Research design
Studies, instruments and methodology completeness.
Who this is for
Research and M&E officers who own a study.
Creating a study
A study carries its type, paradigm, temporality, unit of analysis and research questions. A panel study that calls itself cross-sectional is rejected as a contradiction, because that mismatch changes what analysis is defensible later.
Attaching data
Attaching a campaign to a study is what makes its responses analysable — it links the responses that carry no study yet, scoped to that campaign and your tenant. A response already belonging to another study is never silently reassigned.
Methodology completeness
A methodology statement reports NOT_PROVIDED for anything the study has not declared, rather than omitting the line. An omitted line reads as “not applicable”; a stated NOT_PROVIDED reads as “nobody recorded this”, which is the truth.
What this does not do
Stated here rather than discovered later
- A study’s declared design constrains what you can claim later. Changing it after analysis has run does not retrospectively make an earlier claim defensible.
- The platform does not choose a design for you and will not recommend one. It refuses combinations that contradict each other, which is a different thing.