Sampling Frame
Updated 2026-08-15
INTRODUCTION
English translation pending.
CORE DEFINITION
A sampling frame is the operational list from which sample units are drawn: a register, a directory, a map, or a customer database. It matters because probability sampling requires that every population member have a known, nonzero chance of selection, which only a frame can supply. Coverage error arises when the frame omits members of the target population, includes ineligible ones, or lists the same unit twice. The qualification is that the frame is not the population; it is a proxy, and the gap between them must be measured rather than assumed away. Frames also age, so units that move or appear after listing drop out of scope.
SCAFFOLDING EFFECT
Reduce cognitive load
- Name the frame: write down the actual list being sampled, not the population you intend. - Measure the gap: compare frame contents against the target population to size the omissions. - Repair the edges: add coverage for missing units or adjust the estimate, and deduplicate entries.
Anchor fast decisions
Every inference from a survey is really an inference from the frame, because units outside it can never be selected. Omitted units are therefore weighted at zero, which biases estimates whenever omission correlates with the measured variable, as it usually does. Duplicate listings inflate the selection probability of those units and distort weights. The error is structural: it enters before any data are collected, and no adjustment applied later can recover the unlisted members. Frame quality therefore bounds the credibility of everything downstream.
MINIMUM ACTION
In progress 0/1Practice this model in one real situation:
account_treeGenealogyexpand_more
menu_bookReferencesexpand_more
Source support: Explicit
- en.wikipedia.orghttps://en.wikipedia.org/wiki/Sampling_frameverified
PRIVATE NOTES · Only visible to you
SAVED Q&A
ENTRY Q&A · Private saving available
Ask with a clear boundary
thinkingmodels answers from published entry context only.
Your question is sent to thinkingmodels. The answer uses public entry context only.
RELATED MODELS