Randomized intervention analysis for detecting non-random change and management impact: Dragonfly examples

The quasi-experimental approach of before-after control-impact (BACI) sampling can help decide when changes are due to human activities rather than natural variability. Detailed arguments for and against BACI designs and analytic methods are widespread in the literature, but far less attention has been paid to the mechanics of analyzing a BACI experiment. This paper demonstrates randomized intervention analysis with user-friendly software, where observations are paired in time before and after intervention. We provide examples using dragonfly count data in vegetation removal experiments.