Choosing a Sampling Design

In practice

The question asks whether a repositioning schedule reduces pressure injuries across a hospital, so the claim has to generalise beyond one ward. Convenience sampling on one unit would not support it. A random sample is drawn from a list of every eligible inpatient, which fixes the chance of selection and lets the result travel back to that population. Age is known to matter and the oldest patients are few, so the sample is stratified by age band to keep them represented. Assignment is a separate decision: within each ward, participants are randomised, and wards are used as blocks so that ward differences fall evenly across both arms rather than loading one. Sample size follows from the smallest reduction that would change practice, not from how many patients were available in the study window. Finally the exclusions are written down — patients with more than one qualifying condition, in this case — because that sentence is what a future reader needs to judge whether the result applies to their patients.