An indispensable question to be addressed in any paper, once the authors have specified the sample period, is: what assurance can readers have that the sample period is reasonably stable itself and in relation to the periods in close proximity.
This should be the case so that any test is not distorted by factors specific to, or even close to, the period selected but not germane to the analysis.
The research findings are inevitably the product of the research design, and a key element in the design is the selection of a sample period that is tree from external bias and other confounding factors.
Why did the collection of research data begin in year x and end in year y, as opposed to some other sample period Sometimes the starting or ending point may appear apparent to a reader, as the case in event studies.
But even here the period selected must be carefully chosen so that the event of interest is not affected by extraneous events.
A recent editorial in Psychological Science spells out the requirements that must be acknowledged by the author of a paper submitted for publication and which must include the stopping rule” use for data collection (Rich 2013) We interpret this to mean both the starting and stopping rule in order to justify both the time period selected to start data collection as well as to end data collection.