Within Fallacy Lab
What Evidence Is the Argument Missing?
Many fallacies work by making thin or missing evidence feel stronger than it really is.
On this page
- Unsupported leaps
- Representative evidence
- Questions to ask
Page outline Jump by section
Introduction
Persuasive claims often sound stronger than they are because they hide an evidence gap. The claim may contain a true example, a confident statistic, a named expert, or a vivid story, but the evidence still may not justify the conclusion being drawn. In logical fallacies, this problem sits behind patterns such as hasty generalisation, cherry-picking, appeal to ignorance, weak analogy, false cause and misleading anecdote: the argument asks the reader to travel further than the evidence can safely carry them.
The central question is not “Is there any evidence?” but “Is there enough of the right kind of evidence for this claim, in this context?” Informal logic treats everyday argument as a matter of evidence, proof, burden and justification rather than just formal validity, because real-world claims usually depend on incomplete information and judgement. [stanford]plato.stanford.eduEncyclopedia of Philosophy Informal LogicStanford Encyclopedia of PhilosophyInformal Logic - Stanford Encyclopedia of Philosophyby L Groarke · 1996 · Cited by 97 — Different info… Encyclopedia of Philosophy Evidence gaps matter because they make thin support feel complete, shift attention away from what has not been shown, and encourage people to mistake plausibility for proof.
Where Unsupported Leaps Hide
An unsupported leap happens when the conclusion is broader, stronger or more certain than the evidence allows. The problem is not always that the evidence is false. Often it is real but too narrow, too indirect or too selectively presented.
A familiar example is: “Everyone is unhappy with this policy; I have spoken to five people and they all hate it.” The five conversations may be genuine. They may even reveal something important. But they do not support the claim “everyone”. The argument quietly turns a small convenience sample into a population-wide judgement. Writing and critical-thinking guides commonly describe this as hasty generalisation: drawing a broad conclusion from evidence that is too small, too exceptional or not representative enough. [Excelsior OWL]owl.excelsior.eduExcelsior OWLHasty Generalization Fallacy | Excelsior OWLThis fallacy occurs when an argument is based on a body of evidence that is simp…
Unsupported leaps can take several forms:
- From some to most: a few cases are treated as if they reveal the normal pattern.
- From correlation to cause: two things occur together, so one is assumed to have caused the other.
- From absence to disproof: because no evidence has been produced, the claim is treated as false; or because no disproof has been produced, it is treated as true.
- From expert mention to proof: a source is named, but the argument does not show that the source actually supports the specific claim.
- From vividness to frequency: a memorable example makes something feel common even when its actual rate is unknown.
The strongest warning sign is a mismatch between the evidence and the wording of the claim. A modest dataset may justify “this happened in these cases”. It does not automatically justify “this always happens”, “this is the main cause”, “this proves the whole theory”, or “critics have no answer”.
The Difference Between Evidence and Representative Evidence
Representative evidence is evidence that can reasonably stand for the wider group, pattern or situation being discussed. That is why sampling matters. Pew Research Center describes random sampling as central to probability-based survey research because it gives members of the target population a known chance of selection, reducing the risk that the sample merely reflects whoever was easiest to reach. [Pew Research Center]pewresearch.orgPew Research Center Methods 101: Random SamplingPew Research Center Methods 101: Random Sampling Public-health research guidance similarly distinguishes probability sampling, which supports generalisation more strongly, from non-probability sampling, which may be quicker but is more vulnerable to selection problems. [HealthKnowledge]healthknowledge.org.ukSource details in endnotes.
This matters in fallacy-spotting because many persuasive arguments offer evidence that is emotionally persuasive but not representative. A customer testimonial may show that one person had a good experience. It does not show the typical result. A viral post may show that an incident occurred. It does not show how often it occurs. A poll of readers on a partisan website may reveal the views of that website’s audience. It does not necessarily reveal the views of the wider public.
Representativeness depends on the claim being made. If the claim is “this can happen”, one documented case may be enough. If the claim is “this usually happens”, the evidence must show typicality. If the claim is “this policy caused the change”, the evidence must address alternative explanations, timing, comparison groups and confounding factors. In medical and population research, a sample is commonly treated as representative only in relation to a well-defined target population whose results can reasonably be generalised from that sample. [PMC]pmc.ncbi.nlm.nih.govSource details in endnotes.
