ABSTRACT
The literature generally talks of two overarching types of replication: internal and external. The former would see action upon the original set of data presented by the author and is most often carried out by that author, prior to the publication of results and therefore already incorporated into the original study. Unlike external replication, this process will, therefore, not require the study to be repeated in any way. It provides a very useful revisit to the study and cross-validation, however, and potentially offers an additional element of reliability in the sense that we see the author providing further analysis and scrutiny of the data – either using the same analysis procedures or others (see below). Such an approach can help us see both whether any errors present at that analysis stage or the outcomes were themselves affected by the nature of the initial analysis undertaken.
