Reading Comprehension: Search, Map, Answer

Read for evidence, then choose a supported answer

DAT Reading Comprehension asks you to understand and analyze scientific passages. You do not need prior expertise in their scientific topics. Your task is to distinguish what the text states, what follows from it, and what it does not establish.

The section contains 50 questions across three passages in 60 minutes. That is an overall average of 72 seconds per question, including reading. The ADA's current user manual describes roughly 1,100–1,400 words and 15–20 questions per passage; do not assume the questions divide equally. ADA DAT User's Manual.

1. Choose a reading method through practice

Try a method on timed passages and record completion, accuracy, and the types of errors you make. One useful option is map, locate, verify:

1. Map: read for the topic, the author's main point, and the role of each paragraph. Use short notes rather than copying details.
2. Locate: let the question guide you back to relevant evidence.
3. Verify: compare each candidate answer with the text and the exact question asked.

An example map might be:

1  problem: surface coating wears away
2  experiment: two coatings, same exposure
3  results: lower average loss for B
4  limitation: one laboratory condition
5  next step: vary temperature

This map is an index, not a substitute for comprehension. Detail questions may need a close reread of one paragraph; purpose and inference questions may require connections across several. Some readers do well with an initial close read, others with a lighter pass followed by targeted reading. Neither method is universally best.

For test-style practice, use the supplied-note-board style of brief notes rather than relying on personal scratch paper. Consult the current candidate guide for test-center materials and navigation.

2. Identify the task before searching

These are preparation categories, not an official percentage breakdown. Do not use an unsupported claim about which type dominates to neglect the others.

For detail, locate and read the relevant sentence with its context. For inference, identify the premises and the smallest justified conclusion. For tone, distinguish the author's words from a quoted person's stance. A word such as “however” can introduce a limitation, contrast, or rejection; counting signal words cannot settle the answer.

3. Test the whole answer choice

Absolute words are not automatically wrong. If a passage actually supports “all,” that wording may be correct. Likewise, an inference need not repeat a sentence verbatim. The question and evidence decide; answer length or position does not.

When two choices survive, seek the decisive evidence. It might be one qualifier, several sentences together, or a limitation on the study. Do not insist that exactly one word must resolve every comparison.

4. Work a short scientific passage

The following passage and study are invented for practice. It is deliberately shorter than a DAT passage, and its numbers are not research evidence.

Laboratory researchers compared two protective coatings, A and B, on identical mineral disks. They randomly assigned twenty disks to each coating and exposed every disk to the same acidic solution for ten minutes. Mean mass loss was 4 mg for coating A and 3 mg for coating B. The report did not provide a measure of variability or a statistical significance test.

The researchers described B's lower average loss as promising, but emphasized that their experiment tested only one exposure condition. Temperature, exposure duration, and repeated acid cycles could affect the comparison. They proposed testing those conditions before making a general durability claim. Here, “resistance” refers to a coating's ability to limit mass loss during exposure, rather than an organism's response to a drug.

Detail: which result was reported?

• 1. Every B disk lost exactly 3 mg.
• 2. B's mean loss was 3 mg.
• 3. B prevented all mass loss.
• 4. A performed better under repeated exposure.

Answer: B. A mean is not the value of every observation. C conflicts with the reported loss, and repeated exposure was proposed rather than tested.

Inference: what conclusion is warranted?

• 1. B is more durable under every condition.
• 2. The observed difference is statistically significant.
• 3. B had lower average loss in this experiment, but the passage does not establish a universal advantage.
• 4. The coatings must be equally effective.

Answer: C. It combines the observed difference with the stated scope limitation. No variability or significance result was supplied; do not manufacture one.

Purpose: why mention temperature and repeated acid cycles?

• 1. To describe conditions already tested.
• 2. To identify further conditions needed to assess generality.
• 3. To show random assignment failed.
• 4. To recommend clinical treatment.

Answer: B. The paragraph proposes further tests. A changes future work into completed work, and neither C nor D is supported.

Tone: “promising” followed by explicit limitations suggests cautious optimism. The caveats qualify the result; they do not erase it. In another passage, a quoted critic's pessimism would not automatically be the author's tone.

Vocabulary: “resistance” here means limiting mass loss under the tested exposure. The familiar meaning “drug resistance” would fail because the passage explicitly selects a different sense.

Numerical detail: B's mean loss was 1 mg lower, a 25% reduction relative to A's 4 mg: (4−3)/4 = 0.25. It is not a 25% increase in every disk's durability; the metric and scope matter.

5. Evaluate evidence in context

If a passage reports an association, do not infer causation without support. For instance, an association between study time and scores does not by itself exclude differences in prior preparation. Randomization can strengthen causal inference in a suitable experiment, but study design labels alone do not answer every evidence question.

A meta-analysis of poor or irrelevant studies is not automatically stronger than a rigorous, directly relevant experiment. Look at the question being investigated, methods, bias, uncertainty, and applicability. Oxford CEBM explains why evidence levels need judgment.

Separate two tasks: reporting the author's conclusion and evaluating whether the evidence supports a conclusion. If the question asks what the author claims, identify that claim accurately; if it asks what is justified, examine the support. Neither task licenses replacing the passage with unrelated outside facts.

6. Budget the whole section

Twenty minutes per passage is a useful starting allocation, not a fixed rule: passages can have different numbers and difficulty of questions. A concrete 20-minute practice plan is 4 minutes reading/mapping + 15 answering + 1 checking. Across three passages that totals 60 minutes. Adjust the phases together rather than increasing every phase independently.

If you stall, make a provisional choice and flag the item. Return when you have enough time to seek the missing evidence. There is no guessing penalty. For a hypothetical five-choice question, a uniform blind guess has a 20% chance; if exactly two choices remain and one is correct, a uniform guess has 50%, not more than 50%. The actual number of choices on the item determines the arithmetic.

In review, record the question type, the evidence, and the error. If familiar biology passages go worse than unfamiliar topics, importing outside facts is one possible cause—not a diagnosis established by the score pattern. Examine actual responses for overreach, missed qualifiers, vocabulary, inference errors, or timing before choosing a fix.