Property statistics are useful only when their definitions match the question being asked. Keep the source, reference period and measurement alongside every number in a research file.
Identify the measure
Distinguish asking prices from completed-sale prices, and advertised rents from rents actually paid. The ONS private-rent methodology explains differences in its underlying data across the UK. Two series with similar names may therefore describe different parts of the market.
Check geography and property mix
A regional average is not a valuation of a particular street. Ask whether the sample covers comparable sizes, types and conditions. Note any small sample or missing period before drawing a conclusion from an apparent change.
Keep the time basis consistent
Record when the data were collected and when they were published. Do not combine figures from different periods without explaining the mismatch. Where a source revises previous estimates, retain the version date so that another reader can understand your calculation.
Separate observation from prediction
A measured change describes the period covered by the data. It does not establish what a particular home will earn or sell for later. Label forecasts and personal assumptions separately, and show the arithmetic behind any derived yield. If the underlying evidence cannot be traced, leave the claim out rather than replacing it with an invented estimate.
Editorial review: 14 October 2026. This original Property Perpetuity guide replaces the imported text; the original archive publication date is retained.