The Attention Landscape · established evidence

Media Nations, Digital News Report, ATUS: What the World's Three Best-Documented National Media-Diet Studies Would Look Like Combined

Last reviewed 2026-07-20. Written by Chandranshu Kumar, Founder, Raveneye Global. · 10 min read

Three national instruments currently set the standard for documenting how a population's attention is spent, and each discloses its method in public detail rather than treating it as proprietary. Ofcom's Media Nations report triangulates a hybrid panel of British households with census-scale return-path and streaming data, and separately publishes an as-broadcast series for trend comparability. The Reuters Institute's Digital News Report surveys more than 97,000 online news consumers across 48 markets using quota-weighted access panels, and states plainly where its sample represents only younger, English-speaking, online populations. The US Bureau of Labor Statistics' American Time Use Survey asks one person per household to diary a full day, continuously, since 2003, funded by no advertiser. None of the three is a census. Comparing what each discloses, panel versus census-scale data, quota versus probability sampling, single-medium versus whole-day, is itself a template for how any organization should document its own attention research.

Three instruments, three philosophies of evidence

Most numbers quoted about media consumption come from vendors who do not publish their method in enough detail for an outsider to judge it. Three national instruments are the exception. Ofcom in the United Kingdom, the Reuters Institute for the Study of Journalism at the University of Oxford, and the United States Bureau of Labor Statistics each release a methodology document alongside their headline figures, and each has made a different, defensible choice about how to trade off cost, scale, and rigor.

The three do not measure the same object. Ofcom measures broadcast and streaming viewing time in the UK. The Reuters Institute measures how people say they access news across 48 markets. The BLS measures how Americans report spending an entire day, media included. Treating any one of them as a universal answer to "how much attention does media get" would misread what each was built to do. Reading them side by side, for what each discloses rather than for a single comparable topline, is the more honest exercise.

Ofcom's Media Nations: a hybrid that admits panel and census disagree

Ofcom's 2025 Media Nations report, published 30 July 2025, is built on BARB panel data, itself a hybrid of a people-meter panel and census-level online return-path and streaming data. For its headline as-viewed time-spent figures, Ofcom uses this hybrid feed. But it separately reports an as-broadcast, 28-day consolidated series specifically for long-run trend comparability, because the report's own methodology annex acknowledges that panel data and census-scale data answer related but distinct questions and should be reconciled explicitly rather than blended silently into one number.

That is an unusually candid design choice. Most vendors that combine multiple data sources do not tell you where the sources diverge; Ofcom's annex does. The lesson for any organization building its own measurement is not the UK viewing figures themselves, it is the practice of publishing two series side by side when the underlying methods are not interchangeable, instead of quietly picking whichever number is more convenient.

The Digital News Report: a global read built on quotas, not random selection

The Reuters Institute's Digital News Report is the closest thing available to a global cross-market census of news consumption habits, and its 2025 edition surveyed more than 97,000 online news consumers across 48 markets through the polling firm YouGov. But the sampling method is a non-probability online access panel with nationally representative quotas on variables like age, gender, and region, and sometimes education or political vote, not random probability sampling.

The report's own methodology page states this limitation for specific markets rather than glossing over it: in India, Kenya, Nigeria, and South Africa, the sample explicitly represents only younger, English-speaking, online populations, not each country's national population as a whole. That single disclosed caveat is worth more than the topline percentages, because it tells a reader exactly which claims the data can and cannot support.

ATUS: the one instrument built by neither an advertiser nor a media company

The American Time Use Survey, run continuously by the Bureau of Labor Statistics since January 2003, takes a structurally different approach from both Ofcom and the Reuters Institute. It samples one person aged 15 or older per household, drawn from the outgoing rotation of the Current Population Survey, and has that person report a full 24-hour diary of activities, including media and leisure time, coded to standardized activity categories.

ATUS is not a media-industry instrument at all; it is a government labor statistics survey that happens to capture media time as one category among many daily activities. That gives it a property none of the commercial or academic media instruments share: no advertiser, broadcaster, or platform funds it or has a stake in what it finds. Its limitation is the mirror image of that strength, a single respondent per household and a single diary day, which is a narrow window on any one person's actual pattern.

What combining them honestly would, and would not, look like

A tempting shortcut is to average or stack figures from all three into one number for "how much attention media gets." That shortcut is not available here, and pretending otherwise would misrepresent all three sources. Ofcom measures UK broadcast and streaming time specifically; the Reuters Institute measures self-reported news access habits across a news-specific questionnaire; ATUS measures self-reported time across every daily activity, of which media is one category, for the United States only.

What can honestly be combined is not the numbers but the disclosure practices behind them. Laid side by side, the three instruments demonstrate three separate, named tradeoffs: hybrid panel-plus-census reconciliation (Ofcom), quota-weighted access-panel scale with named population caveats (Reuters Institute), and single-respondent whole-day diary independence from industry funding (ATUS). A fourth instrument attempting to measure a population's attention honestly would need to make its own tradeoff on this same set of choices and state it as plainly as these three do.

