MCAT

1. Turn an unfamiliar passage into a causal chain

Learning goal: Use the stated model before adding outside facts.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

The MCAT science sections test scientific reasoning as well as knowledge; the AAMC reasoning framework explicitly includes applying models to unfamiliar situations [M6]. A useful passage map has a manipulated variable, an intermediate process and a measured outcome. Write only the links the passage supplies. A familiar word can tempt you to import a pathway that the question never established.

In our toy model, an inhibitor reduces enzyme activity, enzyme activity raises product concentration, and product activates a receptor. If all other variables are held fixed, less activity predicts less product and weaker receptor activation. This is conditional reasoning, not a universal biological law. If a later paragraph says receptor activation increases enzyme activity, add that feedback only after reading it; do not silently turn a one-way model into a feedback loop.

Worked example

A passage states: A raises B, B lowers C. Compound X lowers A. Predict C, with other factors fixed.

1. Record A → higher B, and B → lower C.
2. X lowers A, so B is predicted to fall.
3. Less B removes some suppression of C.
4. Therefore C is predicted to rise within this stated model.

Mental model: Two negative links make a positive net prediction.

Common trap: Do not attach an outcome the passage never connects.

2. Conserve the solute

Learning goal: Compute concentration from amount and total final volume.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

Dilution changes the amount per unit volume, not the conserved amount of solute, if no reaction or material loss occurs. Use amount = concentration × initial volume, then divide by total final volume. For these examples assume additive liquid volumes. A millimolar concentration is millimoles per litre, which is also micromoles per millilitre; the unit conversion is a check, not an extra numerical factor.

Adding solvent and diluting to a marked final volume are different instructions. "Add 6 mL" to 2 mL gives 8 mL; "dilute to 6 mL" gives 6 mL. The passage wording sets the denominator. This is an editorial practice target within AAMC chemical/physical problem-solving [M2, M6], not evidence about how many candidates miss dilution.

Worked example

Add 6.0 mL solvent to 2.0 mL of 12.0 mM solution. Find final concentration.

1. Amount = 12.0 × 2.0 = 24.0 micromoles.
2. Final volume = 2.0 + 6.0 = 8.0 mL.
3. Final concentration = 24.0/8.0 = 3.0 mM.
4. Check: a fourfold volume increase produces a fourfold concentration decrease.

Mental model: Concentration × initial volume = concentration × final volume.

Common trap: The added volume is not the final denominator.

3. Powers of ten with a physical check

Learning goal: Use units and proportionality to catch arithmetic errors.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

Time pressure makes scientific notation useful, but the exponent must carry units correctly. A toy constant-speed model has distance = speed × time. Converting milliseconds to seconds changes the time factor by 10^-3. If a result implies thousands of metres in a few milliseconds at ordinary laboratory speeds, check the conversion before accepting it.

Proportional reasoning also lets you predict direction before calculating: doubling speed at unchanged duration doubles distance. AAMC includes quantitative scientific problem-solving in its reasoning framework [M6]. These activities isolate that skill; they do not assert that every passage uses constant speed or that real tissue motion has no acceleration.

Worked example

A marker travels at 2.0 m/s for 5.0 ms at constant speed. Distance in metres?

1. Convert time: 5.0 ms = 5.0 × 10^-3 s.
2. Multiply: 2.0 m/s × 5.0 × 10^-3 s.
3. Seconds cancel, leaving 1.0 × 10^-2 m.
4. This is 1.0 cm, a plausible short displacement.

Mental model: Keep the exponent and the unit together.

Common trap: A correct formula with incompatible units still gives a wrong answer.

4. A control is a comparison, not a decoration

Learning goal: Separate what changed from what was measured.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

AAMC explicitly tests research design and execution [M6]. An independent variable is what the investigator changes; a dependent variable is what is measured. A control makes a comparison interpretable by matching other conditions. In a toy enzyme experiment, changing both inhibitor dose and temperature would make a difference in product ambiguous.

Random allocation reduces systematic group differences in expectation; it does not guarantee identical participants or remove every source of bias. Blinding addresses knowledge-related measurement or behaviour effects. These are different design tools. When a question asks for the best control, choose the comparison that isolates the stated manipulation rather than the group that simply produces the lowest response.

Worked example

Cells receive drug dissolved in solvent. Treated cells show less growth. Choose a control for the drug effect.

1. Identify active drug as the intended independent variable.
2. Recognize that solvent is present only because of delivery.
3. Give control cells the same solvent, culture duration and conditions without drug.
4. Compare growth; the active compound is the intended remaining difference.

