Established model
Non-response bias and total survey error
Robert M. Groves · 1989
Also known as Total survey error, Who answers is not who you asked
Every source of error in a survey, gathered into one frame so they can be traded off against each other.
Its place in the frameworkData Core›Data Collection
What it does
Coverage, sampling, non-response and measurement error all reduce accuracy, and they compete for the same budget. The insight that changed practice is that non-response bias, not sample size, usually dominates — which means a larger survey of the same self-selecting people buys precision around a wrong number.
- Reach for it when
- Before commissioning any survey, and when reading someone else’s.
- Where it stops
- Several of the errors cannot be measured from inside the survey itself. It tells you what to worry about rather than how much.
Robert M. Groves, Survey Errors and Survey Costs, Wiley, 1989.
Why it sits at Data Collection
Where the numbers come from: which sources, how they are captured, whether they can be trusted, and on what basis the organisation is entitled to hold them.
A model is only useful when you reach for it at the right moment. This one answers a question that arises here — so it is filed here, and nowhere else. These are the working areas it serves:
- The sourcesMaster-data practice on source-of-truth designation, and the reconciliation problem where several systems hold the same fact.
- InstrumentationMeasurement-design literature: the distinction between data that accumulates and data that is captured on purpose.
- Data qualityThe data-quality dimensions used in information management: completeness, accuracy, timeliness, consistency and validity.
- Consent and basisData-protection principles of purpose limitation and data minimisation, which constrain collection independently of technical capability.
What it touches elsewhere
Nothing in a business is decided on its own. A conclusion reached with this model at Data Collection lands in these other cores, whether or not anyone follows it there.
- KPI managementA measure is only as good as the collection behind it, and the definition has to be collectable.
- Data GovernanceWhat may be collected and how long it may be held is settled in governance.
- Omni CoreMuch behavioural data is collected at touchpoints, and instrumentation is part of designing them.
- Business CoreRegulatory obligations on data collection are compliance obligations like any other.
Filed at the same place
These answer questions that arise at Data Collection too. Where they disagree with this one, the disagreement is the useful part.
- The six data quality dimensionsQuality is not one property but several — completeness, uniqueness, timeliness, validity, accuracy, consistency.
Elsewhere in Data Core
These are other people’s models, named here so you can go to the source and use them properly. The Omnigoal is not affiliated with their authors and is not endorsed by them; nothing of theirs is reproduced here — no canvas, no diagram, no wording. Each is described in our own words, with the originator credited, because the framework is a place to put thinking, not a replacement for the people who did it. Model names and trademarks belong to their respective owners and are used here only to refer to the work itself.
All 125 models