Replication and Reproducibility
Replication and Reproducibility
- Stefan Schmidt
Abstract
This chapter focuses on replication and reproducibility. A single observation cannot be trusted. Similarly, findings from a single experimental investigation may reflect some regularity but they may also be due to chance, artefacts, or misinterpretations. Therefore, it is necessary to repeat the respective research procedure in order to validate the observations from the first study. Such a repetition is called replication. It is a very basic methodological tool that serves to transform an observation into a piece of validated knowledge. An observation or relationship that is found repeatedly and is also found under various scope conditions fulfils the important scientific criteria of reproducibility. There are different types of replications. The most basic distinction is between a narrow-bounded notion of replication termed direct replication and a wider notion of replication termed conceptual replication.