Critical Reasoning

Module 2.2: Critical Reasoning

What Critical Reasoning Actually Means

Critical reasoning is the ability to evaluate arguments --- to look at a claim and the evidence supporting it and determine whether the conclusion actually follows. It's not about having strong opinions. It's not about being "critical" in the negative sense. It's about thinking clearly, precisely, and logically.

On the STEM Thinking Skills Assessment, critical reasoning questions present you with a scenario, an argument, or a set of facts and then ask you to do something with that information: identify an assumption, evaluate whether a conclusion is supported, find a flaw in reasoning, or determine what additional information would strengthen or weaken a claim.

Students who do well in this domain share a common trait: they slow down and actually analyze what's being said rather than going with what "feels" right. Your gut instinct is useful in a lot of situations, but critical reasoning questions are specifically designed to test whether you can override gut feelings and follow the logic.

Skill 1: Identifying Assumptions

An assumption is something that must be true for an argument to work, but that isn't explicitly stated. Every argument has assumptions --- hidden premises that the arguer takes for granted.

Here's an example:

Argument: "We should plant more trees in the school courtyard because trees improve air quality."

On the surface, this sounds perfectly reasonable. But what's being assumed?

• That the school courtyard currently has an air quality problem worth addressing

• That the number of trees that could fit in a courtyard would make a meaningful difference

• That there aren't other solutions that would be more effective or practical

• That "improving air quality" is a sufficient reason to plant trees (what about cost, maintenance, space?)

None of these assumptions are necessarily wrong. But they're unstated, and if any of them turn out to be false, the argument weakens or collapses entirely.

How to identify assumptions on the test:

Ask yourself: "What has to be true for this conclusion to follow from these premises?" The gap between what's stated and what's concluded --- that gap is filled by assumptions.

Another approach: try to imagine a scenario where the premises are true but the conclusion is false. Whatever you'd have to change to make that happen --- that's an assumption the argument is making.

Practice scenario:

"Students who eat breakfast before school score higher on standardized tests. Therefore, the school should provide free breakfast to improve test scores."

What assumptions are being made?

• That the relationship between breakfast and test scores is causal, not just correlational (maybe students who eat breakfast also tend to have more stable home environments, better sleep, etc.)

• That students would actually eat the breakfast if provided

• That the type of breakfast matters (or doesn't)

• That improving test scores is a priority that justifies the cost

• That the students at this particular school don't already eat breakfast

If the test asks "Which of the following is an assumption underlying this argument?" --- the correct answer will be one of these hidden premises.

Skill 2: Evaluating Strength of Evidence

Not all evidence is created equal. One of the most important critical reasoning skills is the ability to look at a piece of evidence and judge how strongly it supports a given conclusion.

Strong evidence is specific, directly relevant, comes from a reliable source, and is sufficient in scope.

Weak evidence is vague, tangentially related, comes from a questionable source, or is too limited to support a broad conclusion.

Example:

Claim: "Solar panels are a good investment for homeowners."

Evidence A: "My neighbor installed solar panels and said he's happy with them."

Evidence B: "A 10-year study of 5,000 homeowners across 15 states found that solar panel installations yielded an average annual return of 8% through energy savings and tax credits."

Evidence B is dramatically stronger. It's based on a large sample, covers a long time period, spans multiple geographic areas, and provides specific data. Evidence A is a single anecdote from one person in one location.

What to look for on the test:

• Sample size: Was this based on one person's experience or a large study?

• Relevance: Does the evidence actually address the specific claim being made?

• Source: Is the source credible and unbiased?

• Specificity: Does the evidence give concrete data, or is it vague?

• Scope: Does the evidence match the scope of the conclusion? (A study of one school can't support a claim about all schools everywhere.)

Practice scenario:

"A survey of 50 students at Lincoln Middle School found that 80% preferred pizza for lunch. This proves that pizza is the most popular lunch food among American middle schoolers."

What's wrong with this evidence?

The sample is too small (50 students) and too narrow (one school) to support such a broad conclusion. Maybe Lincoln Middle School has particularly good pizza. Maybe the survey was conducted on pizza day. The evidence might support the claim "pizza is popular at Lincoln Middle School," but it can't support a claim about all American middle schoolers.

Skill 3: Drawing Valid Conclusions

A valid conclusion is one that logically follows from the given information --- and only from the given information. This is where the test gets tricky, because several answer choices might seem reasonable, but only one is actually supported by what's stated.

The golden rule: A valid conclusion doesn't require you to add any outside information or make any leaps of logic. Everything you need is in the passage.

Example:

Given facts:

• All members of the robotics club are in 8th grade.

• Sarah is a member of the robotics club.

Valid conclusion: Sarah is in 8th grade.

