All I want for Christmas…is a precisely defined research question

[...]for a binary outcome, a risk ratio or risk difference; for a count outcome, a rate ratio; or for time to event outcome, a hazard ratio may be appropriate; for each of the aforementioned outcome types, other summary measure options exist [7, 8–9], and the most appropriate should be chosen based on the trial objective. The average difference (summary measure) in tiredness score (outcome) between eating a mince pie and a bunch of five carrots (treatment conditions) for flying reindeers (population) regardless of whether all the carrots or mince pies are eaten or cheeky reindeers switching allocations or sneaky reindeers eating other food on route or bold reindeers indulging in Santa’s sherry (intercurrent events—treatment policy strategy). Like optimising performance for flying reindeer, in trials of healthcare interventions multiple different questions can be investigated and the answers to these questions may lead to different conclusions on treatment benefit [5]. In such trials, participants may similarly experience different types of intercurrent events which when handled in different ways can result in different impressions of treatment benefit.