What Did the Post Leave Out?

Introduction

A social media claim can be true in the narrowest sense and still mislead. The trick is often not a fabricated number, but a missing frame: when the event happened, where it happened, what definition was used, what the denominator was, and what baseline the comparison quietly assumes. In an AI-shaped information environment, this problem becomes easier to scale because old clips, isolated statistics and plausible-sounding summaries can be repackaged quickly for new audiences.

The critical question is therefore not only “Is this false?” but “What did the post leave out?” A post that says crime is “up”, a clip that appears to show today’s protest, or a chart comparing two countries may all collapse once the date, comparison group, population size or measurement method is restored. Good statistical communication guidance makes the same point: numbers need context, including limitations, long-term trends, geographical comparisons and explanations of what the figures mean in practice. [Government Analysis Function]analysisfunction.civilservice.gov.ukernment Analysis FunctionWriting about statistics – Government Analysis Function…

Dates and Old Content

One of the simplest ways to mislead online is to make old material look current. The image or video may be real, but its date has been stripped away and replaced with a new caption. This is powerful because viewers often react to the scene before checking whether it belongs to the event being discussed.

Full Fact’s 2025 report identifies the re-use of old footage and imagery in new false contexts as a recurring tactic, especially during breaking news. It gives examples including footage from a Hindu religious festival procession being misrepresented during the UK riots of summer 2024, a 2015 warehouse explosion in Tianjin being presented as an explosion at Mossad headquarters, and 2022 footage being re-used in claims about Ukrainian troops in 2025. [Full Fact]fullfact.orgfull fact report 2025Full FactFull Fact Report 2025 – Full Fact…

This kind of claim works because it exploits the “now” bias of the feed. A video appears between fresh posts, current reactions and breaking-news hashtags, so the platform itself supplies a false sense of recency. The missing date is not a small detail; it is the difference between evidence and atmosphere.

Old content is especially damaging during crises because it competes with real-time updates. When authentic footage from a previous event is repurposed, it can confuse the public and divert attention from reliable current information. [Full Fact]fullfact.orgfull fact report 2025Full FactFull Fact Report 2025 – Full Fact… Full Fact’s earlier guidance on misleading videos makes the same distinction: some videos are not fake in the sense of being staged or edited, but are mislabelled as happening in another country, involving different people, or showing a different incident. [Full Fact]fullfact.orgFull Fact How to spot misleading videos online – Full FactFull Fact How to spot misleading videos online – Full Fact

A useful reader test is: “Would this claim still matter if the video were five years old, from another country, or from an unrelated event?” If the answer is no, the date and location are not background information. They are the claim.

Definitions and Denominators

Numbers are often most misleading when they look most precise. A percentage can feel authoritative while hiding the category being counted. “Most”, “record”, “surge”, “majority” and “risk” all depend on definitions: most of what, compared with whom, during which period, and counted how?

Denominators are the quiet base of a claim. A post saying “10,000 cases” sounds alarming or trivial depending on whether it refers to a city, a country, a hospital system, a week, a decade, or a population of millions. A percentage can hide the same problem in reverse. “A 100% increase” may mean two cases rising to four; “only 1%” may still mean a very large number of people if the population is huge.

The UK Code of Practice for Statistics says producers should present statistics impartially and objectively, avoid misleading communication, and provide clear explanations that support appropriate interpretation. It also says notable misuse should be challenged. [Code of Practice for Statistics]code.statisticsauthority.gov.ukOpen source on statisticsauthority.gov.uk. These are not technical niceties for statisticians; they are the exact missing pieces readers need when a number is posted without its frame.

A clear example is Full Fact’s analysis of viral Facebook posts about UK state pensions. The posts treated National Insurance contributions as if they built up in an individual pension pot with interest, then asked why pensioners were not receiving the resulting large sum. The underlying mechanism was wrong: National Insurance operates broadly on a pay-as-you-go basis, funding current pensions and other contributory benefits rather than building an individual investment account. Full Fact also found that even under the posts’ mistaken assumption, the arithmetic was wrong. [Full Fact]fullfact.orgFull Fact Viral Facebook pension posts get maths wrong again – Full FactFull Fact Viral Facebook pension posts get maths wrong again – Full Fact

That example matters because the claim was not merely a bad calculation. It used the wrong definition of the system. Once the definition changes from “personal investment pot” to “current-year public fund”, the emotional force of the comparison changes too.

Bad Comparisons and Hidden Baselines

Comparisons are persuasive because they seem to answer the reader’s natural question: “Is this big or small?” But a comparison can mislead if the baseline is chosen to exaggerate, minimise or redirect attention.