Mental model: Name the manipulation and measurement before judging a control.

Common trap: A control is not simply an untreated group under different conditions.

5. Read the axis before the result

Learning goal: Distinguish totals, rates and normalized comparisons.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

A plot is an argument whose axis labels are part of the premises. Counts, percentages and per-cell measurements answer different questions. If sample A has 100 units of product across 10 cells and B has 160 across 20 cells, B has the larger total but A has the larger amount per cell. You must match the comparison to the requested denominator.

For AAMC data reasoning [M6], first describe what was measured, then the group ordering, then only the inference the design supports. Error bars without a definition do not establish a particular confidence level or significance test. All bars in this activity are explicitly toy teaching data, not observed biology, exam performance or a treatment claim.

Worked example

Toy cultures: A has 100 product units and 10 cells; B has 160 units and 20 cells. Which has more product per cell?

1. A: 100/10 = 10 units/cell.
2. B: 160/20 = 8 units/cell.
3. Compare 10 with 8, not 100 with 160.
4. A has greater per-cell product; B has greater total product.

Mental model: Read labels before ranking bars.

Common trap: Do not attach a significance claim to undefined error bars.

6. Relative risk is not a percentage-point difference

Learning goal: Keep risk ratios and risk differences separate.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

A risk is the event count divided by the group size. Relative risk compares two risks by division; absolute risk reduction subtracts treated risk from control risk when the treated risk is smaller. A change from 30% to 15% is 15 percentage points but a 50% relative decrease. Saying only "15% reduction" leaves the quantity ambiguous.

This activity is a hypothetical randomized-trial arithmetic exercise for AAMC data interpretation [M6], not evidence about an actual drug. Even a correctly calculated risk ratio does not describe harms, duration, uncertainty or the population to which results apply. Report the denominator and comparison direction before translating numbers into words.

Worked example

Toy trial: 12/80 treated and 24/80 control participants have the event. Find RR and absolute reduction.

1. Treated risk = 12/80 = 0.15.
2. Control risk = 24/80 = 0.30.
3. Treated/control RR = 0.15/0.30 = 0.50.
4. Absolute reduction = 0.30 - 0.15 = 0.15, or 15 percentage points; relative reduction is 50%.

Mental model: Ratios divide; absolute differences subtract.

Common trap: Always specify which group is the reference.

7. Association does not close the causal argument

Learning goal: Identify a plausible common cause or selection effect.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

Science passages may report observational relationships. A link between attending a programme and better outcomes could reflect programme effects, prior motivation, resources or selection into attendance. The measured association does not by itself select a causal explanation. AAMC research-design reasoning includes evaluating alternative explanations [M6].

Use the question stem to decide the task: an alternative explanation weakens a causal interpretation, whereas a better-controlled design makes that interpretation more credible. This is not a rule that observational evidence is useless or that randomization proves every generalization. Specify the causal claim and show how the alternative can produce the same pattern without it.

Worked example

Toy study: students choosing tutoring score higher later. Give an alternative to tutoring causing the entire gap.

1. The claim is tutoring → higher later score.
2. Students self-select, so groups may differ before tutoring.
3. Motivation could increase both attendance and later study.
4. The same association can occur without tutoring explaining the entire gap.

Mental model: Show how the alternative produces the observed association.

Common trap: Naming a variable without a causal route is not an explanation.

8. CARS: the author is not every voice in the passage

Learning goal: Separate quoted viewpoints from the central claim.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

AAMC CARS includes comprehension and reasoning within and beyond a text, without requiring subject-specific knowledge [M3]. A paragraph may quote a critic only to qualify or reject that critic. Track who holds each view and what the author does with it. The topic names what the passage discusses; the main claim tells what the author argues about it.

Use connective words as evidence, but read their scope. "However" may introduce a narrow reservation rather than a total reversal. In our original mini-passage, an author praises archives for preserving evidence but argues that access, not volume alone, determines public value. The main claim must include that qualification; an answer that archives are useless is too strong.

Worked example

Mini-passage: "Archives preserve fragile evidence. Yet a larger archive helps little if the public cannot find its contents. Funding access is therefore as important as collecting." Identify the central claim.

1. The first sentence establishes a benefit.
2. "Yet" introduces a limit: inaccessible collections may have little public value.
3. "Therefore" signals the policy conclusion.
4. The author argues that access deserves attention alongside collection, not that collection should stop.

