What you'll learn
Every day you hear claims: "This drink boosts energy." "Music helps plants grow." "This paper towel absorbs the most." Science asks one question about every claim: What is the evidence?
In plain words: don't just believe a claim — check what backs it up.
Think of a judge in a courtroom. The judge listens to both sides before deciding. The GED Science test asks you to act like that judge. You weigh the evidence for and against a claim.
In this lesson you will learn to:
- decide whether data support, challenge, or are irrelevant to a claim;
- judge the strength of evidence (design, sample size, controls, blinding, what was actually measured);
- spot overgeneralizations, conclusions that go beyond what was tested;
- identify what additional evidence would best test a claim;
- compare two studies and decide which gives stronger evidence.
How it shows up on the GED: typical questions read:
- "Which evidence best supports the claim?"
- "Which finding would challenge this conclusion?"
- "Why is the conclusion not supported?"
- "Which claim is NOT supported by the data?"
- "What additional information is needed?"
These questions build on investigation design (SP.2) and reasoning from data (SP.3).
Key terms
- claim — a statement someone says is true ("Brand C absorbs the most water").
- evidence — data or observations used to judge a claim (mean absorption of 35 mL).
- conclusion — a judgment drawn from evidence.
- support — evidence that agrees with and strengthens a claim.
- challenge (refute, weaken) — evidence that disagrees with a claim or points to another explanation.
- irrelevant evidence — true information that has nothing to do with the claim. Example: paper-towel color, when the claim is about how much water it soaks up (absorbency).
- overgeneralization — a conclusion that goes beyond the people, conditions or amounts tested. Example: "works for everyone" from a study of adults only.
- scope — the range of conditions a study actually covered (one dose, adults, 30 minutes after drinking).
- observational study — researchers record what already happens, without assigning treatments (a survey).
- controlled experiment — researchers assign treatments and control other variables.
- self-report — data based on what people say about themselves ("I feel very alert").
- bias — anything that unfairly pushes results one way. Examples: a company testing its own product, or volunteers choosing their own group.
- replication — repeating a study to see whether the results hold up.
- anecdote — a single personal story used as evidence ("My uncle drank it and felt great").
1. Support, challenge or irrelevant?
In plain words: first say exactly what the claim says. Then check whether the data agree, disagree, or don't matter.
To evaluate any claim:
- Restate the claim exactly, including its limits ("of the brands we tested", "alone", "on average").
- Find the data that address it.
- Decide: do the data agree (support), disagree or offer another explanation (challenge), or not address it (irrelevant)?
Example: Claim: "All metals are attracted to magnets."
| Metal | Attracted to magnet? |
|---|---|
| Iron | Yes |
| Nickel | Yes |
| Copper | No |
| Aluminum | No |
Iron and nickel agree with the claim, but copper and aluminum challenge it. One clear counterexample is enough to challenge a claim with all in it.
Absolute words matter. Claims with all, every, only, always, never, alone need very strong evidence. A single exception can challenge them. If a friend says "this bus is never late," one late bus proves them wrong. Claims with some, most, tends to, on average are easier to support.
2. How strong is the evidence?
In plain words: some evidence is solid, and some is shaky.
Think about buying a used car. A mechanic's inspection tells you more than the seller saying "it runs great." Not all evidence is equal. Ask these questions:
| Question | Stronger evidence | Weaker evidence |
|---|---|---|
| What was measured? | exactly what the claim is about (reaction time) | something related (feeling alert) |
| How was it measured? | instruments, objective tests | opinions, self-report, anecdotes |
| How many? | large sample, repeated trials | a few subjects, one trial |
| Was there a control? | control group under same conditions | no comparison group |
| Were groups fair? | random assignment | self-selected groups |
| Was it blind? | participants and testers do not know groups | everyone knows who got what |
| Who did it? | independent researchers, replicated | one study by someone who profits from the result |
Example: compare two studies. Both seem to favor caffeine.
- Study 1: a double-blind experiment with 200 randomly assigned adults. It measures reaction time.
- Study 2: an online survey of 50 people about how alert they feel.
Study 1 is much stronger.
3. Staying within the scope
In plain words: a study only tells you about what it actually tested.
If a jacket kept you warm at 40 °F, that does not prove it will keep you warm at 0 °F. A conclusion is supported only for the conditions that were tested. Watch for these overgeneralizations:
- People/organisms: a study of adult volunteers → a claim about children or about "everyone."
- Amounts: one dose tested (200 mg) → a claim about higher doses.
- Outcomes: reaction time measured → a claim about memory or health.
- Averages vs. individuals: a group mean improved → "every person improves."
- Time: effects measured after 30 minutes → a claim about all day.