A hidden baseline often appears in phrases such as “up by”, “more than”, “worse than”, “highest since” or “only”. These phrases are incomplete unless the reader knows the starting point. “Spending doubled” sounds dramatic, but may be unsurprising if the original figure was unusually low, the population grew, prices rose, or a new duty was added. “Country A has more cases than Country B” may be meaningless without population size, testing rates, age structure or reporting rules.

PolitiFact’s check of a claim that “95%” of gun violence occurs in “inner cities” shows how a denominator can steer interpretation. The source claim relied on county-level concentration, but PolitiFact noted that using all US counties as the baseline can mislead because many counties have very small populations; collectively, large numbers of small counties account for only a tiny share of the population. It also noted that “inner city” has no standard federal definition. [PolitiFact]politifact.comPoliti Fact Is 95% of gun violence occurring in 'inner cities'?No10 May 2023 — However, using the nation's 3,145 counties as a baseline can be misleading, because the country includes many small-popul…Published: May 2023

This is the mechanism behind many viral charts: the comparison unit is technically countable, but not necessarily fair. Counties, countries, age bands, schools, hospitals, social media accounts and time periods can all be used as units. The question is whether the unit matches the claim being made.

A bad comparison may also smuggle in a moral judgement. For example, comparing raw totals between two places can imply one place is more dangerous, wasteful or successful, when the difference may largely reflect population size. Comparing two dates without explaining a policy change, data revision or change in testing can imply a trend that the data cannot support.

Why AI Makes Missing Context Easier to Miss

Generative AI does not create the missing-context problem, but it can make it more fluent. A chatbot summary, AI-written caption or synthetic explainer can present an old statistic with clean wording and confident structure, even when the key date, source, denominator or uncertainty has disappeared.

Research on AI-generated misinformation on X, using a dataset of 91,452 misleading posts flagged through Community Notes, found that AI-generated misinformation was more likely to be entertainment-centred, more likely to come from smaller accounts, and more likely to go viral than non-AI-generated misinformation, while being judged slightly less believable and harmful. [arXiv]arxiv.orgarXiv Characterizing AI-Generated Misinformation on Social MediaCharacterizing AI-Generated Misinformation on Social MediaMay 15, 2025…Published: May 15, 2025 That finding is important for missing context because a post does not have to look like formal propaganda to spread. It may look like a witty summary, a neat “explainer”, or a shareable comparison.

AI systems also struggle with temporal context. Recent research on temporal fact conflicts in large language models describes how models can struggle with outdated or evolving information in their training data, and how external context does not always resolve those conflicts reliably across datasets and methods. [arXiv]arxiv.orgTemporal Fact Conflicts in LLMs: Reproducibility Insights from Unifying DYNAMICQA and MULANMarch 16, 2026…Published: March 16, 2026 For readers, the practical lesson is simple: when a claim depends on who currently holds a role, what the latest figures show, or whether a rule has changed, the date of the evidence matters.

The Reuters Institute’s 2025 work on how people check information found that when users suspect something may be wrong, they rely on a mixed set of sources: news media, official sources, search engines, fact-checkers, Wikipedia, specialist experts, online personalities, social media and AI chatbots. Younger adults were more likely than older adults to mention comments, social media and AI chatbots as places they would use to check. [reutersinstitute.politics.ox.ac.uk]reutersinstitute.politics.ox.ac.ukOpen source on ox.ac.uk. That makes context literacy more important, because verification is no longer confined to professional sources.

How to Read a Claim for Missing Context

The fastest way to test a claim is to rebuild the frame around it. This does not require advanced statistics. It requires asking what a careful caption, chart note or source paragraph would have included.

Use these checks before sharing a date-sensitive or comparison-heavy post:

These questions are especially useful because missing context often survives ordinary fact-checking instincts. A person may search the headline figure, find that the number appears somewhere real, and stop. The better question is whether the number is being used in the same way as the original source.

What a Better Post Would Show

A trustworthy post does not need to include every caveat, but it should give readers enough frame to interpret the claim. For a statistic, that means the source, date range, population, definition and comparison point. For a video, it means when and where it was filmed. For a chart, it means labelled axes, units, baselines and any changes in method.

The Government Statistical Service’s writing guidance says commentary should provide a full picture and help readers understand strengths, limitations, long-term trends and geographical comparisons. [Government Analysis Function]analysisfunction.civilservice.gov.ukernment Analysis FunctionWriting about statistics – Government Analysis Function… The same principle applies to social media, even in compressed form. A short post can still say “per 100,000 people”, “in England and Wales”, “in 2023/24”, “compared with the five-year average”, or “old footage from 2022”.

