2 Evaluating Evidence and Reasoning

How arguments connect claims and evidence

Arguments connect claims—the statements offered for acceptance—to evidence and reasoning. To evaluate an argument, identify its main claim, the reasons offered in support, and any assumptions that link those reasons to the conclusion. Then assess whether the evidence is trustworthy, relevant, and sufficient.

Assessing evidence quality

Examine evidence from several related angles. Consider who produced it, what expertise or method supports it, whether its facts and references can be checked, and whether it is represented in its original context. The source's purpose and possible bias also matter; no single sign, such as a professional-looking website, guarantees accuracy.

Ask whether the evidence directly bears on the claim. A reliable source can still offer evidence that is irrelevant to a particular question, and a relevant detail can still be unreliable.

Assess whether there is enough evidence for the claim's scope and certainty. Consider sample size, representativeness, methods, limitations, and relevant counterevidence. A broad claim generally needs broader and stronger support than a narrow one.

Check currency and context: determine whether the information is recent enough for the topic, and what was measured, when and where it was measured, and how key terms were defined. Finally, ask whether the argument explains how the evidence supports the claim or leaves an important assumption unstated. These qualities work together: evidence can be accurate but irrelevant, relevant but too limited, or plentiful but collected using a weak method.

For example, a survey of 20 students in one class finds that 16 prefer later school start times. This supports the limited claim that most surveyed students in that class prefer a later start. By itself, it does not establish that most students across the whole school prefer later starts, or that later starts improve academic performance: the survey may not represent the whole school, and it measures preference rather than academic outcomes.

Comparing sources

When sources disagree, compare what each actually claims and what evidence it uses. Examine the authors, methods, dates, definitions, populations, and purposes of the sources. Look for independent sources that corroborate a finding, but do not treat the number of sources as decisive when they all repeat the same original report.

A careful evaluation notes genuine limitations or disagreement. Match the conclusion to the strength of the available evidence rather than choosing only sources that confirm an existing view.

Interpreting charts and written information

Before interpreting a chart, read its title, labels, units, dates, source, and notes. Check what its values count and whether they are totals, percentages, rates, or estimates. Inspect the scale: a bar chart whose vertical axis begins above zero can make small differences look dramatic, and different scales can distort comparisons.

Describe what the data show, then distinguish that observation from an explanation of why it happened. A pattern or correlation alone does not prove causation.

For written material, separate stated facts, interpretations, and opinions. Check whether the conclusion follows from the evidence, whether important qualifications or counterarguments have been omitted, and whether the wording expresses more certainty than the evidence warrants.

Common reasoning errors

A reasoning error weakens the connection between evidence and conclusion. An unsupported claim is one for which adequate evidence or reasoning has not been provided; it may be true or false, but the argument has not established it.

Common errors include:

• Hasty generalization: Drawing a broad conclusion from too few or unrepresentative examples.

• Post hoc reasoning: Treating the fact that one event followed another as proof that the first caused the second. Establishing a cause requires additional evidence and consideration of other explanations.

• False dilemma: Presenting two choices as the only possibilities when other options exist.

• Straw man: Misrepresenting a position to make it easier to criticize.

• Ad hominem: Attacking the person making an argument instead of assessing its evidence and reasoning.

• Red herring: Shifting attention to a point that does not answer the central claim.

• Circular reasoning: Restating a claim as if it were evidence for itself.

• Cherry-picking: Selecting supportive examples or data while leaving out relevant contrary evidence.

Recognizing a possible fallacy is a prompt to examine the reasoning, not a substitute for doing so. Judge a conclusion by its actual support.

A practical evaluation sequence

Use this sequence to reach a measured judgment:

1. State the main claim in your own words.

2. Identify the evidence offered and the source of each piece.

3. Test the evidence for credibility, relevance, sufficiency, and context.

4. Explain how each piece of evidence is meant to support the claim, and identify any assumptions in that reasoning.

5. Compare other credible evidence and consider limitations or counterarguments.

6. Decide whether the claim is supported, partly supported, or not established by the evidence presented.

The judgment should reflect the strength and limits of the available evidence. Treat an unsupported claim as unproven rather than automatically false.