Mental model: Map voices before picking a thesis.

Common trap: Outside history of the subject cannot replace textual evidence.

9. CARS: extend the logic, not the topic

Learning goal: Choose a warranted application without adding premises.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

Reasoning beyond the text in CARS applies passage ideas to a new context [M3]. It does not license any statement about the same topic. Restate the passage principle with its condition, then test the new case. If an author says a collection has public value only when accessible, a digitized archive behind an unusable interface does not automatically satisfy the access condition.

An inference can be modest and still be correct. "May improve access" is different from "guarantees understanding." Preserve qualifiers such as some, often and only when. This editorial exercise uses short original passages to isolate transfer; it is not an official CARS set or a claim that passage reading can be reduced to keyword matching.

Worked example

Principle: a public dataset is useful only when users can interpret its labels. A city releases a larger unlabeled dataset. What can be concluded?

1. Useful → interpretable labels is a necessary condition.
2. The new dataset is larger, but labels remain uninterpretable.
3. Size does not establish the necessary condition.
4. The passage principle does not support claiming that larger release is useful.

Mental model: Apply the principle with its conditions intact.

Common trap: A familiar scenario can still be an unsupported answer.

10. Operational definitions and a sustainable plan

Learning goal: Translate labels into measurable behaviours and respect the real exam format.

Practice focus: editorial, not a measured error-frequency claim. This lesson uses original examples.

The psychological/social section tests applying concepts to scenarios [M5]. An operational definition states how a construct is measured in a particular study. "Engagement" could mean attendance, completed tasks or self-report; these are not interchangeable. A study can count behaviour accurately while still measuring only one part of the construct.

The official MCAT order is Chemical/Physical, CARS, Biological/Biochemical, Psychological/Social; content times are 95, 90, 95 and 95 minutes [M1]. These sum to 375 minutes, not the entire seated day. Use this module for reasoning drills, not a scaled-score predictor. Section scores are 118-132, with form-specific conversion [M7]; no homemade raw-score table is supplied. Personal notes and electronics are restricted [M11].

Worked example

A toy study defines engagement as tasks submitted out of ten. Student A submits eight and B submits six. Compute their measured engagement and state a limit.

1. A = 8/10 = 80%; B = 6/10 = 60%.
2. The measured gap is 20 percentage points.
3. This establishes a difference in task submission under this definition.
4. It does not by itself establish a difference in motivation or learning quality.

Mental model: Name the measurement before interpreting the construct.

Common trap: A practice percentage cannot be converted into an official scaled score.

11. Enzyme rates need a controlled comparison

Learning goal: Use rate measurements to distinguish observations from a proposed mechanism.

A rate is a change per unit time. Before interpreting an enzyme experiment, identify what was measured, what was manipulated, and which variables were held constant. Substrate concentration, enzyme amount, temperature and pH can all matter to the model. Changing several inputs at once makes it harder to attribute a difference to one cause. AAMC includes scientific inquiry and reasoning in the science sections.

In an original experiment, identical enzyme samples produce ten units of product in five minutes under condition A and fifteen units in five minutes under condition B. The measured average rates are two and three units per minute. Condition B has a higher observed rate under these conditions. That observation alone does not identify the molecular reason for the change.

If temperature is the intended manipulation but enzyme concentration also differs, either difference may contribute. A control comparison should match the relevant conditions except the variable being tested. Repeated measurements help characterize variation but do not repair a systematically confounded comparison. Do not label a higher rate as higher affinity without the additional evidence required for that model.

Practice set: twelve product units form in four minutes and twenty form in five minutes. Calculate average rates three and four. Then compare ten units in two minutes with twenty in four minutes: the rates are equal despite different totals. In an exam passage, use the stated model and axes rather than importing an unstated enzyme mechanism. The numbers are invented for practice and do not describe a particular real enzyme.

Source alignment: [M1/M6]. All numerical data and practice questions here are original teaching examples, not official exam questions or observed results.

Worked example

12 product units form in 4 minutes. Calculate average rate.

1. Define rate as product/time.
2. Use 12 product units and 4 minutes.
3. 12/4=3.
4. The units are product units per minute.

Mental model: Use rate measurements to distinguish observations from a proposed mechanism.

Common trap: Compare product totals without the elapsed-time denominator.

12. Energy bookkeeping uses the stated boundary

Learning goal: Apply a supplied conservation model before interpreting biological efficiency.

