Non-Sampling Error
Updated 2026-09-10
INTRODUCTION
English translation pending.
CORE DEFINITION
Non-sampling error covers every inaccuracy in a survey or study that does not come from the random selection of respondents: faulty measurement instruments, coverage gaps, non-response, interviewer effects, coding, and data entry mistakes. The core claim is that these errors are systematic rather than random, so increasing the sample size does not reduce them and may make a biased estimate look more precise. The qualification is that non-sampling error can rarely be eliminated, only controlled through design, pretesting, and follow-up, so its likely direction should be reported alongside the estimate.
SCAFFOLDING EFFECT
Reduce cognitive load
- Error inventory: list every step between the concept and the recorded value, and name its failure mode. - Precision check: ask whether the apparent accuracy comes from sample size or from a biased instrument. - Direction estimate: state which way the likely error pushes the result rather than only its size.
Anchor fast decisions
Random sampling error shrinks as the sample grows because independent errors cancel, but a bias in the instrument or the coverage does not cancel, it applies to every observation equally. Adding more observations therefore narrows the confidence interval around a value that is wrong. Because the bias is invisible in the interval, the study appears more reliable as the sample grows, which is why design controls matter more than sample size.
MINIMUM ACTION
In progress 0/1Practice this model in one real situation:
account_treeGenealogyexpand_more
menu_bookReferencesexpand_more
Source support: Explicit
- wiki.mbalib.comhttps://wiki.mbalib.com/wiki/%E9%9D%9E%E6%8A%BD%E6%A0%B7%E8%AF%AF%E5%B7%AEverified
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