The aim is not to make every reader suspicious of every number. It is to make the missing frame visible. In the age of social media and AI, many misleading claims are not lies in disguise. They are fragments presented as if they were the whole picture. Critical thinking begins when the reader asks what the fragment was cut away from.

Amazon book picks

Further Reading

Books and field guides related to What Did the Post Leave Out?. Use these as the next step if you want deeper reading beyond the article.

Book

How to Lie with Statistics

By Darrell Huff

Directly addresses missing context, denominators, and misleading comparisons.

See on Amazon

Book

Calling Bullshit

By Carl T. Bergstrom, Jevin D. West

Explains how context omissions distort conclusions.

See on Amazon

Book

Factfulness

By Hans Rosling, Ola Rosling et al.

Shows why trends and statistics need context.

See on Amazon

Book

Data Detective

By Tim Harford

First published 2021. Subjects: Statistics, Methodology, Social sciences, Statistical methods, Business.

See on Amazon

Browse more on Amazon: How to Lie with Statistics Calling Bullshit Factfulness

As an Amazon Associate I earn from qualifying purchases.

eBay marketplace picks

Marketplace Samples

Example marketplace items related to this page. Use the search link to explore similar finds on eBay.

Shop location

Using USA USA USAUKCanadaAustraliaIreland

Example eBay listing

Critical thinking the other national deficit Vinyl Car Window Wall Decal Sticker

Search eBay.co.uk: critical thinking sticker

Browse similar on eBay.co.uk

Example eBay listing

Critical Thinking The Other National Deficit 3x10 Vinyl Bumper Sticker

Search eBay.co.uk: critical thinking sticker

Browse similar on eBay.co.uk

Example eBay listing

Critical Thinking The Other National Deficit 2.5x8 Vinyl Bumper Sticker

Search eBay.co.uk: critical thinking sticker

Browse similar on eBay.co.uk

Browse more on eBay.co.uk

Example items shown for inspiration; availability and pricing can change. Branchoria may earn a commission if you purchase through outbound eBay links.

Endnotes

1. Source: politifact.com
   Title: Politi Fact Is 95% of gun violence occurring in ‘inner cities’?
   Link: https://www.politifact.com/factchecks/2023/may/10/marc-short/is-95-of-gun-violence-occurring-in-inner-cities-no/
   Source snippet

   No10 May 2023 — However, using the nation's 3,145 counties as a baseline can be misleading, because the country includes many small-popul...

   Published: May 2023

2. Source: arxiv.org
   Title: arXiv Characterizing AI-Generated Misinformation on Social Media
   Link: https://arxiv.org/abs/2505.10266
   Source snippet

   Characterizing AI-Generated Misinformation on Social MediaMay 15, 2025...

   Published: May 15, 2025

3. Source: arxiv.org
   Link: https://arxiv.org/abs/2603.15892
   Source snippet

   Temporal Fact Conflicts in LLMs: Reproducibility Insights from Unifying DYNAMICQA and MULANMarch 16, 2026...

   Published: March 16, 2026

4. Source: reutersinstitute.politics.ox.ac.uk
   Link: https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2025/how-public-checks-information-it-thinks-might-be-wrong

5. Source: arxiv.org
   Link: https://arxiv.org/html/2603.11058v1

6. Source: arxiv.org
   Link: https://arxiv.org/html/2503.05565v1

7. Source: arxiv.org
   Link: https://arxiv.org/html/2505.10266v1

8. Source: facebook.com
   Link: https://www.facebook.com/freepressunltd/posts/-a-photo-without-context-an-old-video-presented-as-new-or-a-claim-designed-to-pr/1480089297484125/

9. Source: Wikipedia
   Title: Base rate fallacy
   Link: https://en.wikipedia.org/wiki/Base_rate_fallacy

10. Source: analysisfunction.civilservice.gov.uk
    Link: https://analysisfunction.civilservice.gov.uk/policy-store/writing-about-statistics-2/
    Source snippet

    ernment Analysis FunctionWriting about statistics – Government Analysis Function...

11. Source: fullfact.org
    Title: full fact report 2025
    Link: https://fullfact.org/policy/reports/full-fact-report-2025/
    Source snippet

    Full FactFull Fact Report 2025 – Full Fact...

