Reading the measurement
Screen time is not the same as social-media time
Before comparing two totals, check what each one counts.
A screen-time total cannot automatically tell you how long someone spent on social media. It may include other activities, use a different reporting period, or cover a different set of devices. A self-reported social-media estimate also depends on the question's exclusions and the people surveyed.
Measurement guide · U.S. survey examples · Fictional arithmetic illustrations, not population estimates
For example, CDC/NCHS's NHIS-Teen screen-time question includes programs, games, internet and social media but excludes schoolwork. Primack and colleagues' young-adult question asks about personal social-media time and excludes work. These are different measures.
One day, two valid totals
Imagine one person uses one device for 60 minutes of social media and 120 minutes of other screen activities. Assume the categories are mutually exclusive, with no simultaneous activity. The total is 180 minutes; only the first part is social-media time.
One fictional day · minutes
- Social-media activity: 60 min
- Other screen activity: 120 min
Source: Social Usage illustration and methods. These are invented inputs for explaining arithmetic, not observed behavior or recommended limits.
Read the example table and calculation
| Activity | Minutes |
|---|---|
| Social-media activity | 60 |
| Other screen activity | 120 |
| Total | 180 |
60 + 120 = 180 minutes. This subtraction works only because the example supplies a compatible total and a complete, non-overlapping breakdown.
A general internet or video category does not, by itself, identify social-media activity. Define the activities you want to count before comparing totals. Do not estimate one population's “other screen time” by subtracting a social-media average from a different survey's screen-time figure.
Two screens can occupy the same clock time
In a second fictional example, a phone session runs from minute 0 to 30, while a tablet session runs from minute 20 to 50. Adding the session durations gives 60 device-minutes. The time during which at least one device is in use is 50 elapsed minutes. The sessions overlap for 10 minutes.
Minutes from the same starting clock
Source: Social Usage illustration and methods, not a description of how a particular operating system combines devices.
Read the session table and formula
| Device | Start | End | Duration |
|---|---|---|---|
| Phone | 0 | 30 | 30 |
| Tablet | 20 | 50 | 30 |
Sum each end minus start for device-minutes. Merge overlapping intervals for elapsed time. Here, 60 − 10 = 50 minutes.
This is why a sum of app or device durations needs an explicit overlap rule. It does not establish that any published survey overcounts by this amount. If a report does not explain how it handles simultaneous use, keep that limitation visible.
Real questions draw different boundaries
| Measure | Who and when | What is counted |
|---|---|---|
| NHIS-Teen screen use | U.S. teens aged 12–17; July 2021–December 2023 | Most-weekday screen activity across several devices and activities; schoolwork excluded. Ordered hour categories with an open-ended top category. |
| Young-adult social-media use | U.S. noninstitutionalized adults aged 19–32; October–November 2014 | Self-reported personal use on an average day; work excluded. Hours and minutes requested. |
Neither definition is an all-adult device log. Their age ranges, periods, activity coverage and response formats differ. This guide makes no numerical comparison between the studies. The time-spent guide keeps the separate source findings and their limitations beside the numbers.
Do not turn the highest screen-time category into an exact duration. In NHIS-Teen, “4 or more hours” has no stated upper bound. Treating every response in that category as exactly four hours would add an assumption and cannot recover an observed mean.
Source: CDC/NCHS, Methods → Measures → Screen time. Original article is freely available; no CDC endorsement of this guide.
A short comparison checklist
- Activity: all screen activities, selected platforms, or personal social-media use?
- Coverage: one device or several, apps or websites, work or school included?
- Window: one dated day, most weekdays, an average day, or a weekly average?
- Method: reported answers or recorded sessions; if sessions overlap, how are they combined?
- People and summary: the same ages and population? A mean, median, category share or individual total?
If those answers do not match, explain the difference before presenting numbers together. A personal total is not a population average, and an older survey does not measure today's behavior.
Sources, dates and reuse
Source dates and what was checked
CDC article published 2025-07-10; data collected July 2021–December 2023. Methods and the screen-time question were checked on 8 October 2026 UTC using a retained connector response from earlier that day. Primack and colleagues' article was published 2017-03-06; its October–November 2014 question definition is reused from the reviewed 6 October evidence, with no fresh article retrieval. These review dates do not update the survey periods.
No health-outcome associations, new population estimates, confidence intervals or causal effects are reported here. The two charts are original hypothetical examples; their minute inputs, assumptions and formulas are recorded and checked deterministically. They do not validate a device product or assess anyone's behavior.
Attribution and reuse boundaries
CDC/NCHS method facts are attributed to the free original article. The historical social-media question keeps the existing source-specific citation scope. Source graphics, document copies, respondent data and product instructions are not reproduced. No dataset, chart download or embed is offered. This is an explanatory guide, not a dataset or a screen-time recommendation.