The null hypothesis: closing the gap between good intentions and good studies

Some papers appearing in the two leading clinical journals of our specialty have fundamental defects in design, statistical analysis, and reporting. Actual publications are used to illustrate what investigators should and should not do, from initial planning through to their finished manuscript. A more detailed discussion of the most common statistical methods used in reproductive research is given as an online supplement. Some of these deficiencies illustrate the importance of registry of all study parameters for randomized clinical trials, a process covered in greater detail in an accompanying review and required for those manuscripts to be considered. Authors and reviewers need to be more aware of these requirements and common errors and pitfalls so that only quality manuscripts are published.