What counts as “news”
Statements such as “most people get their news from social media” depend on what the surveyor meant by news and by getting it. A person who sees a headline in a feed while scrolling, a person who opens a news app on purpose and a person who watches a full evening bulletin are all “consuming news” under a loose definition, but their experiences differ greatly.
Researchers typically make choices about at least four things:
- Topic. Is news limited to politics and current affairs, or does it include sport, entertainment, local events and weather?
- Source. Does a post from a friend about a news event count, or only material from a recognised news organisation?
- Depth. Does a glance at a headline count, or does it require reading an article?
- Intent. Is incidental exposure included, or only deliberate seeking?
A survey that counts any brush with a headline will usually show higher and more social-media-heavy results than one asking only about deliberate use. Neither is wrong, but they measure different things.
How news use is usually measured
Self-report surveys. The most common approach is to ask people which sources they used in the last week, how often, and which they consider main. Questions can be asked by phone, in person or online. Sample design matters: who is invited, who responds and how results are weighted all shape the outcome, which is why survey panels and non-response bias are relevant background.
Diary and time-use methods. Some studies ask people to record activities over a day or week. These can reduce reliance on memory but are burdensome, so participants may be unusual in their willingness to complete them. Time use surveys are explored in their own right elsewhere.
Passive tracking. With permission, software on phones or browsers can record which sites and apps a person uses. This avoids recall problems, but it usually covers a limited set of devices, may miss television, radio and print, and generally involves volunteers who agreed to be monitored.
Platform and publisher analytics. News organisations and platforms can count page views, clicks and watch time. These describe activity on their own services, not people’s whole news diet, and a click does not show that anything was read.
Why people misreport
Self-reports about news have several well-known weaknesses.
Recall errors. Many news encounters are brief and unmemorable. Asked to estimate time spent or number of sources, people typically rely on rough impressions, and they may forget incidental exposure entirely.
Social desirability. Following the news is widely seen as part of being an informed citizen. Some respondents may therefore overstate their attention, particularly to serious topics, a pattern discussed in coverage of social desirability and honest answers. Others may understate time spent on entertaining or lower-status content.
Confusing the source with the route. A person who read an article that a friend shared on a social app may say they got the news from the publisher, or from the app. Responses can depend on how the question separates the platform from the outlet.
Mixing up habit and memory. People asked about a “typical” week describe an idealised routine, not what they did. Major events can also distort recall, since a dramatic week may make general habits seem more news-heavy than usual.
Question wording. Changing a list of response options, or the order in which they appear, can shift results, as described in the discussion of question wording.
Studies that compare self-reports with tracked behaviour often, though not always, find that the two do not line up neatly. The general gap between self-reported behaviour and observed behaviour is a recurring theme in trend data.
Why the headline findings shift
Several features of the online environment complicate comparisons over time.
First, the line between news and other content has blurred. Feeds mix journalism, commentary, entertainment and personal posts, and recommendation algorithms decide what appears, so respondents may not be clear which of their exposure counts.
Second, survey questions and categories change as new platforms appear. A series that tracked a handful of sources a decade ago may now list many more, making direct comparison difficult.
Third, different surveys sample different populations. Online surveys may reach people more comfortable with the internet, while telephone or face-to-face surveys can reach different groups. Results often differ between countries partly because of methodology and not only because of real differences in habits.
Fourth, claims that news use has moved to a particular platform describe the share of respondents mentioning it, not the share of total news time. A platform can be used by many people for a small part of their news and still look dominant in a “do you ever use it?” question.
Reading a news habit statistic
When a figure is quoted, helpful questions include:
- Was it a survey, a tracking study or platform analytics?
- How was news defined, and was incidental exposure counted?
- Does the figure refer to the proportion of people using a source at all, their main source or time spent?
- Who was sampled, and how were results adjusted?
- Is the comparison across years using the same questions?
Findings about news habits are best treated as approximate pictures of what people say and do under specific definitions. They can show broad direction, such as the growing presence of digital sources, more reliably than they can show precise levels or the reasons behind them.