OnCo
ideasIdea

Pre-specified sample sizes for animal studies; no more 'representative' experiments

Many mouse experiments use so few animals that the results are unreliable, and papers show one 'representative' result out of several tries. Funders should require proper sample-size planning.

Underpowered animal studies produce exaggerated effects when positive and are uninformative when negative. Sample-size justification is required by ARRIVE and by ethics committees in principle but is rarely checked. Funders and journals could require a pre-specified power calculation, the number of independent repeats performed, and reporting of all repeats rather than a representative one.

Hypothesis
Enforcement increases the median group size in published in vivo efficacy studies and reduces reported effect sizes towards those seen in adequately powered replications.
Rationale
Effect-size inflation from small samples is a mathematical certainty under publication bias; the fix is procedural and cheap.
What would test it
Adopt at two funders; audit group sizes, repeat reporting and effect sizes in funded publications before and after.
Maturity
early clinical
Who has to act
policy
Cost to try
Small (under $1M)
Years to first evidence
1
Bottlenecks it attacks

Connected

4top

Pages like this

not linked directly; found by shared links