Method & limitations

What a snapshot is

A pulsur snapshot measures public online conversation about transport in a city: posts and comments that people published publicly, collected, labelled and then counted. It shows what is being said, how it feels, and who is saying it.

What it is not

It is not a survey. The people in it chose to post, which is not the same as a representative sample of a city’s residents. Groups who post rarely are under-represented, and a single event can dominate a month. Read a snapshot as a signal worth investigating, not as a measurement of public opinion.

How a snapshot is built

  1. Collection. Public posts and comments about transport in the city, over the period shown on the snapshot itself.
  2. Labelling. Each one is labelled with a mode, a topic, an emotion and, where it is clear, the kind of person speaking.
  3. Filtering. Your question narrows the set to one city, one mode and one topic. The snapshot shows how many remain at each step.
  4. Counting. Shares of emotion, persona mix, volume by month and the words used most are counted from that set.
  5. Writing. The short insight and the quotes are written from those labelled summaries, and reviewed by us before they are published.

The 15-conversation minimum

We only show insight where at least 15 relevant posts and comments match the city, mode and topic. Below that the snapshot says so and shows nothing else, because a handful of posts cannot support a claim about a city. Between 15 and 100 we mark the sample as small, and the patterns should be read as directional.

Emotion

Emotions are placed on a circumplex: how positive or negative the feeling is, and how activated. Twelve emotions are plotted. Posts with no clear emotion are counted and named under the chart, but not plotted, so the percentages you see are shares of the emotionally expressive posts rather than of everything collected.

Quotes

Quotes are reworded, never reproduced. Each one is a paraphrase of a labelled summary, which keeps the meaning while making the original post hard to trace back to a person. If you need verbatim evidence, that belongs in a tailored study with its own ethical review.

What a tailored study adds

Lines, stops and hours; change over time and around specific events; persona by emotion; comparisons with peer cities; continuous monitoring; and the evidence behind each finding.

Questions

If a number in a snapshot does not look right to you, tell us. We would rather hear it.