Invalid conclusion: Sarah is good at building robots. (Being in the club doesn't mean she's good at it --- maybe she just joined.)

Invalid conclusion: Sarah is the best student in her grade. (Nothing in the facts says anything about academic performance.)

This seems obvious with a simple example, but on the test, the scenarios are more complex and the invalid conclusions are designed to sound reasonable.

A more complex example:

Given facts:

• Enrollment at Westfield High decreased by 15% over the past five years.

• During the same period, three new housing developments were built within Westfield's school boundary.

• Westfield's boundary was redrawn two years ago, moving the Oak Ridge neighborhood to a different school zone.

Which conclusion is supported?

A) Westfield is a declining school that families are fleeing.

B) The boundary change may have contributed to the enrollment decrease.

C) The new housing developments haven't attracted families with school-age children.

D) Westfield needs to improve its programs to attract more students.

The answer is B. It's the only conclusion that's directly supported by the given facts --- we know the boundary was redrawn (removing a neighborhood) and enrollment went down. That's a plausible causal link based on the evidence.

A is an emotional interpretation, not a logical conclusion. C contradicts the fact that housing was built (but doesn't address whether those families have kids --- we don't know). D is a recommendation, not a conclusion from the data.

How to approach these questions:

For each answer choice, ask: "Can I point to specific facts in the passage that directly support this?" If you have to add assumptions, outside knowledge, or logical leaps, it's probably not the best answer.

Skill 4: Distinguishing Correlation from Causation

This is one of the most commonly tested concepts in critical reasoning, and it's one that trips up even smart, analytically-minded students.

Correlation means two things tend to happen together.

Causation means one thing actually causes the other.

The classic example: ice cream sales and drowning deaths both increase in the summer. Ice cream sales are correlated with drowning deaths, but ice cream doesn't cause drowning. A third factor --- hot weather --- causes both.

Three ways a correlation can exist without direct causation:

1. Third variable (confounding factor): Something else causes both. Hot weather causes both ice cream sales and swimming (which leads to drowning).

2. Reverse causation: The cause and effect are flipped. "Students who study more get higher grades" seems straightforward, but what if students who get higher grades become more motivated to study? The causation might run in both directions.

3. Coincidence: With enough data points, some things will correlate by pure chance. There's a real statistical correlation between the number of Nicolas Cage movies released in a year and the number of people who drowned in swimming pools. That's obviously coincidence.

Example on the test:

"A study found that students who participate in extracurricular activities have higher GPAs than students who don't. Therefore, participating in extracurricular activities causes students to earn higher grades."

What's the flaw? The study shows correlation, not causation. Possible alternative explanations:

• Students with higher GPAs may be more likely to join extracurriculars (reverse causation)

• Students from families with more resources may both participate in activities and perform better academically (third variable)

• Schools may require minimum GPAs for extracurricular participation, filtering out lower-performing students (selection bias)

How to spot this on the test:

Whenever you see a conclusion that says X causes Y, ask: "Is there another explanation for why X and Y go together?" If you can think of one, the causal conclusion is not fully supported.

Skill 5: Recognizing Logical Fallacies

A logical fallacy is a flaw in reasoning that makes an argument invalid. You don't need to memorize the Latin names of every fallacy, but you should be able to recognize the most common patterns of bad reasoning.

Appeal to popularity (ad populum): "Everyone is doing it, so it must be right."

Example: "Most students at our school think the dress code is too strict, so it must be too strict." The number of people who believe something has no bearing on whether it's true.

False dichotomy (either/or fallacy): Presenting only two options when more exist.

Example: "Either we ban all cell phones in school, or we accept that students will be distracted all day." There are plenty of middle-ground options --- phone pouches, designated phone times, phone-free zones, etc.

Hasty generalization: Drawing a broad conclusion from too little evidence.

Example: "I visited two restaurants in that city and both were bad. That city has terrible food." Two data points can't support a conclusion about an entire city.

Ad hominem (attacking the person): Dismissing an argument by attacking the person making it rather than addressing the argument itself.

Example: "You can't trust Dr. Smith's nutritional advice --- she's overweight." Dr. Smith's personal health has nothing to do with whether her nutritional research is scientifically sound.

Slippery slope: Arguing that one action will inevitably lead to extreme consequences without evidence for the chain of events.

Example: "If we allow students to retake one test, soon they'll expect to retake every test, then they'll stop studying altogether, and eventually no one will learn anything." Each step in this chain is a separate claim that needs its own evidence.

Circular reasoning: Using the conclusion as a premise.

Example: "This is the best school in the county because no other school is as good." That's just restating the conclusion --- it doesn't provide any independent reason to believe the claim.

How to spot fallacies on the test:

Ask: "Does this conclusion actually follow from the reasoning provided, or is there a logical gap?" If the reasoning feels "off" but you can't immediately pinpoint why, check for the patterns above.

