About the evidence

Sources & reuse rights

U.S. statistics lead this release. Each source keeps its own population, definition and reuse conditions.

NTIA / U.S. Census Bureau · U.S. participation

The U.S. social-network-use dataset powers our default explorer. It selects 75 published aggregate observations from NTIA’s November 2023 Internet Use Survey: a national estimate, demographic breakdowns, and the 50 states plus D.C.

The population is internet users aged 15 and older in the civilian noninstitutionalized population. The measure covers at least occasional use of online social networks during the past six months. Fieldwork was November 12–18, 2023. It does not measure individual platforms or all U.S. adults, and the available breakdowns do not support arbitrary combinations such as state by age.

Source: NTIA Analyze Table and documentation, November 2023 / socialNetworkUser / adultInternetUser. The source’s standard errors accompany the estimates.

NTIA access, publication dates and reuse

The import uses NTIA’s public aggregate CSV without credentials; it does not process individual survey responses. NTIA announced its 2023 explorer update on June 6, 2024. A later file update or our October 2026 evidence check does not turn these into 2026 observations. No published 2025 social-network-use table was established in this release.

Selected NTIA-produced federal statistics support versioned CSV downloads and original SocialUsage SVG charts. Exports retain the source, period, population and presentation notes. The rights basis is U.S. federal government statistical work, public domain in the United States; no agency logos, third-party source graphics or blanket worldwide licence are included.

NTIA release announcement · U.S. government works provision · NTIA government-work and third-party rights explanation

NCI HINTS · U.S. adult visit frequency

The adult social-media frequency dataset connects five national response categories from HINTS 7, collected March 25–September 16, 2024. It describes visiting a social-media site in the previous 12 months among U.S. civilian noninstitutionalized adults aged 18+. The source’s “Almost everyday” category is not an exact daily-use count.

The point estimates come from NCI’s supplied public codebook. Standard errors are SocialUsage analysis of NCI public-use data, calculated using 50 supplied replicate weights and the documented delete-one jackknife method. We retain missing and invalid item responses in the denominator, as the codebook does; the five substantive categories therefore do not sum to 100%.

Source: NCI HINTS 7 public-use package, updated August 2025, Public Codebook PDF page 80 (printed 79), SocMed_Visited / B12a. Questionnaire, PDF page 5 (printed 4) · Methodology, section 5.5.

HINTS analysis, access and reuse

The offline importer reads one survey item and its weights from the public package, verifies the published estimates and writes national aggregates. It does not publish respondent records or add unsupported demographic or historical breakdowns. The package revision, source hashes, missingness and calculation inputs are retained in each release.

NCI’s reuse guidance and its public-use analysis guidance support the selected aggregate facts and attributed original analysis. The dataset provides versioned aggregate CSV downloads; copied source graphics, agency logos and respondent redistribution are excluded. Exports preserve NCI attribution and identify SocialUsage’s uncertainty calculations. NCI does not endorse this presentation.

Pew Research Center · U.S. platform adoption

The Social Media Fact Sheet supports five limited platform briefs: Facebook, Instagram, Reddit, TikTok and YouTube. These are ever-use estimates for all U.S. adults aged 18+, including non-internet users.

Source: 2025 National Public Opinion Reference Survey, 5,022 adults, February 5–June 18, 2025; published November 20, 2025. Original topline (PDF) · Methodology.

Pew selection and reuse limits

The selection contains eleven source values: five overall estimates and two age groups for each of Instagram, YouTube and Reddit. The U.S. age-gap report calculates differences from those existing excerpts. Facebook and TikTok have no imported age or historical breakdowns.

Pew’s terms allow attributed use, including commercial use, while restricting principal-part republication, mirroring and unauthorized automated harvesting. They specify no numerical safe harbor. Our limited selection is an editorial reuse decision; it is not a complete Pew table library. We offer no Pew source files, CSV downloads, image exports or scheduled ingestion.

Pew Research Center bears no responsibility for the analyses or interpretations of the data presented here. The opinions expressed herein, including any implications for policy, are those of the author and not of Pew Research Center.