A conservation question needs a system boundary. Energy entering a system can appear as useful output, stored energy or other transfers such as heat. In a simplified model with no change in storage, input equals the sum of outputs. The word lost usually means transferred outside the useful-output category, not destroyed. Read the passage definition of useful work before calculating an efficiency.

Original model: a device receives one hundred joules, delivers thirty joules as defined useful work and transfers seventy joules as heat. Its useful-work efficiency is thirty percent. The seventy joules have not ceased to exist; they belong to another output pathway in this model. If storage changes, include that term rather than forcing the same equation.

Biological and physical passages may use different boundaries and conventions. A whole-organism energy budget differs from one reaction step or one piece of equipment. A statement that a reaction requires an energy input does not by itself specify how every surrounding process is coupled. Apply only the relationships the problem provides, and state the missing assumptions when interpreting a result.

Practice set: two hundred joules enter, fifty become useful output and storage does not change. Other outputs total 150 and useful efficiency is twenty-five percent. Next, one hundred enter, thirty are useful and twenty are stored; the remaining transfers sum to fifty. These are bookkeeping models, not biological measurements. AAMC science reasoning asks for integration of knowledge with evidence; dimensional consistency and a clear boundary make that integration testable.

Source alignment: [M1/M6]. All numerical data and practice questions here are original teaching examples, not official exam questions or observed results.

Worked example

Input 200 J, useful output 50 J, no storage change. Find efficiency.

1. Define efficiency as useful/input.
2. 50/200=0.25.
3. Convert to 25%.
4. Other outputs account for 150 J.

Mental model: Apply a supplied conservation model before interpreting biological efficiency.

Common trap: Treat energy outside useful work as destroyed.

13. Genetic probability names the inheritance model

Learning goal: Calculate a probability only after specifying genotype and independence assumptions.

A genetics calculation depends on an inheritance model. In a simple single-locus diploid model with two alleles A and a, each parent contributes one allele. An Aa parent has a one-half chance of transmitting A and a one-half chance of transmitting a under the stated Mendelian model. These probabilities are assumptions of the teaching example, not claims about every biological inheritance system.

Crossing two Aa parents gives four equally weighted combinations: AA, Aa, aA and aa. If A is completely dominant for the trait in the example, three combinations have the dominant phenotype and one has the recessive phenotype. Genotype and phenotype are different: the two heterozygous combinations count separately in the probability tree even though they have the same genotype label after order is ignored.

A passage may change the assumptions through incomplete dominance, linkage, penetrance or a different inheritance pattern. Use the supplied model instead of forcing a familiar ratio. For independent births in the simple model, the probability of two recessive-genotype outcomes is one-quarter times one-quarter, or one-sixteenth. Independence belongs in the explanation, not only the multiplication.

Practice set: Aa crossed with aa gives a one-half chance of aa. Two independent offspring both aa have probability one-quarter. Now ask for at least one aa: use one minus the probability neither is aa, yielding three-quarters. The event both is different from at least one. AAMC biological science scope provides the exam context; the small probability models are original exercises and not real family-risk counseling.

Source alignment: [M1/M6]. All numerical data and practice questions here are original teaching examples, not official exam questions or observed results.

Worked example

Aa×Aa in the stated simple model. Probability of aa?

1. Each parent contributes a with probability 1/2.
2. Assume independent parental transmissions.
3. Multiply 1/2 by 1/2.
4. Probability is 1/4.

Mental model: Calculate a probability only after specifying genotype and independence assumptions.

Common trap: Use a memorized phenotype ratio without checking the stated model.

14. Diagnostic evidence needs the right conditional

Learning goal: Distinguish sensitivity from the probability of disease after a positive result.

A conditional probability has a specified reference group. Sensitivity describes positive results among people with the condition. A positive predictive proportion describes people with the condition among those who test positive. These reverse the conditioning and are not generally equal. A two-by-two table makes the denominator visible and avoids relying on an ambiguous phrase such as test accuracy.

In an original sample, ten people have the condition. Eight test positive and two negative. Ninety people do not have the condition; nine test positive and eighty-one negative. Sensitivity is eight divided by ten, or eighty percent. Among all seventeen positives, eight have the condition, so the positive predictive proportion is eight-seventeenths, about forty-seven percent. The numerical difference comes from different denominators.

The example does not describe a real clinical test or establish what any individual should do. Its purpose is to show how base rates and errors combine. Specificity uses true negatives among those without the condition: eighty-one divided by ninety, or ninety percent. A statement about one conditional should not be silently converted into another without prevalence or counts.

