TempoLife › Features › Sample size, power and the study of twelve people
Sample size, power and the study of twelve people
Sample size, power and the study of twelve people: pages on what statistical power means, why underpowered studies do not just fail but actively exaggerate…
This launch guide maps the topic, the source base and the practical questions. Check the linked official authority before acting on a rule, deadline, price or safety requirement.
What this guide covers
Pages on what statistical power means, why underpowered studies do not just fail but actively exaggerate (the winner's curse), how confidence-interval width shrinks with n, margin of error in surveys, the effect of dropouts, crossover and repeated measures, cluster designs, and why nutrition research is so full of n=12. Two calculators run both directions: how many people you would need to detect a given effect, and what size of effect a study of n could ever have found.
Thirteen pages, and the "is a study of 20 people reliable" query is asked in every language on the list.
How to use it
Use Sample size, power and the study of twelve people as a structured learning path: begin with the vocabulary, follow the evidence map, and keep the uncertainty labels attached to claims as you compare sources.
The page keeps the main question, its evidence base and its limits together in one canonical place. That makes it useful as a quick orientation and avoids splitting a broad subject into shallow pages before each narrower answer has enough verified material of its own.
Search vocabulary in four languages
Terminology changes across borders. These are example phrases people use for the same subject in Estonian, Finnish, Russian and English; they are included as a practical vocabulary aid, not as hidden keywords.
| Language | Example phrase |
|---|---|
| Estonian | valimi suurus uuringus |
| Finnish | otoskoko ja tutkimuksen luotettavuus |
| Russian | размер выборки исследования сколько нужно |
| English | is a study of 20 people reliable |
Evidence and provenance
The evidence map below records the primary, official or openly licensed material chosen for this guide. It is deliberately visible so a reader can tell the difference between a rule published by an authority, a measured dataset and an editorial explanation.
Source: none, computed from standard power formulas; Button et al. 2013 "Power failure" cited. · checked 2026-09-09
Source names are preserved even where an authority changes its web address. No figure is added to this launch page unless it is already present in the cited source description or can be computed from a named dataset.
Scope and boundaries
/experiments supplies protocols for self-experiments; this is the arithmetic of how many people or how many days it would take.
Connected TempoLife guides
The following paths are the intentional hand-offs named in the brief. They keep adjacent subjects separate while making the next useful reference easy to find.
Research
Continue to the related TempoLife reference.
Experiments
Continue to the related TempoLife reference.
Why it belongs on TempoLife
/research, /experiments.
Frequently asked questions
What does this guide cover?
Pages on what statistical power means, why underpowered studies do not just fail but actively exaggerate (the winner's curse), how confidence-interval width shrinks with n, margin of error in surveys, the effect of dropouts, crossover and repeated measures, cluster designs, and why nutrition research is so full of n=12. Two calculators run both directions: how many people you would need to detect a given effect, and what size of effect a study of n could ever have found. Thirteen pages, and the "is a study of 20…
What evidence does it use?
none, computed from standard power formulas; Button et al. 2013 "Power failure" cited.
Where does this guide stop?
/experiments supplies protocols for self-experiments; this is the arithmetic of how many people or how many days it would take.
Turn the reference into a routine
TempoLife brings food, movement, sleep and everyday health habits into one private place when you want to track what happens next.