Designing experiments

We are interested in how to design experiments to target specific hypotheses and model-fitting goals. We have explored sampling strategies to minimize uncertainty in parameter estimates in greenhouse mesocosm experiments.

This type of investigation is important, since investing in ecological experiments is often costly and inherently risky. Some things we consider to minimize those risks are listed below.

  • Model sensitivity to changing parameters
  • Constraints due to available time and labor
  • Information lost through reduced sampling strategies
  • Temporal variation in sensitivity and noise

Future work for this project involves planning experiments for different systems - we are in preliminary stages for a problem related to dung beetles.


Publication List

Virtual Office (drop ins welcome) - Schedule a Meeting (no need to ask) - - Math 117B (by appt. only)