Print4.2 Empirical Adequacy and Simplicity
How Much Observation, How Simple, and Explanation?
Conducting and publishing research is a process of interpreting observations and describing the results. Questions about research and hypothesis formation point us in a specific direction and guide the interpretation of results. But how do we determine what we are seeing adequately supports our claims? How many observations do we need to make to assume our interpretation is correct? As well, does our research apparatus adequately support our ability to answer our research question in the detail or resolution necessary? How does an observation count if it does not fit our expected results?
Systems are Complex
Complex phenomena require complex models and descriptions. Not adding enough complexity to a research hypothesis could result in oversimplification of a situation, leaving out crucial thresholds or other limits in the system(s) under consideration. Often, in research, a compromise needs to be made, even for reasons of cost, between adequate observations and extremely comprehensive observations (such as sampling across a large site.) All of these choices can potentially lead to a false confidence in projections of model adequacy, which can result in real-world impacts.
The method of science depends on our attempts to describe the world with simple theories: theories that are complex may become untestable, even if they happen to be true. Science may be described as the art of systematic over-simplification—the art of discerning what we may with advantage omit. – Karl Raimund Popper
Empirical Adequacy and Consistency
Were adequate tests conducted to assure the phenomena observed are consistent, is the study reproducible, or is the instrumentation working within viable parameters and/or limits of observation? As nanotechnology is an emerging field with increasingly finer tolerance, many observations and conceptions of adequacy can change over time.
Simplicity/Scope
What is the scope of the study under consideration? Is the study significantly comprehensive to be relevant to various conditions? Is there detail being lost through the over-simplification of a model or representation?