CDC YRBS · U.S. high-school frequency

The teen-frequency brief uses CDC’s 2025 Youth Risk Behavior Survey and its published, matching 2023 comparison. The measure is social-media use about once an hour or more often among students in grades 9–12 attending U.S. public, parochial and other private schools. CDC supplies the confidence intervals and the statistical comparison; these are not SocialUsage calculations.

The hourly measure remains separate from the broader several-times-daily measure in the older 2023 MMWR report. Neither threshold measures hours spent or covers all U.S. teenagers.

Source: CDC, Digital Behaviors, September 18, 2026; technical Tables 1d, 35 and 44. Original materials are available free of charge on CDC.gov. Use by SocialUsage does not imply endorsement by CDC, HHS or the U.S. Government.

CDC access, dates and reuse

These are reviewed published aggregates, not a live API or a new analysis of individual responses. The companion article is dated September 18, 2026; the technical tables cite publication year 2026. Exact 2025 fieldwork start and end dates were not established in the checked documents.

CDC’s agency-materials policy governs the selected CDC-authored statistical facts and requires credit, nonendorsement, unchanged meaning and free-original notice. We do not reproduce source photographs or agency logos. This editorial brief offers no dataset download or source-chart export.

Gallup · One U.S. teen time estimate

The time-spent brief cites Gallup’s source-reported mean across seven tested platforms among U.S. teenagers aged 13–19, surveyed June 26–July 17, 2023. The supporting report supplies an analytic sample and standard error. The figure remains a 2023 reported-time estimate, with Snapchat excluded; it is not an all-American average or a passive device measurement.

Source: Gallup’s original article, October 13, 2023; IFS/Gallup report, October 11, 2023, Figure 2 p8 and methods pp7–9.

Gallup citation and reuse limits

Gallup’s permissions guidance provides for accurately attributed citations and separately requires permission for reproducing its material. We make one limited factual citation in original prose, with scope and source links. There is no Gallup ingestion feed, open-data licence, chart reproduction, CSV download or image export.

The exact verbatim time-question wording was not located in the checked article and report. We label the known question description as a paraphrase. We do not sum individual platform estimates or infer a confidence interval that the cited figure does not supply.

Eurostat · Additional European comparisons

The European dataset and European explorer use the public Statistics API for isoc_ci_ac_i. The saved selection covers 2023–2025, all 27 EU countries and the EU aggregate, for individuals aged 16–74 and supported age bands.

Its private-purpose participation measure uses a three-month activity window and all individuals in each age group. It is kept separate from NTIA’s six-month measure among internet users aged 15+.

Eurostat access and reuse

Public API access needs no credentials. Eurostat’s reuse policy allows attributed commercial statistical reuse, subject to exceptions. Our selected EU CSV and SVG exports preserve credit, disclose changes and state that Eurostat is not responsible for the modified presentation. The source’s missing values and quality flags remain visible.

Company disclosures · Additional worldwide measures

The Pinterest and Snapchat briefs select operating metrics from original Q1/Q2 2026 releases, with two compatible quarters and company-defined regional figures. Pinterest reports monthly active users at quarter end; Snap reports a quarterly average of daily active users. Neither is a U.S. adult survey estimate.

Original releases: Pinterest Q2 2026 (PDF) · Snap Q2 2026. Each brief links its earlier-quarter evidence and definitions.

Company source access and reuse limits

These are attributed factual summaries in original wording, not connected ingestion APIs or licensed source datasets. Company-source CSV downloads, copied graphics and image exports are disabled. Public access to a release does not grant blanket republication rights. Pinterest terms · Snap terms.

The Reddit Q2 2026 disclosure is a definitional example in the user-count guide, not a connected operating-metric feed.

Researched for future coverage

Other U.S. and international sources
  • Pew teen research: a separate population and survey, requiring its own limited selection and review.
  • SEC EDGAR: original filing discovery; Companyfacts is not a universal MAU/DAU service.
  • Ofcom: UK research within each study’s population; third-party audience data may need separate permission.
  • World Bank / ITU: internet-use context, not a substitute social-media count.
  • DataReportal: research reference, with no assumed commercial ingestion licence.

U.S. source access and terms were checked October 1, 2026 UTC; later release timestamps are recorded separately. Initial source checks began September 30, 2026. Checks were assisted by automation and are not independent human editorial approval. Reference periods stay attached to the original findings.