Using evidence

Intuition [Beginner]

Evidence is the ground an argument stands on. Without it, you have opinions. With it, you have a case. But not all evidence is equal. A single dramatic story is not as strong as a controlled study. A quote from a celebrity is not as credible as a quote from a subject-matter expert. Learning to use evidence well means learning both how to find it and how to judge it.

Four main types of evidence appear in most writing:

1. Statistical / quantitative: Numbers, percentages, measurements. "Traffic fatalities decreased 23 percent after the law took effect." Strong when the data is from a reliable source; weak when cherry-picked or misleadingly presented.

2. Expert testimony: Statements from authorities in the relevant field. "According to Dr. Chen, a virologist at Stanford, the variant is no more severe than previous strains." Strong when the expert's field matches the claim; weak when the expert is speaking outside their area.

3. Anecdotal / illustrative: Specific stories or examples. "Maria, a single mother in Detroit, lost her job when the factory closed." Not proof of a general trend, but powerful for showing what a trend means in human terms.

4. Analogical: Comparisons to similar situations. "When Ireland taxed plastic bags in 2002, usage dropped 90 percent; a similar policy here would likely produce comparable results." Useful for prediction, but only if the situations are genuinely similar in relevant respects.

Good writers do not pick one type and rely on it exclusively. They combine types: statistics to show the scale, an anecdote to make it real, expert testimony to explain why it is happening.

Evidence must be integrated, not dropped in. A quotation without context or explanation is not evidence -- it is decoration. After presenting evidence, explain what it means and how it supports your claim. The reader should never have to guess why a piece of evidence is there.

Visual [Beginner]

Evidence types and their strengths:

  TYPE              BEST FOR                      WATCH OUT FOR
  ---------------   ---------------------------   ---------------------------
  Statistical       Showing scale, trends         Cherry-picked data,
                                                   misleading graphs

  Expert testimony  Establishing authority        Expert outside their field,
                                                   outdated opinions

  Anecdotal         Illustrating human impact     Generalizing from one case

  Analogical        Predicting from parallels     False analogies


The evidence sandwich:

  1. CLAIM       "Carbon emissions from transportation are rising."
  2. INTRODUCE   "According to the EPA's 2023 inventory,"
  3. EVIDENCE    "transportation overtook power generation as the
                  largest source of U.S. emissions in 2017."
  4. EXPLAIN     "This shift occurred because power plants switched
                  to natural gas and renewables while vehicle
                  emissions continued to grow."
        |
  Claim -> Introduce -> Evidence -> Explain -> (repeat or conclude)


Evaluating evidence quality:

  Source credibility:  Who produced it? Are they qualified? Biased?
  Recency:             When was it published? Is it still current?
  Relevance:           Does it actually address the specific claim?
  Representativeness:  Is this one case, or a pattern?
  Corroboration:       Do other sources confirm it?

Worked example [Beginner]

Example 1: Integrating statistical evidence

Weak (dropped in): Climate change is serious. The temperature has risen 1.1 degrees. Something must be done.

Strong (integrated): Global average temperature has risen approximately 1.1 degrees Celsius above pre-industrial levels (NASA, 2023). While 1.1 degrees may sound small, it represents a massive energy imbalance: the additional heat absorbed by the planet is equivalent to detonating several Hiroshima-sized atomic bombs per second. This warming has already contributed to a 13 percent decline in Arctic sea ice extent per decade.

The strong version names the source, explains why the number matters, and connects it to a concrete consequence.

Example 2: Using anecdotal evidence responsibly

Weak: John lost his home in the hurricane, so climate change is destroying communities.

The anecdote of one person does not prove a general claim about climate change. It illustrates what the general claim means at the human level.

Strong: Hurricane Maria destroyed an estimated 400,000 homes in Puerto Rico (FEMA, 2018). Among those homes was the house where John Rodriguez had lived for thirty years. Rodriguez's story is not unique: it is one of hundreds of thousands of similar losses that together illustrate the human cost of increasingly severe weather events.

The anecdote serves as illustration after the statistical scale is established.

Example 3: Expert testimony used correctly vs. incorrectly

Incorrect: Dr. Smith says vaccines are dangerous. (Who is Dr. Smith? What is their field? Are they a relevant expert?)

