Minimizing the cost of environmental management decisions by optimizing statistical thresholds

Abstract

Environmental management decisions are prone to expensive mistakes if they are triggered by hypothesis tests using the conventional Type I error rate (a) of 0.05. We derive optimal a-levels for decision-making by minimizing a cost function that specifies the overall cost of monitoring and management. When managing an economically valuable koala population, it shows that a decision based on a ¼ 0.05 carries an expected cost over $5 million greater than the optimal decision. For a species of such value, there is never any benefit in guarding against the spurious detection of declines and therefore management should proceed directly to recovery action. This result holds in most circumstances where the speciesÕ value substantially exceeds its recovery costs. For species of lower economic value, we show that the conventional a-level of 0.05 rarely approximates the optimal decision-making threshold. This analysis supports calls for reversing the statistical Ôburden of proofÕ in environmental decision-making when the cost of Type II errors is relatively high.

Monitoring freshwater fish in Florida lakes using electrofishing: Lessons learned

“…However, these same individuals are often frustrated because no 20-year historical record exists for an aquatic system to help make management decisions and achieve desired management goals. Thus, monitoring is a necessity that has to be accomplished as cost effectively as possible yet be rigorous enough to provide information needed for management (Field et al 2004). …”

Evaluating the efficacy of adaptive management approaches: Is there a formula for success?

“…Despite the differences between schools discussed above, a recurrent theme in all adaptive management approaches is the ongoing monitoring of measurable objectives while also imple- menting selected actions (Walters and Holling, 1990;Field et al, 2004;Gerber et al, 2005). With active learning and continuous monitoring, uncertainty decreases and forecast management outcomes are more easily predicted (Walters, 1986(Walters, , 1997.…”

L’assurance qualité (AQ) dans les programmes de surveillance des forêts : activités, problèmes et perspectives pour l’Asie de l’Est et l’Europe

“…When the hypothesis testing is of concern, the statistical power (power = 1.0 minus the probability of a Type II error) of the test becomes important (for a discussion about Type I and Type II errors see e.g. Di Stefano, 2003;Field et al, 2004;Mapstone, 1995;Peterman, 1990). However, the power depends on effect size (the change the monitoring is requested to detect), survey design and statistical test applied, sample size and the Type I error rate.…”