Environmental Monitoring

Environmental Monitoring

This can be considered in a number of ways:

  • You might want to use explicit decision-making strategies that link your results to decisions (such as decision trees);
  • You can use a statistical analysis that links the results to action more dynamically ( Bayesian statistics, or "run control" charts);
  • You may need to monitor implementation of a program as well as sampling the environment for the effects (ie you need to know whether any "failure" is due to the program design or the lack of implementation).

The first of these helps you define how well you understand your system and the kinds of risks that are involved in different options.

Bayesian statistics are often used in the analysis of decision trees. Sometimes called "subjectivist" statistics they do allow you to build on what you know (however imprecisely) or believe based on the feedback from your studies. In comparing the actual results with your current knowledge ("posts" versus "priors") you can see iteratively whether your confidence in your knowledge/beliefs is increasing or decreasing and can take appropriate action as needed rather than waiting for the "experiment" to be completed. This fits comfortably with the concepts of adaptive management.

While the last may seem self-evident many programs do not explicitly require measures of effectiveness, but if you don't know if a policy or program is being implemented, how can you interpret any (lack of) environmental effects? You may be able to build in an assessment of all actions but more often you may need to do randomised surveys, which will require its own experimental design!

How do I get the best outcomes if I can't design an ideal program?


Hosted by Keysoft Pty Ltd