Practice Section

Work through these scenarios. For each one, identify the critical reasoning skill being tested and determine the best answer.

Scenario 1:

"A local newspaper reported that test scores at Riverside Elementary improved by 12% after the school adopted a new math curriculum. The school board is now considering adopting the same curriculum at all 15 elementary schools in the district."

Which of the following, if true, would most weaken the case for adopting the curriculum district-wide?

A) The new curriculum costs 30% more than the current one.

B) Riverside Elementary also hired five new math teachers the same year it adopted the new curriculum.

C) Several parents at Riverside Elementary have praised the new curriculum.

D) Test scores at other schools in the district have remained flat.

Analysis: The answer is B. If Riverside also hired new teachers at the same time, we can't tell whether the improvement was caused by the new curriculum or the new teachers (or both). This is a correlation vs. causation question --- there's a confounding variable. A is relevant to the decision but doesn't weaken the evidence that the curriculum works. C is weak anecdotal evidence (and supports the curriculum, not weakens it). D actually strengthens the case slightly (Riverside improved while others didn't).

Scenario 2:

"Dr. Patel argues that students should be required to take a computer science course before graduating high school. However, Dr. Patel owns a software company and would benefit financially if more people learned to code."

Which logical fallacy is being committed here?

Analysis: This is an ad hominem attack. The speaker is dismissing Dr. Patel's argument by pointing to a potential personal motive rather than addressing whether the argument itself has merit. Even if Dr. Patel would benefit financially, that doesn't mean the argument is wrong. The argument should be evaluated on its own logic and evidence, not on who's making it.

Scenario 3:

"In a survey of 200 high school students, those who reported sleeping 8 or more hours per night had an average GPA of 3.6, while those who reported sleeping fewer than 6 hours had an average GPA of 2.8."

Which of the following conclusions is best supported by this data?

A) Sleeping more causes students to earn better grades.

B) There is an association between sleep duration and academic performance among the surveyed students.

C) All students should sleep at least 8 hours per night to improve their grades.

D) Students with lower GPAs should be required to go to bed earlier.

Analysis: The answer is B. The data shows a correlation between sleep duration and GPA among these specific students. A commits the correlation-causation error. C goes beyond the data by prescribing a recommendation to "all students" based on one survey of 200 students. D is a policy recommendation that the data doesn't support --- and it also assumes causation.

Scenario 4:

"All AOS students take advanced math courses. Kevin takes advanced math courses."

Which of the following can be validly concluded?

A) Kevin is an AOS student.

B) Kevin might be an AOS student.

C) Kevin is not an AOS student.

D) Kevin will apply to AOS next year.

Analysis: The answer is B. We know all AOS students take advanced math, but that doesn't mean everyone who takes advanced math is in AOS. Kevin could be an AOS student, or he could be a regular student who chose advanced math. This is a classic logic trap --- just because all A are B doesn't mean all B are A. We can say Kevin "might" be an AOS student, but we can't confirm or deny it.

Scenario 5:

"The city installed bike lanes on Main Street last year. Since then, traffic accidents on Main Street have decreased by 20%. Clearly, the bike lanes have made the street safer."

Identify at least two assumptions in this argument.

Analysis:

Assumption 1: The bike lanes caused the decrease in accidents (and not some other factor, like reduced overall traffic, a new traffic signal, lower speed limits, or seasonal variation).

Assumption 2: The 20% decrease is statistically significant and not due to normal year-to-year fluctuation.

Assumption 3: "Safer" means fewer total accidents (rather than, say, fewer serious accidents --- maybe there are more minor fender-benders but fewer serious crashes, or vice versa).

Assumption 4: The decrease applies to all types of road users, not just one group (maybe car accidents went down but cyclist accidents went up).

Building Your Critical Reasoning Skills

Critical reasoning is not something you learn once and master forever. It's a habit of mind that gets sharper with practice. Here are concrete things you can do to improve before the test:

Read opinion pieces and editorials actively. Don't just read for content --- read to evaluate. After every argument you encounter, ask: What's the conclusion? What evidence supports it? What's assumed but not stated? Is there a logical flaw?

Practice the "what else could explain this?" game. Whenever someone presents a cause-and-effect claim --- in conversation, in the news, in a textbook --- challenge yourself to come up with at least two alternative explanations. This trains your brain to automatically check for confounding variables.

Argue the other side. When you have a strong opinion about something, force yourself to construct the strongest possible argument for the opposite position. This builds the mental flexibility that critical reasoning demands.

Slow down on test questions. The most common critical reasoning mistake is rushing. Read the question stem carefully. Read every answer choice. Eliminate wrong answers before selecting the right one. The correct answer is the one that's logically supported --- not the one that sounds most impressive or feels most intuitively right.