A large sample does not automatically solve the problem. A huge but biased dataset can still mislead if important groups are missing or overrepresented. The gap is not just quantity; it is fit. The reader should ask whether the evidence matches the people, time period, conditions and outcome named in the conclusion.
Why “Some Evidence” Can Still Be Misleading
Persuasive claims often survive because they do contain something that looks like proof. The fallacy lies in what has been left out, not only in what has been included.
Cherry-picking is the clearest example. It presents selected evidence that supports a conclusion while omitting relevant evidence that weakens, complicates or reverses it. The individual facts may be accurate, but the overall picture is distorted. This is why cherry-picking is often described as a fallacy of incomplete evidence: the missing cases are part of the argument’s real evidential meaning. [Wikipedia]WikipediaCherry pickingCherry picking
A health claim, for instance, might cite one favourable study but ignore larger trials, failed replications, adverse findings or reviews that put the result in context. A political claim might show one dramatic local example while ignoring national data. A business claim might advertise “nine out of ten users saw improvement” without explaining how users were selected, what counted as improvement, how many dropped out, or whether there was a comparison group.
This is also why regulators focus on substantiation in advertising. The US Federal Trade Commission says objective advertising claims should have a reasonable basis before they are made, and that advertisers must substantiate both express and implied claims conveyed to reasonable consumers. [Federal Trade Commission]ftc.govSource details in endnotes. That principle is useful beyond advertising: the strength of the evidence should match the strength and practical significance of the claim.
Missing Counter-Evidence Is Part of the Evidence Gap
A one-sided argument is not always fallacious. In a short speech, essay or advert, nobody can include every detail. The problem begins when the omitted evidence is the evidence a fair assessment would need.
A claim such as “this programme works because three participants improved” leaves several obvious gaps. How many participants did not improve? Were the improvements measured before and after? Did similar people improve without the programme? Were the reported participants typical or specially chosen? The missing information is not a decorative extra; it determines whether the claim is weak, promising or well supported.
Critical-thinking guidance often frames this as a duty to seek both confirming and opposing information. The Foundation for Critical Thinking’s standards emphasise that reasoning rests on data, information and evidence, and that claims should be restricted to what the available data support while actively looking for information that opposes one’s position. [criticalthinking.org]criticalthinking.orgOpen source on criticalthinking.org.
This is especially important when a claim relies on:
- Anecdotes: vivid stories can reveal possibilities but cannot establish rates by themselves.
- Before-and-after comparisons: improvement after an intervention does not prove the intervention caused it.
- Expert quotations: a quotation may be accurate but taken out of context or applied beyond the expert’s claim.
- Single metrics: one number may hide trade-offs, uncertainty, subgroup differences or measurement problems.
- Absence of reports: lack of visible evidence may reflect poor detection, low reporting, secrecy, weak records or simply the fact that nothing happened.
The practical test is whether the argument would still sound persuasive if the missing evidence were placed beside it.
Absence of Evidence Is Not Always the Same Thing
The phrase “absence of evidence is not evidence of absence” is useful but easy to overuse. Sometimes missing evidence tells us very little. At other times, it tells us a lot. The difference depends on whether we would reasonably expect the evidence to exist if the claim were true.
The appeal to ignorance fallacy occurs when a person treats lack of proof as proof in the opposite direction: “No one has proved this is false, so it must be true,” or “No one has proved this is true, so it must be false.” Fallacy Files notes that this often appears as a burden-of-proof shift: the person making the claim demands that critics disprove it rather than supplying adequate support. [fallacyfiles.org]fallacyfiles.orgSource details in endnotes.
But absence can sometimes matter. If a claim predicts clear, repeated, detectable evidence and careful searches keep failing to find it, that absence may weaken the claim. In law, science and public policy, evidential standards often depend on what records, tests, witnesses or measurements should be available. The key is not a slogan but a question: would reliable evidence probably be visible by now if the claim were correct?
That distinction prevents two opposite mistakes. One mistake is believing extraordinary claims merely because nobody has conclusively disproved them. The other is dismissing uncertain claims too quickly when the right evidence has not yet been gathered.
