Premium
epower : An r package for power analysis of Before‐After‐Control‐Impact (BACI) designs
Author(s) -
Fisher Rebecca,
Shiell Glenn R.,
Sadler Rohan J.,
Inostroza Karina,
Shedrawi George,
Holmes Thomas H.,
McGree James M.
Publication year - 2019
Publication title -
methods in ecology and evolution
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 3.425
H-Index - 105
ISSN - 2041-210X
DOI - 10.1111/2041-210x.13287
Subject(s) - computer science , statistical power , r package , bayesian probability , replication (statistics) , sampling (signal processing) , range (aeronautics) , monte carlo method , hierarchy , sample size determination , statistical hypothesis testing , data mining , sample (material) , statistical model , statistics , machine learning , artificial intelligence , mathematics , engineering , computational science , chemistry , filter (signal processing) , chromatography , economics , market economy , computer vision , aerospace engineering
Before‐After‐Control‐Impact (BACI) designs are widespread in environmental science, however their implicitly hierarchical nature complicates the evaluation of statistical power. Here, we describe epower , an r package for assessing statistical power of BACI designs. The package uses Bayesian statistical methods via the r ‐package INLA to fit the appropriate hierarchical model to user supplied pilot survey data. A posterior sample is then used to build a Monte Carlo simulation to test statistical power specifically for the Before/After × Control/Impact interaction term in the BACI model. Power can be assessed for any number of user‐specified effect sizes for the existing design, or across a range of levels of replication for any part of the sampling design hierarchy. The package offers a user friendly robust approach for assessing statistical power of BACI designs whilst accounting for uncertainty in parameter values within a fully generalized framework.