Correct: According to Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases from 1984 to 2022, the mRNA vaccines underwent the largest clinical trials in vaccine history and showed 94-95 percent efficacy at preventing symptomatic COVID-19.

The correct version identifies the expert, their credentials, and their specific claim.

Example 4: Avoiding cherry-picking

Cherry-picked: Economic growth was 4.2 percent in Q3, proving the policy works.

Complete picture: Economic growth spiked to 4.2 percent in Q3 but fell to 1.8 percent in Q4 and averaged 2.1 percent for the year -- roughly in line with the pre-policy average of 2.0 percent. The Q3 spike appears to reflect a one-time inventory adjustment rather than a sustained change in trend.

Check your understanding [Beginner]

Formal definition [Intermediate+]

Evidence is any material presented in support of a claim. In argumentative and expository writing, evidence functions as the grounds (in Toulmin's terms) upon which the argument's reasoning operates. The four major categories of evidence are:

1. Statistical / quantitative evidence: Numerical data derived from systematic observation, measurement, or sampling. Includes surveys, experiments, censuses, and meta-analyses. Evaluated by sample size, methodology, source credibility, and whether the data is presented in context or cherry-picked.

2. Expert testimony: Statements by individuals recognized as authorities in a relevant field. Evaluated by the expert's credentials, the relevance of their expertise to the specific claim, and whether their view represents a scholarly consensus or a minority position.

3. Anecdotal evidence: Specific instances, stories, or examples. Evaluated by representativeness (does this case typify the general pattern?) and by whether it is presented as illustration (legitimate) or as proof of a general claim (fallacious).

4. Analogical evidence: Reasoning from a known case to a similar case. Evaluated by the relevance and number of similarities between the cases and the absence of critical differences that would undermine the comparison.

Evidence integration follows a pattern sometimes called the "evidence sandwich" or "ICE" (Introduce, Cite, Explain):

• Introduce: Provide context (who, what, when, where) before the evidence.
• Cite / Present: State the evidence, with attribution.
• Explain: Show how the evidence supports the claim. Never assume the reader will make the connection independently.

Cherry-picking is the selective presentation of evidence that supports a predetermined conclusion while omitting evidence that contradicts it. It is an epistemic error (it misrepresents the state of knowledge) and a rhetorical error (it undermines credibility when the omission is discovered).

The hierarchy of evidence (from evidence-based medicine, applicable more broadly) ranks evidence types by reliability:

• Systematic reviews and meta-analyses (strongest)
• Randomized controlled trials
• Cohort studies
• Case-control studies
• Case series and case reports
• Expert opinion / anecdote (weakest)

This hierarchy is a guide, not an absolute rule. A well-designed observational study can be more informative than a poorly designed randomized trial.

Key concepts [Intermediate+]

• Evidence types: Statistical, expert testimony, anecdotal, analogical. Each has strengths and limitations; effective arguments combine types.
• Evidence integration: The practice of introducing, presenting, and explaining evidence so its relevance to the claim is explicit.
• Cherry-picking: Selecting only evidence that supports a conclusion while ignoring contradictory evidence. Undermines both accuracy and credibility.
• Source credibility: The trustworthiness of an evidence source, determined by expertise, institutional affiliation, peer review status, and track record.
• Representativeness: Whether a sample or example accurately reflects the larger population or pattern it is invoked to represent.
• Evidence hierarchy: A ranking of evidence types by reliability. Systematic reviews outrank expert opinion, but context matters.
• Corroboration: Whether multiple independent sources support the same finding. Single-source claims are inherently weaker than multi-source claims.

Rhetorical theory [Master]

The use of evidence in argument connects to ancient questions about the nature of proof. Aristotle distinguished between entechnic (intrinsic) proofs, which the rhetorician discovers through argument, and atechnic (extrinsic) proofs, which exist independently of the argument (witnesses, documents, contracts). This distinction maps roughly onto the modern difference between reasoning (how you connect evidence to claims) and evidence itself (the data you work with).

The scientific revolution transformed Western expectations about evidence. Before the seventeenth century, authority (citation of Aristotle, Scripture, or legal precedent) was a primary mode of proof. The empirical method, as articulated by Bacon and practiced by the Royal Society, established observation and experiment as the gold standard. This shift created the modern evidence hierarchy: empirical data outranks authority, and replicated studies outrank single findings.

