← All studies

Ofcom Children and Parents Media Literacy Tracker 2025/26

Study details from the Studies index, with sources and citations for each field.

Geo
UK
Year
2026
N
3,426
Ages
8-17
Per-app coverage
Harms
Offline

Notes · Non-probability; Child self-report frame: ages 8–17, n=3,426. The full parent/child survey contains 5,008 sampled child units, including 1,582 parent-only younger units. Under-13 subgroup ages 8–12: n=1,665; ages 8–9 n=649 (weighted base 567), ages 10–12 n=1,016 (weighted base 889). QC20 is in the child core. Bullying is in randomized Module E (n=1,724), with consenters n=1,342 and personal victims n=335; weighted bases 1,438 / 1,112 / 259. Parent QP20 ages 3–7 is outside this entry.

References

  1. Ofcom / BMG Research (2026) Main-set Tables— 2025/26 Main-set Data Tables

    Ofcom / BMG Research (2026). Children and Parents Media Literacy Tracker 2025/26: Data Tables, Main Set. Updated 21 May 2026.

    crosstabs
    Retrieved on:
    2026-10-03
    Archived source:
    Download XLSX
    File verification
    File verification details

    SHA-256: 092b45299bf4b41f32e21843e3e37e33071ab7d5eb10764d3e9ce2f8c7397b09

  2. Ofcom / BMG Research (2026) Questionnaire— 2025/26 Questionnaire

    Ofcom / BMG Research (2026). Children and Parents Media Literacy Tracker 2025/26: Questionnaire.

    instrument
    Retrieved on:
    2026-10-03
    Pages:
    105
    Archived source:
    Download PDF
    File verification
    File verification details

    SHA-256: 7e6b1406bc69ed975c2dc09f56c5404c7e12c1a04998c4d9837297f982fa6047

  3. Ofcom (2026) Children and Parents report— Report

    Ofcom (2026). Children and Parents: Media Use and Attitudes Report. Published 21 May 2026.

    report
    Retrieved on:
    2026-10-03
    Pages:
    63
    Archived source:
    Download PDF
    File verification
    File verification details

    SHA-256: 7af93c19017e6b52b278a31b86b1ca8e369dbf3ac4336b35cd144c6af3e46cae

  4. Ofcom / BMG Research (2026) Technical Report— 2025/26 Technical Report

    Ofcom / BMG Research (2026). Children and Parents Media Literacy Tracker 2025/26: Technical Report. Published 20 March 2026.

    methodology
    Retrieved on:
    2026-10-03
    Pages:
    12
    Archived source:
    Download PDF
    File verification
    File verification details

    SHA-256: 4607cb56318ea2ddd9dcea114ebf8ea77e2099f12f406dc629e72dc608bd5a26

  5. Ofcom / BMG Research (2026) Subset Tables— 2025/26 Subset Data Tablesadditional source

    Ofcom / BMG Research (2026). Children and Parents Media Literacy Tracker 2025/26: Data Tables, Subset. Updated 21 May 2026.

    crosstabs
    Retrieved on:
    2026-10-03
    Archived source:
    Download XLSX
    File verification
    File verification details

    SHA-256: 8d6919e170158cbda92f8f2ea8062ee152710d5ce3a2769c73ea71a4a7e280b9

Study details & methodology

Name
“Ofcom Children and Parents Media Literacy Tracker 2025/26”
Country
“Ofcom is keen to understand how children across the UK are using and thinking about media”
Publication year
“Published 21 May 2026”

Publication year 2026; 2025/26 is the wave name.

Fieldwork dates
“Fieldwork for online panel started on November 19th 2025 and finished on February 22nd 2026.”“Fieldwork for face-to-face started on December 4th 2025 and finished on February 20th 2026.”

Online: 19 November 2025–22 February 2026. Face-to-face CASI: 4 December 2025–20 February 2026.

