Improve the Reproducibility of Experiments Using Randomized Block Designs


Too many experiments involving laboratory animals are giving results that cannot be reproduced, resulting in expenses estimated at $28 billion a year in the US alone.

Rather than repeating experiments and incurring further costs, a viable option is to switch from the common practice of completely randomized (CR) experimental designs, which are popular due to historical reasons, to randomized block (RB) designs, which are widely used in other disciplines.

The power and logic of RB designs are less subject to bias, and in some cases, such as within-animal or within-litter experiments, RB is the only suitable design. This webinar discusses the approach and statistical analysis of RB designs and explains why they should be the default design in laboratory animal research.

Presenter

  • Michael FW Festing, Ph.D., D.Sc., Cstat.