Practice set: of twenty with a condition, sixteen are positive. Of eighty without it, four are positive. Sensitivity remains eighty percent, but now sixteen of twenty positives have the condition, giving eighty percent positive predictive proportion in this particular sample. Equality here is a feature of the chosen numbers, not a universal identity. AAMC research reasoning includes reading and interpreting data; explain each denominator before drawing a health-related conclusion.

Source alignment: [M1/M6]. All numerical data and practice questions here are original teaching examples, not official exam questions or observed results.

Worked example

8 true positives,9 false positives. What fraction of positive results are true positives?

1. Define the positive-result group.
2. Its size is8+9=17.
3. True positives number8.
4. Fraction is8/17.

Mental model: Distinguish sensitivity from the probability of disease after a positive result.

Common trap: Reverse a conditional probability without changing its denominator.

15. Psychology studies need an operational measure

Learning goal: Check whether the measured variable represents the concept the claim names.

A theoretical concept and its operational measure are different. A study may discuss stress but measure questionnaire responses, heart rate or task performance. The operational definition says how the variable was recorded. Evaluating the conclusion requires asking whether that measure fits the intended concept and whether alternative influences on the measurement were controlled.

An original study defines concentration as the number of correct responses on a five-minute task. A group completes more correct responses after a break. The observation concerns performance on this task under the study conditions. It does not automatically establish improved concentration in every setting, or prove that the break is the only cause. Prior familiarity, selection and changed instructions may affect performance.

Reliability concerns consistency of measurement; validity concerns whether the interpretation fits the intended construct. A device that gives the same wrong reading can be consistent without measuring the target accurately. Random assignment, where appropriate, addresses allocation of conditions rather than guaranteeing that the outcome measure is valid. Keep these concepts separate when reading an experimental design.

Practice set: a researcher uses hours of attendance as a measure of learning. Attendance may be reliably recorded but does not directly show understanding. Suggest a task testing knowledge or application and note that it too has limitations. Then compare a self-report of fatigue with a measured reaction time; neither alone defines all aspects of fatigue. AAMC psychological and social section provides the scope; the scenarios are invented and do not make empirical claims about breaks or learning.

Source alignment: [M1/M6]. All numerical data and practice questions here are original teaching examples, not official exam questions or observed results.

Worked example

A study calls attendance hours a measure of understanding. Identify the issue.

1. Name the target:understanding.
2. Name the recorded variable:attendance.
3. Attendance does not directly demonstrate knowledge.
4. Construct validity needs support.

Mental model: Check whether the measured variable represents the concept the claim names.

Common trap: Treat a convenient measure as identical to the concept.

16. Mixed science practice checks units and claims

Learning goal: Solve a compact set, then limit each conclusion to the stated evidence.

An integrated science problem may mix calculation with interpretation. Separate the two tasks. A correctly calculated rate does not identify the cause of a rate change. A conditional probability does not become its reverse. A measured performance difference does not establish an unlimited claim about a psychological construct. This original set practices these boundaries rather than reproducing official MCAT timing.

First, twelve units of product form in four minutes. Find average production rate. Second, a sample contains sixteen true-positive and four false-positive results. Find the proportion of positives that are true. Third, a study records higher scores after a break but also changes the task to an easier one. Is the break uniquely identified as the cause? Try each with a short explanation.

The rate is three units per minute. The predictive proportion is sixteen over twenty, or eighty percent. The causal answer is no because task difficulty changed too. A number can be exact within the example while a causal inference remains unresolved. State the numerator, denominator or competing explanation in the answer, rather than simply writing a result.

For a fresh pass, use twenty product units over five minutes, eight true positives with twelve false positives, and identical task difficulty but no random allocation. The first two answers are four and forty percent. The third still does not by itself prove causation; holding one factor constant does not remove every selection or design concern. Record your error as units, conditioning, model or inference. Use AAMC resources separately for the full official content framework and interface. No practice-set result here predicts an exam score or a clinical outcome.

Source alignment: [M1/M6]. All numerical data and practice questions here are original teaching examples, not official exam questions or observed results.

Worked example

16 true positives and4 false positives. Find true-positive fraction among positives.

1. Use all positive results as denominator.
2. 16+4=20.
3. 16/20=0.8.
4. That is80% of positive results.

Mental model: Solve a compact set, then limit each conclusion to the stated evidence.

Common trap: Treat a numerical answer as permission for a broader causal claim.