factual question asks how long the restoration took. The student answers "three years", which appears in paragraph 1. The correct answer, in paragraph 3, is eleven years. What trap is this?
The search term found a number early, in a sentence about the survey and funding phase, and the student stopped there. The sentence contained a duration but not the duration the stem asked for. The counter is to check that the sentence answers the question rather than merely containing a plausible-looking figure.Nothing is ranked. Both figures are durations, presented without comparison.Three years is stated in the passage, so the option is not unsupported. It is supported by the wrong sentence, which is a different failure.No hedge has been removed. Both sentences state their figures plainly.
Once the fault ruptured, the propagation stopped abruptly at a boundary in the rock. Seismologists describe the rupture as arrested at that point, and the arrest is the reason the earthquake was a magnitude six rather than a magnitude seven.
A vocabulary question on "arrested" in this passage. The student picks "taken into custody". What trap is this?
The student answered from the sense of the word they know best rather than from the sentence, which is about a rupture stopping at a rock boundary. Substituting "taken into custody" into the sentence produces nonsense in under three seconds, which is why the substitution test exists.Word matching is choosing an option because it repeats the passage vocabulary. Here the option repeats nothing from the passage. It comes from outside it.Plausible but unstated describes an option that is reasonable in context. Custody is not reasonable in a sentence about fault ruptures, so this option is not plausible at all, merely familiar.No qualifier is involved. The passage makes no hedged claim here.
A student finishes with six items unanswered and reports having spent about two minutes on a single inference question. Which fix should come first?
Six unanswered items is six guaranteed zeros, and the student has named the cause: refusing to abandon a one point question. The cap addresses the actual bottleneck directly, and it is the fix with the largest and fastest payoff for this specific profile.The content traps are worth learning and they are not this student problem. They fix accuracy on items attempted, and this student is losing points on items never reached.General reading speed is the right target for a student who is evenly slow across everything. This student is not evenly slow, they had one catastrophic item, so the diagnosis does not match.The coverage test improves one question type by perhaps one item. The cap recovers up to six. Order the fixes by size.
The pre-test reminder card
Decide your answer in your own words before reading the options.
For every answer, put your eyes back on the line. No exceptions.
Hedges are load-bearing: most, some, believe, appears, does not.
An option you cannot attach to a sentence is out, however true it sounds.
A contrast is not a ranking.
A vivid detail is not a thesis.
Skim once. Never read a passage twice before the questions.
Fifty seconds maximum on any item, then choose and move.
Fill every blank before the section ends. Nothing is deducted for a wrong answer.
Where this course leaves you
Twelve lessons, and the honest summary is shorter than it should be. Reading is 30 minutes and 50 items across two modules. Every item is worth one point. Most students understand more of the passages than their score suggests, and lose marks to a small number of repeated habits rather than to gaps in their English.
If you have read the whole course, the three things worth carrying into the exam are these. Fund the expensive questions with the cheap ones, because the letter-completion paragraphs are where the points are. Go back into the text for every answer, because your memory of a passage does not store the details distractors are built from. And treat Module 1 as the half that matters, because it scores points and it decides which version of Module 2 you get.
Everything else in this course is detail underneath those three.
The last thing to say is about how to use the practice tests. Sitting them is only half the work, and it is the half that feels productive. The other half is the review: naming the trap for every wrong answer, counting the clusters, picking one thing to change. A student who sits four tests and reviews them properly will move further than one who sits twelve and looks only at the band. Reading and Listening results are objective and come back immediately, which means you can run that loop as often as you have the patience for. Writing and Speaking go through expert evaluation and take longer, which is a good reason to do your Reading work while you wait.
Monthly & Premium
7 more sections in this lesson
You have read the opening. The rest covers the method in full, with worked examples and practice questions that explain why each wrong answer is wrong.
All 45 lessons: Reading, Listening, Writing and Speaking
Real exam audio and record-yourself speaking drills
Almost always because you answered from memory of the passage rather than from the text, or because the option you picked repeated passage wording while changing one element. Both are fixed by deciding your answer in your own words and then returning to the line.
Not finishing. Unanswered items score zero with certainty, and the two usual causes are reading each passage twice and refusing to abandon a single hard question.
No. Decide the meaning first and let overlap be irrelevant. Sometimes the correct answer is close to verbatim, particularly on Daily Life texts. Rejecting overlap on principle is just a different way of guessing.
After a timed section, name the trap for every wrong answer before reading the explanation, then count. Almost everyone finds two or three recurring patterns rather than nine.
Never. Nothing is deducted for a wrong answer, so a blank is a zero you chose over a guess with a real chance of scoring.
Fewer than you think, reviewed harder than you expect. Four tests properly reviewed, with a named trap for every wrong answer and one habit changed between attempts, beats twelve tests where you only look at the band.