| tags: [ replication reproducibility ecology evolution generalisation validation ] categories: [reading ]

Nakagawa and Parker 2015

Why do we replicate?

  1. Assess the validity of prior findings
  2. probe the generality of those findings

Levels of replication

  1. Exact (also known as “direct”): highest fidelity to the original work. But in ecology, usually can only be ‘close’ replications.
  2. Partial: there is a spectrum of partial replications, from close to limited. These have slight procedural differences.
  3. Conceptual: uses distinctly different study designs to test the same hypotheses.
  4. Quasi-replication (cross-species or system)

Assessing validity

Probing generality

Conceptual replications: when results concur, we can define generality. If they conflict, we cannot draw robust inferences about why this occurs.

The majority of large-scale meta-analyses

Incentives to encourage replication in ecology and evolution