Sample size
“3426 with children themselves aged 8 to 17”

Child self-report sample: 3,426 aged 8–17, within the full parent/child survey of 5,008. QC20 is in the child core. Bullying questions are in randomized Module E and require consent; the personal-victim follow-up has a smaller base. Parent-proxy responses for younger children are outside this entry. Under-13 child sample: 1,665 (649 aged 8–9 and 1,016 aged 10–12). Bullying consent base: 1,342; personal-victim follow-up: 335. These are unweighted bases.

Age range
“SECTION C TO BE SHOWN TO ALL CHILDREN AGED 8-17 (ALLOCATEDCHILD = 17-36)”

The entry covers child self-reports aged 8–17. The under-13 age bands in the published tables are 8–9 and 10–12.

Sampling
“The sample for the online version of the survey was drawn using an online panel blend.”“Quotas were set for both face-to-face and online panels to ensure targets for child age, gender, region, social grade, and urbanity.”

Non-probability quota recruitment. Online panel blend plus face-to-face CASI is a mixture of interview modes, not evidence of mixed probability/non-probability sampling.

Harms measured

Bullying
“People can be nasty or hurtful. It could be behind someone’s back, to their face, through calls or texts. It could be by being nasty through social media, games or other websites. It could be by calling people names, leaving them out, or through sharing photos or videos that upset them. It could be threatening to hurt or actually hurting them. It could be done on purpose or as a joke that goes too far.…” “And has this happened to you in the last 12 months?”Response options: Yes / No / Don’t know / Prefer not to say

QC53B asks about the respondent’s own experiences in the past 12 months. Asked of children randomly assigned to Module E, after consent. The definition includes online and in-person conduct. The questionnaire labels the consent question inconsistently.

Underage Use
“Below is a list of some of the apps/ sites that you said earlier that you use. Some can be used by anyone, but others need you to set up a profile first. This may involve choosing a username, password and a picture for your profile. You might have done this on your own or someone might have helped you do it.…” “Please click on those where you have your own profile.”

QC20 child own-profile item, combined with age and app-specific policy eligibility. The header refers to QCNEW2, but option routing on p59 uses QCNEW1. Own-profile ownership is not automatically a policy violation; the index counts 20 named apps.

Offline or broader-context measurement

Bullying
“When somebody was nasty or hurtful to you did it happen in any of these ways? Please choose as many of the options below that apply…” “Face to face”

QC55 asks respondents who reported a personal experience in the past 12 months where it happened. They may select more than one answer, including face to face.

Question sources

COBA_2025_26_QC20
“QC20. Below is a list of some of the apps/ sites that you said earlier that you use. Some can be used by anyone, but others need you to set up a profile first. This may involve choosing a username, password and a picture for your profile. You might have done this on your own or someone might have helped you do it. Please click on those where you have your own profile. Please choose all the apps / sites that you have a profile on”Response options: BeReal / DailyMotion / Discord / Facebook / Imgur / Instagram / Kik / Pinterest / Reddit / Signal / Snapchat / Threads / TikTok / Triller / Tumblr / Twitch / Vimeo / WhatsApp / Wink / X/ Twitter / YouNow / YouTube / YouTube Kids / Yubo / I do not have a profile on ANY of these / Don’t know

QC20 asks which listed services the child has an own profile on. Options depend on earlier app-use answers; the option instructions refer to QCNEW1, while the heading refers to QCNEW2.

COBA_2025_26_QC53B
“QC53B. Please read the examples of upsetting things that can happen to people and then answer the TWO questions underneath. People can be nasty or hurtful. It could be behind someone’s back, to their face, through calls or texts. It could be by being nasty through social media, games or other websites. It could be by calling people names, leaving them out, or through sharing photos or videos that upset them. It could be threatening to hurt or actually hurting them. It could be done on purpose or as a joke that goes too far.…” “And has this happened to you in the last 12 months?”Response options: Yes / No / Don’t know / Prefer not to say

QC53B asks whether the respondent personally experienced the described behavior in the past 12 months. The definition includes online and in-person conduct.