Example:
| Statement | |
|---|---|
| Study result | "On average, 200 mg of caffeine shortened reaction time by 20 ms in adults." |
| Supported | "In adults, 200 mg of caffeine improved average reaction time." |
| Unsupported | "Caffeine makes everyone faster." "More caffeine is always better." |
4. What additional evidence is needed?
In plain words: the best new evidence fixes what is missing from the old evidence.
GED questions often ask which additional data would best test, support or challenge a claim. The best answer:
- matches the claim exactly. A claim about dose needs several doses. A "wolves alone" claim is challenged by evidence of a different cause;
- fixes a weakness of the existing evidence (adds a control, removes a confound, enlarges the sample);
- produces measurable data, not opinions.
To support a claim, look for evidence that rules out other explanations. To challenge a claim, look for one of three things:
- a counterexample;
- another possible cause;
- a better-designed study with a different result.
Example: Claim: "Music makes plants grow taller." Evidence so far: plants with music sat in a sunny kitchen. Silent plants sat on a shaded porch. The music plants grew more. Best additional evidence: many plants in the same room, same light and water, half with music. That is the only design that isolates music.
5. Comparing studies and changing conclusions
In plain words: trust the better-built study, and be ready to change your mind when new evidence comes in.
Sometimes two sources disagree or differ in quality. Give more weight to the one with the stronger design. Also favor the one that most directly measures what the claim is about. A large, controlled, randomized study outweighs a small survey, even if the survey has a dramatic result.
Science is self-correcting. Conclusions can change when better evidence appears. A conclusion that once seemed supported may be challenged later. A new study with a larger sample or better controls can do this. On the GED, you may be asked how a new finding affects an earlier conclusion. Does it strengthen it, weaken it, or not affect it?
Worked example 1: Is the claim supported?
Stimulus: "A gym claims its 8-week program lowers resting heart rate. Thirty members joined. Their average resting heart rate was 76 beats per minute before and 70 after. There was no comparison group."
Question: Which additional evidence would most strengthen the gym's claim? A. Testimonials from members who enjoyed the program B. Data from a similar group of people who did not do the program during the same 8 weeks, showing no change in heart rate C. The program's price D. A photo of members exercising
Think-aloud:
- The data show heart rates fell, but without a comparison group, we cannot rule out other causes (seasonal change, members also changing diets).
- B adds a control group. If non-members did not change, the program becomes the most likely cause.
- A and D are anecdotes and pictures; C is irrelevant.
Answer: B.
Worked example 2: Evidence that challenges
Stimulus: A student claims, "Heavier objects always fall faster than lighter objects." Her data: in air, a feather fell more slowly than a baseball.
Question: Which observation would most strongly challenge her claim? A. A bowling ball falls faster than a feather in air. B. In a vacuum chamber with no air, a feather and a bowling ball dropped together land at the same time. C. A baseball is heavier than a feather. D. Feathers are used in pillows.
Think-aloud:
- The claim uses always, so one exception challenges it.
- B shows equal falling when air resistance is removed. The weight difference did not cause faster falling; air did.
- A agrees with her claim. C is a fact without a test, and D is irrelevant.
Answer: B.
Worked example 3: Spotting an overgeneralization
Stimulus: "Researchers gave 300 adult men a new allergy medicine or a placebo, randomly assigned and double-blind. Sneezing decreased by 40% on average in the medicine group compared with the placebo group."
Question: Which conclusion is best supported? A. The medicine reduces sneezing in all people, including children. B. On average, the medicine reduced sneezing in adult men compared with a placebo. C. The medicine cures allergies permanently. D. The medicine reduced sneezing in every man who took it.
Think-aloud:
- Scope: adult men, one medicine, sneezing measured, results are averages.
- A extends to children and women (not tested). C claims a cure (not measured). D ignores that 40% is an average.
- B stays within the scope.
Answer: B.
Worked example 4: Comparing two sources
Stimulus: Source 1: a blog post where the author says a new fertilizer doubled her tomato harvest. Source 2: a university study that grew 400 tomato plants in identical greenhouse conditions, half with the new fertilizer; harvests differed by 3%, which the researchers said could be due to chance.
Question: Which statement best evaluates the claim that the fertilizer doubles harvests? A. The claim is supported because the blogger saw it happen. B. The claim is weakened because the larger, controlled study found little difference. C. Both sources are equally reliable. D. The claim is proven because two sources discussed it.
Think-aloud:
- Source 1 is an anecdote: one garden, no control, possibly biased.
- Source 2 is a large controlled experiment that measured harvests directly.
- The stronger evidence shows no large effect, which challenges the claim.
Answer: B.
Common mistakes to avoid
- Accepting evidence that's only related. Wrong: "People feel more alert, so reaction time improves." Right: only a direct measurement of reaction time tests that claim.
- Confusing support and challenge. Wrong: choosing "wolves hunted deer" to challenge "wolves alone caused the decline." Right: a challenge points to another cause, such as a harsh winter.
