ABSTRACT
The Statistical Analysis Plan (SAP) is probably the most important document for a collaborating biostatistician to work on, regardless of whether it is for a grant application or a protocol for a project leading to a publication. It is a structured proposal that outlines statistical tests and procedures that would be applied in a project. It often includes the hypothesis to be tested, the statistical test to be conducted, details on how the variables are measured and coded (i.e. the exposure, outcome and confounding variables), the statistical software as well as the level of significance at which the p-values will be evaluated against. Information is also provided here on how any complex data analysis problems are handled (e.g. missing data imputation techniques and adjustment for multiple comparisons).
