(f) “Which individuals would disapprove (i.e., think it was a bad
idea)?”, and (g) “Are there any other individuals or groups of people
who would approve or disapprove of you driving over the speed
limit?”. Control beliefs were probed by asking “What things (i.e.,
factors or circumstances)?” would make the target behaviour (h)
more and (i) less likely and (j) whether there were other things
that would make the target behaviour more or less likely. Two
raters independently coded the generated beliefs. Coding agreement
ranged from 89% to 95% across the four violations studied.
Commonly identified beliefs (identified by more than 3 participants),
were used to populate the belief questionnaires in the main
study (see Tables 1–4).