RT Journal Article
SR Electronic
T1 Effect size and statistical power in the rodent fear conditioning literature – a systematic review
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 116202
DO 10.1101/116202
A1 Clarissa F. D. Carneiro
A1 Thiago C. Moulin
A1 Malcolm R. Macleod
A1 Olavo B. Amaral
YR 2017
UL http://biorxiv.org/content/early/2017/09/13/116202.abstract
AB Proposals to increase research reproducibility in basic science frequently call for focusing on effect sizes instead of p values, as well as for increasing statistical power. To study how these two concepts are taken into account in behavioral neuroscience, we performed a systematic review to evaluate the distribution and description of effect sizes and statistical power in studies on rodent fear conditioning learning. Amnesia caused by memory-impairing interventions was nearly always partial, and mean statistical power to detect the average effect size observed in well-powered experiments was 65%. Effect size correlated with textual descriptions of results only when findings were non-significant, and neither effect size nor power correlated with article citations. In summary, effect sizes and statistical power have a wide distribution in the literature on rodent fear conditioning acquisition, but do not seem to have a large influence on how results are described or cited.