12. Source: fullfact.org
    Title: Full Fact How to spot misleading videos online – Full Fact
    Link: https://fullfact.org/blog/2018/aug/how-spot-misleading-videos-online/

13. Source: code.statisticsauthority.gov.uk
    Link: https://code.statisticsauthority.gov.uk/standards-of-the-code-of-practice/standards-for-official-statistics-with-required-practices/

14. Source: fullfact.org
    Title: Full Fact Viral Facebook pension posts get maths wrong again – Full Fact
    Link: https://fullfact.org/economy/pensions-maths-facebook-viral-repeat/

15. Source: fullfact.org
    Title: missing information disappearing archives data sources
    Link: https://fullfact.org/technology/missing-information-disappearing-archives-data-sources/

16. Source: fullfact.org
    Link: https://fullfact.org/about/frequently-asked-questions/

17. Source: fullfact.org
    Title: a year of fact checking the middle east conflict
    Link: https://fullfact.org/blog/2024/oct/a-year-of-fact-checking-the-middle-east-conflict/

18. Source: fullfact.org
    Title: covid vaccine rap battle
    Link: https://fullfact.org/health/covid-vaccine-rap-battle/

19. Source: fullfact.org
    Link: https://fullfact.org/about/[corrections

20. Source: fullfact.org
    Title: china gaza planes aid pyramids claimfalse
    Link: https://fullfact.org/conflict/china-gaza-planes-aid-pyramids-claimfalse/

21. Source: fullfact.org
    Title: penalty charge notice
    Link: https://fullfact.org/online/penalty-charge-notice/

22. Source: osr.statisticsauthority.gov.uk
    Title: statisticsauthority.gov.uk Guidance
    Link: https://osr.statisticsauthority.gov.uk/guidance/

23. Source: osr.statisticsauthority.gov.uk
    Title: short guide to code of practice standard six be rigorous
    Link: https://osr.statisticsauthority.gov.uk/guidance/short-guide-to-code-of-practice-standard-six-be-rigorous/

24. Source: osr.statisticsauthority.gov.uk
    Link: https://osr.statisticsauthority.gov.uk/wp-content/uploads/2023/02/OSR_Statistical_Literacy_Research_Report.pdf

25. Source: GOV.UK
    Title: guidance on accessing and using comparable datasets for uk government officials
    Link: https://www.gov.uk/government/publications/guidance-on-accessing-and-using-comparable-datasets-for-uk-government-officials/guidance-on-accessing-and-using-comparable-datasets-for-uk-government-officials

26. Source: blog.ons.gov.uk
    Title: the truth behind the numbers spotting statistical misuse
    Link: https://blog.ons.gov.uk/2025/02/17/the-truth-behind-the-numbers-spotting-statistical-misuse/

Additional References

1. Source: youtube.com
   Title: How (Not) to Lie with Data Visualisations
   Link: https://www.youtube.com/watch?v=2204fMpL84E
   Source snippet

   OSMC 2024 | The Subtle Art of Lying with Statistics by Dave McAllister...

2. Source: youtube.com
   Title: The Media Literacy Crisis Is Crisis-ing
   Link: https://www.youtube.com/watch?v=dfaMdpyaAaU
   Source snippet

   How (Not) to Lie with Data Visualisations - Curtis Wilson - NIDC 2025...

3. Source: globalfactcheck.org
   Link: https://www.globalfactcheck.org/documents/2019-06-27%20RoyalStatScandal.%20%20Documentary%20evidence%20-%20Royal%20Statistical%20Society%20false%20and%20misleading%20claims.%20%20Matt%20Berkley%20draft%20192.pdf

4. Source: researchgate.net
   Link: https://www.researchgate.net/publication/396009805_Generative_AI_and_misinformation_a_scoping_review_of_the_role_of_generative_AI_in_the_generation_detection_mitigation_and_impact_of_misinformation

5. Source: researchgate.net
   Link: https://www.researchgate.net/publication/362867698_Identifying_the_Drivers_Behind_the_Dissemination_of_Online_Misinformation_A_Study_on_Political_Attitudes_and_Individual_Characteristics_in_the_Context_of_Engaging_With_Misinformation_on_Social_Media

6. Source: scilit.com
   Link: https://www.scilit.com/publications/d7b22abc420a0732ca16c277f0b54822

7. Source: mtu.edu
   Link: https://www.mtu.edu/badinfo/

8. Source: researchgate.net
   Link: https://www.researchgate.net/publication/340810923_Causes_of_Misleading_Statistics_and_Research_Results_Irreproducibility_A_Concise_Review

9. Source: stratcomcoe.org
   Link: https://stratcomcoe.org/cuploads/pfiles/nato_stratcom_coe_fact-checking_and_debunking_02-02-2021-1.pdf

10. Source: medium.com
    Link: https://medium.com/%40khaoujai/how-large-language-models-detect-misinformation-391f84fd5f7b