publicmood

Method

What our numbers mean, and what they don't

Public Mood reports online sentiment among people who post publicly. Every number describes analysed posts, not people, and every analysis states its platforms, counts, period and caveats.

Analysed posts, not people

We write "63% of analysed posts were critical", never "63% of people think". Online posts come from people who chose to post, so they are not a representative picture of a population.

Why there is no margin of error

A margin of error requires a random sample. Instead we show how many posts were analysed per platform, the period, and how accurately posts were classified.

Public content only

We only analyse publicly available content: posts and comments anyone can see without being invited. Private messages, private profiles, closed groups and private channels are never included, on any platform.

Anonymous quotes

Quotes are shown with platform and date only. We never show usernames of private individuals, and only data needed for the analysis is kept.

Checked facts

All facts used to explain shifts, such as dates, results and events, are checked against credible sources such as official sources and established media.

Every analysis has a method section

Period, posts analysed per platform, scope per platform and caveats. See it at the bottom of the sample analysis.

News coverage

News coverage is based on written online news sources we can find through search. We may miss articles behind paywalls, in smaller outlets or in languages outside the analysis scope, so treat the outlet count as a minimum, not a complete list.