Limits to statistical method: The mechanics of generating data and undertaking statistical analysis and modeling with that data are relatively straightforward. What is less clear is the process of structuring the scope and content of an empirical study and interpreting the results of data collection, statistical analysis and modeling. The “science” of statistics inevitably needs some of the “art” of interpretation.
The first and most important point to investigate is whether the fundamental question being investigated is properly specified (i.e., can a hypothesis be rejected or is the method relevant to the objective) and whether the scope of the data and methods being used are a reasonable approximation of the problem under investigation.
Depending on whether the study involves using deductive or inductive reasoning to derive a conclusion, the authors may require more or less statistical and modeling rigor.