Why the disclosure, not the topline, is the real finding

Every methodology document reviewed here says, in effect, "here is what we can and cannot claim from this sample." That habit is rarer than it should be in commercial media-consumption reporting, where a "hours per day" figure is frequently published without any accompanying account of how it was derived. The existence of three separately funded, methodologically transparent instruments, run by a UK regulator, an Oxford research institute, and a US government agency, shows that rigorous disclosure is achievable at national and cross-national scale. It is a choice, not a constraint of the subject matter.

For any organization building a model of where attention actually sits, the operational lesson is not to copy any single instrument's numbers, since none of the three answers the same question a business needs answered. The lesson is to copy the discipline: name the sample, name the method, disclose where the data does and does not generalize, and never present a modeled estimate as a census.

The evidence

Key findings, with their sources

Ofcom's Media Nations 2025 uses BARB hybrid panel and census-level online return-path and streaming data for as-viewed time-spent figures, and separately reports as-broadcast 28-day consolidated data for long-run trend comparability, treating panel and census-scale data as answering different questions that must be reconciled, not blended silently.

established Ofcom, "Media Nations 2025" and "Media Nations 2025: Annex Methodologies," published 30 July 2025.
The Reuters Institute Digital News Report 2025 surveyed more than 97,000 online news consumers across 48 markets via YouGov, using nationally representative quotas rather than probability sampling; in India, Kenya, Nigeria, and South Africa the sample explicitly represents only younger, English-speaking, online populations, not the national population.

established Reuters Institute for the Study of Journalism, University of Oxford, "Methodology," Digital News Report 2025.
The BLS American Time Use Survey has run continuously since January 2003, sampling one person aged 15 or older per household drawn from the outgoing Current Population Survey rotation, and has them report a full 24-hour activity diary coded to standardized categories including media and leisure.

established Bureau of Labor Statistics, American Time Use Survey User's Guide; ATUS News Release.
No instrument among the national and cross-national media-measurement studies reviewed here produces a true population census of digital or media exposure; every figure reviewed is sample-based, panel-based, or self-reported, with disclosed or discoverable coverage limits.

established Cross-source synthesis of Ofcom, Reuters Institute, and BLS methodology documentation.

Reference

Glossary

Probability sampling: A recruitment method in which every member of a population has a known, nonzero chance of selection, allowing formal statistical inference to the whole population.
Non-probability access panel: A panel recruited by invitation or opt-in rather than random selection, typically corrected afterward with quota weighting rather than probability-based inference.
Quota weighting: Adjusting a non-probability sample's composition to match known population benchmarks on variables like age, gender, or region, without claiming the underlying selection was random.
BARB: The Broadcasters' Audience Research Board, the UK body whose hybrid people-meter panel and census-level return-path data underpin Ofcom's Media Nations viewing figures.
Time-use diary: A survey instrument in which a respondent reports a chronological account of activities across a fixed period, typically 24 hours, rather than answering recall questions about a specific medium.

Straight answers

Frequently asked questions

What is the Ofcom Media Nations report?

Media Nations is Ofcom's annual report on UK broadcast, streaming, and media consumption, built on BARB panel data, a hybrid of a people-meter panel and census-level online return-path and streaming data. The 2025 edition was published 30 July 2025 and reports both an as-viewed hybrid figure and a separate as-broadcast series for trend comparability.

How does the Reuters Institute's Digital News Report sample 48 countries?

It surveys online news consumers, over 97,000 in the 2025 edition, through the polling firm YouGov, using a non-probability online access panel weighted with nationally representative quotas rather than random probability sampling. Its own methodology page discloses that in several markets the sample represents only younger, English-speaking, online populations, not each country as a whole.

What is the American Time Use Survey?

ATUS is a continuous Bureau of Labor Statistics survey, running since January 2003, that samples one person aged 15 or older per household and has them report a full 24-hour diary of daily activities, including media and leisure time. It is a government labor statistics instrument, not a media-industry or advertiser-funded survey.

Could these three studies just be added together into one number?

No. They measure different objects on different populations: UK broadcast and streaming time, self-reported global news access habits, and US whole-day activity time respectively. Averaging or stacking their figures would misrepresent all three. What generalizes across them is the disclosure discipline, not the numbers.

Why does method disclosure matter more than the topline statistic?

Because the topline number is only interpretable in light of its method. A figure with no disclosed sample, weighting, or coverage caveat cannot be checked or generalized responsibly. All three instruments here name their method plainly, which is what makes them citable rather than merely quotable.

Provenance

Sources

Ofcom, "Media Nations 2025" and "Media Nations 2025: Annex Methodologies," published 30 July 2025 (established)
Reuters Institute for the Study of Journalism, University of Oxford, "Methodology," Digital News Report 2025 (established)reutersinstitute.politics.ox.ac.uk
Bureau of Labor Statistics, American Time Use Survey User's Guide (established)
Bureau of Labor Statistics, American Time Use Survey News Release (established)

Every figure above is attributed to a real, dated source and tagged with its evidence tier. Where a claim could not be verified to a primary source, it is not stated as fact.