In the twentieth century, the field of informal logic developed systematic accounts of evidence evaluation. The critical thinking movement, associated with Robert Ennis, John McPeck, and Richard Paul, argued that evaluating evidence quality is a core cognitive skill that education should develop. In composition studies, this led to pedagogical emphasis on source evaluation, especially in the context of research writing.

The rise of the internet and digital media has complicated evidence evaluation. Information abundance means that supporting evidence is available for virtually any claim, including false ones. The challenge has shifted from finding evidence to evaluating it. Pedagogical responses include the "CRAAP test" (Currency, Relevance, Authority, Accuracy, Purpose), lateral reading strategies (checking what other sources say about a source), and the "SIFT" method (Stop, Investigate the source, Find better coverage, Trace claims to their origin).

In rhetoric and composition, the concept of kairos -- the timeliness and appropriateness of an argument for its moment -- applies to evidence as well. Evidence that was current and compelling ten years ago may be outdated. Evidence that is persuasive in one cultural context may carry no weight in another. Effective arguers attend not only to the logical strength of their evidence but to its rhetorical appropriateness for the specific audience and situation.

Postmodern and social-constructionist critiques question whether evidence can ever be "neutral." Thomas Kuhn's The Structure of Scientific Revolutions (1962) argued that what counts as evidence is shaped by the prevailing paradigm -- scientists working within different frameworks literally see different things. This critique does not entail that all evidence is equally valid, but it complicates the assumption that evidence speaks for itself independent of interpretive frameworks.

Historical context [Master]

The use of evidence has evolved alongside Western epistemology. In classical rhetoric, evidence (pisteis) included both logical argument and extrinsic proofs such as laws, witnesses, and documents. Roman law developed formal rules of evidence that influenced legal systems for centuries.

The medieval period saw the development of the scholastic method, which used evidence in the context of authoritative texts. A typical scholastic argument began with a question, cited authoritative sources on both sides, and resolved the question through reasoning grounded in those authorities. Evidence was textual and authority-based rather than empirical.

The scientific revolution introduced experimental evidence as a new standard of proof. Robert Boyle's air pump experiments (1660s) were significant not only for their findings but for establishing the practice of public, repeatable experimentation as a form of evidence. The Royal Society's motto, Nullius in verba ("Take nobody's word for it"), encapsulated the new emphasis on evidence over authority.

In the nineteenth and twentieth centuries, statistical evidence gained prominence. The development of probability theory, sampling methods, and statistical inference (Fisher, Neyman, Pearson) created tools for drawing conclusions from numerical data. The rise of social science brought statistical evidence into domains previously governed by qualitative argument.

In the twenty-first century, evidence evaluation has become both more important and more difficult. The internet provides instant access to vast quantities of information, but it also enables the rapid spread of misinformation. Fact-checking organizations, media literacy initiatives, and critical thinking curricula have emerged in response. In academic writing, the expectation for evidence has intensified: a research paper in 2025 is expected to engage with more sources, from more diverse perspectives, than was typical fifty years ago.

Bibliography [Master]

• Aristotle. Rhetoric. Book I on proofs (pisteis).
• Kuhn, Thomas S. The Structure of Scientific Revolutions. 2nd ed. Chicago: University of Chicago Press, 1970.
• Walton, Douglas. Argumentation Schemes for Presumptive Reasoning. Mahwah, NJ: Lawrence Erlbaum, 1996.
• Wineburg, Sam, and Sarah McGrew. "Lateral Reading and the Nature of Expertise: Reading Less and Learning More When Evaluating Digital Information." Teachers College Record 119, no. 11 (2017): 1-40.
• Ennis, Robert H. "A Taxonomy of Critical Thinking Dispositions and Abilities." In Teaching Thinking Skills: Theory and Practice, edited by Joan Boykoff Baron and Robert J. Sternberg, 9-26. New York: W. H. Freeman, 1987.
• Shapin, Steven, and Simon Schaffer. Leviathan and the Air-Pump: Hobbes, Boyle, and the Experimental Life. Princeton: Princeton University Press, 1985.