The Dataset Questions That Expose Weak Claims
The quickest way to find an evidence gap is to translate the persuasive claim into a dataset question. Instead of asking whether the claim sounds plausible, ask what kind of evidence would have to exist for the claim to be well supported.
For a broad claim, ask: What is the target population? A claim about “students”, “voters”, “patients”, “customers” or “the public” needs evidence that actually reaches the relevant group. If the data only cover one school, one website, one clinic or one self-selected survey, the conclusion should be narrower.
For a frequency claim, ask: What is the denominator? “Thousands of complaints” sounds serious, but its meaning changes depending on whether it comes from ten thousand users or ten million. A raw count without a base rate can make rare events seem common or common events seem exceptional.
For a causal claim, ask: What else could explain the pattern? Public-health and epidemiology sources treat confounding as a routine concern because an observed association may be distorted by a third factor linked to both the supposed cause and the outcome. [HealthKnowledge]healthknowledge.org.ukSource details in endnotes. If an argument does not address plausible alternative causes, it has not yet shown causation.
For a statistical claim, ask: What does the number actually measure? The American Statistical Association’s statement on p-values warned that a p-value, by itself, does not provide a good measure of evidence for a model or hypothesis and must be interpreted with context and other evidence. [Taylor & Francis Online]tandfonline.comSource details in endnotes. A claim can therefore misuse a technically real statistic by giving it more argumentative weight than it can bear.
For a source-based claim, ask: Does the cited source support the exact conclusion? Recent research on fallacies in health misinformation has found that credible biomedical publications can be superficially cited in support of false claims, even when the original passages do not actually justify the conclusion being drawn. [arXiv]arxiv.orgarXiv Grounding Fallacies Misrepresenting Scientific Publications in EvidencearXiv Grounding Fallacies Misrepresenting Scientific Publications in Evidence The evidence gap is hidden inside the citation itself: the source exists, but the inference from source to claim is faulty.
A Practical Checklist for Reading Persuasive Claims
Evidence-gap detection works best as a short sequence of questions rather than as a hunt for fallacy names. The label can be useful later, but the first task is to identify what has and has not been shown.
- What exactly is the claim?
Rewrite it in plain terms. Watch for words such as “proves”, “everyone”, “never”, “the real reason”, “guaranteed”, “scientifically shown” or “no evidence”. These often raise the evidential standard.
- What evidence is actually offered?
Separate examples, statistics, expert opinion, documents and personal experience. Do not let the most vivid item stand in for the whole evidence base.
- What would stronger evidence look like?
A representative survey, a controlled comparison, a systematic review, a transparent dataset, a full quotation, a disclosed method or a clear denominator may be needed, depending on the claim.
- Is the evidence typical, selective or exceptional?
Ask whether the examples were chosen because they are representative or because they are persuasive.
- What relevant evidence is missing?
Look for absent comparison groups, missing timeframes, unreported failures, ignored counterexamples, undisclosed sampling methods, and alternative explanations.
- Does the conclusion need to be narrowed?
Often the honest repair is not to reject the claim completely but to restate it more carefully: “This happened in these cases”, “This suggests a possible link”, or “This is consistent with the claim but does not prove it.”
How Evidence Gaps Connect Different Fallacies
Evidence gaps are not a single fallacy so much as a common engine behind many fallacies. In a hasty generalisation, the gap is between the sample and the population. In cherry-picking, it is between selected evidence and the fuller record. In appeal to ignorance, it is between lack of proof and positive proof. In false cause, it is between association and causal explanation. In weak analogy, it is between superficial similarity and relevant similarity.
This is why fallacy labels should not be used as shortcuts for dismissal. A claim supported by limited evidence may be wrong, partly right, or simply underdeveloped. The better response is to name the gap and state what would close it. “That is a hasty generalisation” is less useful than “The evidence comes from three cases, but the claim is about the whole market; we would need a representative sample or a broader dataset.”
The strongest critical readers therefore ask evidence questions before reaching for rhetorical labels. They do not merely ask whether an argument is persuasive. They ask whether the evidence is sufficient, relevant, representative and fairly presented for the conclusion being sold.
Amazon book picks
Further Reading
Books and field guides related to What Evidence Is the Argument Missing?. Use these as the next step if you want deeper reading beyond the article.
