Keep the core questions stable and record the events around each wave.
A clear point on the timeline
The Bank of Canada reduced its policy rate to 2.75% on 12 March 2025. For a research tracker, the announcement is part of the context in which interviews occur. It is not evidence that a particular attitude changed.
Protect the comparison
A tracker is most useful when its wording, population and measurement approach remain comparable. Changing the questionnaire every time the headlines change can make it difficult to distinguish a real movement from a design effect. Our preference is to retain a stable core and add clearly identified topical modules.
Keep a fieldwork record
Log significant events that occur before and during interviews. If an event falls inside a field period, report that fact and consider whether the sample supports a cautious timing comparison. Do not infer causation from a convenient sequence of dates. This retrospective sets out a measurement approach; it does not claim Northbound operated a tracker in March 2025.
Evidence note. This retrospective was first published on 25 October 2026. Its period label refers to the underlying event or data, not an earlier publication. Source material may include final figures published after the period covered. Northbound’s interpretation is identified separately from reported findings. Read our publication policy.