- Overgeneralizing from averages. Wrong: "Every person reacted faster." Right: "On average, the caffeine group reacted faster."
- Going beyond the scope. Wrong: "Higher doses work even better" from a one-dose study. Right: "Only 200 mg was tested."
- Trusting anecdotes and opinions. Wrong: "Gardeners say music works, so it does." Right: opinions are not controlled measurements.
- Ignoring the claim's exact wording. Wrong: "Not enough brands were tested" when the claim is about "the brands we tested." Right: judge the claim on its own terms.
GED test strategy
- Underline the claim and circle its key limits and absolute words (all, only, alone, best, always).
- Point to the number. For "supported" choices, find the exact data value that backs them up. If you cannot point to it, it is probably not supported.
- For "NOT supported" questions, look for the choice that brings in something never measured (price, safety, taste, memory, long-term effects).
- For "additional evidence" questions, ask: "Does this new data change only the variable in the claim, and does it produce measurements?"
- For comparing studies, rank them by design, strongest first:
- randomized, controlled, blinded experiment
- controlled experiment
- observational study
- survey or self-report
- anecdote
- Eliminate extremes. Choices that say proves or every or that predict untested conditions are rarely correct.
- Time: these questions take careful reading. Spend about 1.5 minutes, and reread the claim before choosing.
Check yourself
Testing Paper Towels
A paper-towel company advertises: "Brand C absorbs the most water of the brands we tested." A consumer group decided to check the claim. They tested three brands. For each trial, they dipped one sheet (all sheets the same size) into water for 30 seconds, let it drip for 10 seconds, and then squeezed the sheet into a measuring cylinder to find how much water it had held. They ran three trials for each brand.
| Brand | Trial 1 (mL) | Trial 2 (mL) | Trial 3 (mL) | Mean (mL) |
|---|---|---|---|---|
| A | 22 | 24 | 23 | 23 |
| B | 30 | 29 | 31 | 30 |
| C | 35 | 36 | 34 | 35 |
Do the consumer group's results support the company's claim that Brand C absorbs the most water of the brands tested?
Hint
Compare the means in the last column, and check the individual trials too.
Step-by-step solution
Compare the claim with the data. The claim is that "Brand C absorbs the most water of the brands we tested." The Mean (mL) column shows A = 23, B = 30 and C = 35.
Step by step:
- Brand C has the highest mean (35 mL), higher than B (30 mL) and A (23 mL).
- Check the trials too: Brand C's lowest trial (34 mL) is still higher than Brand B's highest trial (31 mL). So C absorbed the most in every trial.
- Choice B's reason is false, since C was highest in every trial. Choice C misreads the claim: it is only about "the brands we tested", so testing three brands is enough. Choice D's reason (strength) was never measured.
The correct answer is A.
Tip: A claim is supported when the data directly match it. Read the exact wording of the claim. Its limits (here, "of the brands we tested") tell you how much evidence it needs.
Check yourself
Testing Paper Towels
A paper-towel company advertises: "Brand C absorbs the most water of the brands we tested." A consumer group decided to check the claim. They tested three brands. For each trial, they dipped one sheet (all sheets the same size) into water for 30 seconds, let it drip for 10 seconds, and then squeezed the sheet into a measuring cylinder to find how much water it had held. They ran three trials for each brand.
| Brand | Trial 1 (mL) | Trial 2 (mL) | Trial 3 (mL) | Mean (mL) |
|---|---|---|---|---|
| A | 22 | 24 | 23 | 23 |
| B | 30 | 29 | 31 | 30 |
| C | 35 | 36 | 34 | 35 |
Which claim is NOT supported by the consumer group's data?
Hint
Three of these statements can be checked with numbers in the table. Which one needs information the test did not collect?
Step-by-step solution
Check each statement against the table. Only one needs information that is not there.
Step by step:
- A: Brand A's mean is 23 mL, the lowest. Supported.
- B: Brand B's mean (30) minus Brand A's mean (23) = 7 mL. Supported.
- C: Each brand's trials are close together: A 22–24, B 29–31, C 34–36, all within 2 mL. Supported.
- D: "Best value for the money" requires prices and possibly other qualities, and the test measured neither. Not supported.
The correct answer is D.
Tip: On "NOT supported" questions, the right answer usually brings in a new idea (price, strength, safety, taste) that was never measured. Supported statements can be checked with numbers you can point to.
Summary
- To evaluate a claim, restate it exactly, find the data that address it, and decide: support, challenge or irrelevant.
- Strong evidence measures exactly what the claim is about, with large samples, controls, random assignment and blinding.
- Supported conclusions stay within the study's scope: who was tested, what amount, what outcome, and averages rather than every individual.
- The best additional evidence matches the claim and fixes a weakness. A challenge often shows a counterexample or another cause.
- When studies differ, trust the better-designed study that measured the claim directly.