A. Background & Definitions
This Toolbox item refers to one of the Core Requirements (Core Requirement 9 - "Reported data must disclose all repetitions of a study, an experiment, or a test regardless of the outcome").
There is no formal definition what "repetition" is. One has to acknowledge that there are often methodological differences between two or more consecutive repetitions of a study that may be too significant to be considered as repetitions.
EQIPD offers the following basic guidance on how to define repetitions:
Repetitions are studies, experiments or tests that:
- answer the same research question (even if this involves certain variations in the study conditions, experimental variables and outcomes)
- are run consecutively
- are technically valid (i.e. no documented failure of hardware, software, research tool or reagent)
- failure of positive or negative control may usually not be recognized as a technical failure justifying the exclusion of a repetition from being disclosed
- failure of a positive or negative control may justify declaring the test results not valid, but
- failure of a positive or negative control does not justify failing to disclose the repetition
B. Guidance & Expectations
In all cases, it is up to scientists to decide whether a study (experiment, test) was repeated or not, and, if yes, decide whether reports should disclose data for all repetitions or simply acknowledge the repetitions.
EQIPD expects that a Process Owner conducts regular spot checks on non-public reports and manuscripts prepared for journal submission to make sure that all repetitions are disclosed.
It is up to a Process Owner to decide which method is used to make sure that she/he can identify all repetitions of a study, e.g.:
- use a common 2.1.2 Unique study ID for all repetitions
- add to a repetition study protocol reference(s) to previous studies
- maintain a central (for a research unit) list of studies planned and completed
to be added
back to Toolbox
Next item: 2.4.1 Non-public reporting