Thinking, Fast and Slow
Explains why people accept weak evidence and unsupported leaps.
eBay marketplace picks
Marketplace Samples
Example marketplace items related to this page. Use the search link to explore similar finds on eBay.
Endnotes
-
Source: plato.stanford.edu
Title: Encyclopedia of Philosophy Informal Logic
Link: https://plato.stanford.edu/entries/logic-informal/Source snippet
Stanford Encyclopedia of PhilosophyInformal Logic - Stanford Encyclopedia of Philosophyby L Groarke · 1996 · Cited by 97 — Different info...
-
Source: owl.excelsior.edu
Link: https://owl.excelsior.edu/argument-and-critical-thinking/logical-fallacies/logical-fallacies-hasty-generalization/Source snippet
Excelsior OWLHasty Generalization Fallacy | Excelsior OWLThis fallacy occurs when an argument is based on a body of evidence that is simp...
-
Source: pmc.ncbi.nlm.nih.gov
Link: https://pmc.ncbi.nlm.nih.gov/articles/PMC10193086/ -
Source: Wikipedia
Title: Cherry picking
Link: https://en.wikipedia.org/wiki/Cherry_picking -
Source: criticalthinking.org
Link: https://www.criticalthinking.org/pages/the-national-council-for-excellence-in-critical-thinking/406 -
Source: fallacyfiles.org
Link: https://www.fallacyfiles.org/ignorant.html -
Source: arxiv.org
Title: arXiv Grounding Fallacies Misrepresenting Scientific Publications in Evidence
Link: https://arxiv.org/abs/2408.12812 -
Source: Wikipedia
Title: Argument from ignorance
Link: https://en.wikipedia.org/wiki/Argument_from_ignorance -
Source: Wikipedia
Title: List of fallacies
Link: https://en.wikipedia.org/wiki/List_of_fallacies -
Source: criticalthinking.org
Link: https://www.criticalthinking.org/data/pages/25/2e51737bb044ed99a88bdc1220fe8cb660f7687377bed.pdf -
Source: criticalthinking.org
Link: https://www.criticalthinking.org/pages/a-brief-history-of-the-idea-of-critical-thinking/408 -
Source: criticalthinking.org
Link: https://www.criticalthinking.org/data/pages/99/579d6b9722ed3418d0b4a7514f0a9f1d513655794857b.pdf -
Source: criticalthinking.org
Link: https://www.criticalthinking.org/pages/center-for-critical-thinking/401 -
Source: criticalthinking.org
Link: https://www.criticalthinking.org/store/get_file.php?inventories_id=159 -
Source: criticalthinking.org
Link: https://www.criticalthinking.org/data/pages/14/c4b33ac92ec0940ec11171a6d52d6627519d0ee10b37e.pdf -
Source: criticalthinking.org
Link: https://www.criticalthinking.org/files/White%20PaperAssessmentSept2007.pdf -
Source: criticalthinking.org
Link: https://www.criticalthinking.org/data/pages/86/2059cb576fc40aa34de9c999042fcf445f7cdc2d30ae6.pdf -
Source: criticalthinking.org
Link: https://www.criticalthinking.org/pages/glossary-of-critical-thinking-terms/496 -
Source: plato.stanford.edu
Title: logic informal
Link: https://plato.stanford.edu/archives/sum2022/entries/logic-informal/ -
Source: plato.stanford.edu
Link: https://plato.stanford.edu/entries/fallacies/ -
Source: plato.stanford.edu
Title: logic informal
Link: https://plato.stanford.edu/archives/fall2019/entries/logic-informal/ -
Source: plato.stanford.edu
Title: logic informal
Link: https://plato.stanford.edu/archives/fall2014/entries/logic-informal/ -
Source: plato.stanford.edu
Title: descartes epistemology
Link: https://plato.stanford.edu/entries/descartes-epistemology/ -
Source: plato.stanford.edu
Title: plato ethics
Link: https://plato.stanford.edu/archives/fall2015/entries/plato-ethics/ -
Source: plato.stanford.edu
Title: plato ethics
Link: https://plato.stanford.edu/archives/fall2007/entries/plato-ethics/ -
Source: youtube.com
Title: 31 logical fallacies in 8 minutes
Link: https://www.youtube.com/watch?v=Qf03U04rqGQSource snippet
Logical Fallacies...