COBA_2025_26_QC55
“QC55. When somebody was nasty or hurtful to you did it happen in any of these ways? Please choose as many of the options below that apply”Response options: Face to face / By text or messaging apps (like WhatsApp) / On social media sites or apps (like TikTok, Instagram, Snapchat) / Through other websites or apps / Through phone calls / Through video calls (like FaceTime or Zoom or WhatsApp video calling) / In online games / Through some other way – (please type in) / Don’t know / Prefer not to say

QC55 asks respondents who reported a personal experience in the past 12 months where it happened. They may select multiple answers. App names are examples within communication categories, rather than separate app measures.

Platforms measured

Platform list
“Please click on those where you have your own profile.…” “1 BeReal QCNEW1=2”“2 DailyMotion QCNEW1=3”“3 Discord QCNEW1=4”“4 Facebook QCNEW1=5”“6 Imgur QCNEW1=11”“7 Instagram QCNEW1=12”“8 Kik QCNEW1=13”“9 Pinterest QCNEW1=16”“10 Reddit QCNEW1=17”“11 Signal QCNEW1=18”“12 Snapchat QCNEW1=20”“13 Threads QCNEW1=22”“14 TikTok QCNEW1=23”“15 Triller QCNEW1=24”“16 Tumblr QCNEW1=25”“17 Twitch QCNEW1=26”“18 Vimeo QCNEW1=28”“19 WhatsApp QCNEW1=29”“20 Wink QCNEW1=30”“21 X/ Twitter QCNEW1=31”“22 YouNow QCNEW1=32”“23 YouTube QCNEW1=33”“25 YouTube Kids QCNEW1=36”“24 Yubo QCNEW1=34”

The questionnaire names 24 distinct profile-ownership options; the index counts 20 eligible apps for under-13 measurement. YouTube and YouTube Kids are separate options.

Published app coverage
“QC20. Below is a list of some of the apps/ sites that you said earlier that you use. Please click on those where you have your own profile”

Main-set Tables worksheet, Table 120 (B3970–B4063), reports profile ownership for ages 8–9 and 10–12. Some cells use *% or -*% in place of a numeric estimate; the table does not state the threshold for these markers. These options remain part of the question regardless of whether a percentage is published.

Per-app measurement by harm

Bullying
“QC55. When somebody was nasty or hurtful to you did it happen in any of these ways? Please choose as many of the options below that apply…” “Face to face”“By text or messaging apps (like WhatsApp)”“On social media sites or apps (like TikTok, Instagram, Snapchat)”“Through other websites or apps”“Through phone calls”“Through video calls (like FaceTime or Zoom or WhatsApp video calling)”“In online games”“Through some other way – (please type in)”“Don’t know”“Prefer not to say”

QC55 asks about types of communication, including face to face, messaging, social media and video calls. App names are examples within these categories, so the answers do not identify individual apps.

Underage Use
“Please click on those where you have your own profile.…” “1 BeReal QCNEW1=2”“2 DailyMotion QCNEW1=3”“3 Discord QCNEW1=4”“4 Facebook QCNEW1=5”“7 Instagram QCNEW1=12”“8 Kik QCNEW1=13”“9 Pinterest QCNEW1=16”“10 Reddit QCNEW1=17”“11 Signal QCNEW1=18”“12 Snapchat QCNEW1=20”“13 Threads QCNEW1=22”“14 TikTok QCNEW1=23”“16 Tumblr QCNEW1=25”“17 Twitch QCNEW1=26”“18 Vimeo QCNEW1=28”“19 WhatsApp QCNEW1=29”“20 Wink QCNEW1=30”“21 X/ Twitter QCNEW1=31”“22 YouNow QCNEW1=32”“24 Yubo QCNEW1=34”

QC20 has 24 named ownership options; conservative Underage Use coverage is 20 apps. YouTube permits supervised child routes and YouTube Kids is designed for children; QC20 does not distinguish these permitted routes. Imgur’s current terms permit supervised under-13 use, but an applicable fieldwork-period account prohibition was not verified. Triller’s applicable operator age rule could not be verified. Some operator evidence is current/undated rather than an archived fieldwork version, so this does not prove that each profile violated the exact fieldwork-period terms. WhatsApp’s 17 March 2025 terms apply to fieldwork; its parent-managed route begins after fieldwork on 17 August 2026. QC55 channel categories add no individually measured apps.