-
Source: youtube.com
Title: Logical Fallacies
Link: https://www.youtube.com/watch?v=HpP71KmmlYQSource snippet
Critical Listening, Logical Fallacies, and Evaluating Evidence...
-
Source: pewresearch.org
Title: Pew Research Center Methods 101: Random Sampling
Link: https://www.pewresearch.org/methods/2017/05/12/methods-101-video-random-sampling/ -
Source: healthknowledge.org.uk
Link: https://www.healthknowledge.org.uk/public-health-textbook/research-methods/1a-epidemiology/methods-of-sampling-population -
Source: ftc.gov
Link: https://www.ftc.gov/legal-library/browse/ftc-policy-statement-regarding-advertising-substantiation -
Source: ftc.gov
Title: advertising faqs guide small business
Link: https://www.ftc.gov/business-guidance/resources/advertising-faqs-guide-small-business -
Source: healthknowledge.org.uk
Link: https://www.healthknowledge.org.uk/public-health-textbook/research-methods/1a-epidemiology/biases -
Source: tandfonline.com
Link: https://www.tandfonline.com/doi/full/10.1080/00031305.2016.1154108 -
Source: writingcenter.unc.edu
Link: https://writingcenter.unc.edu/tips-and-tools/fallacies/ -
Source: pewresearch.org
Title: for weighting online opt in samples what matters most
Link: https://www.pewresearch.org/methods/2018/01/26/for-weighting-online-opt-in-samples-what-matters-most/ -
Source: pewresearch.org
Link: https://www.pewresearch.org/methods/2018/08/06/video-explainer-what-are-nonprobability-surveys/ -
Source: pewresearch.org
Title: [social media]({{ ‘social-media/’ | relative_url }}) use 2025 methodology
Link: https://www.pewresearch.org/internet/2025/11/20/social-media-use-2025-methodology/ -
Source: askattest.com
Title: representative sample
Link: https://www.askattest.com/blog/articles/representative-sample -
Source: methods.sagepub.com
Link: https://methods.sagepub.com/hnbk/edvol/sage-hdbk-public-opinion-research/chpt/sampling -
Source: unr.edu
Link: https://www.unr.edu/writing-speaking-center/writing-speaking-resources/logical-fallacies -
Source: checkbox.com
Title: representative sample
Link: https://www.checkbox.com/blog/representative-sample -
Source: quillbot.com
Title: appeal to ignorance fallacy
Link: https://quillbot.com/blog/reasoning/appeal-to-ignorance-fallacy/
Additional References
-
Source: iep.utm.edu
Link: https://iep.utm.edu/fallacy/Source snippet
Internet Encyclopedia of PhilosophyFallaciesSo, [informal fallacies]({{ 'informal-logic/' | relative_url }}) are errors of reasoning that cannot easily be expressed in our standar...
-
Source: purdueglobalwriting.center
Link: https://purdueglobalwriting.center/hasty-generalizations-and-other-logical-fallacies/Source snippet
Purdue Global Success CenterHasty Generalizations and Other Logical FallaciesHasty generalizations are committed when a person draws a co...
-
Source: ftc.gov
Link: https://www.ftc.gov/sites/default/files/attachments/training-materials/substantiation.pdf -
Source: cdc.gov
Link: https://www.cdc.gov/field-epi-manual/php/chapters/analyze-interpret-data.html -
Source: labxchange.org
Link: https://www.labxchange.org/library/pathway/lx-pathway%3Ae14a306c-99b2-4ae8-9d59-b2169c30b54b/items/lb%3ALabXchange%3A75f36ede%3Ahtml%3A1/178803 -
Source: burtthompson.net
Link: https://www.burtthompson.net/uploads/9/6/8/4/9684389/wasserstein-2016__asa_p-value_statement.pdf -
Source: logicallyfallacious.com
Link: https://www.logicallyfallacious.com/logicalfallacies/Argument-from-Ignorance -
Source: errorstatistics.com
Link: https://errorstatistics.com/wp-content/uploads/2016/03/4_berry.pdf -
Source: scribd.com
Link: https://www.scribd.com/document/701425677/6-Session-Six-Evidence-Evaluation-Causality -
Source: logicallyfallacious.com
Link: https://www.logicallyfallacious.com/critical-thinking-fallacies
Topic Tree