Elusive or Illusive Truth? The Role of Sensors and Sampling Strategies in DataCollection and its Implication for Scientific Knowledge and the Pursuit of Truth
A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Environmental Science, Policy and Geography College of Arts and Sciences University of South Florida St. Petersburg
The work presented here represents a practical investigation into the role of truth in the development of scientific knowledge. Truth is commonly understood in the natural sciences to exist as a single, objective reality, independent of human perception. The role of science, then, is to elucidate that truth through observations and experiments in the endeavor to uncover scientific knowledge. However, one's access to scientific truth is only possible through the lens of the "perceptual apparatus"- the combination of available data, observational tools, theory, prior experience, and cultural bias which influence the interpretation of any scientific result. The perceptual apparatus, however, inherently constrains one's interpretation of the truth to a single perspective, leaving open the possibility of other perspectives which potentially hold different interpretations of the same truth. The realization of this would suggest that there lies some value in understanding what interpretations are held by an alternate perspective. The point of the study presented is to explore, on a practical level, how perspective and interpretation shift under different perceptual apparatuses and what consequences that might hold for scientific knowledge developed within the field of environmental science. Chapter II examines how the concept of the paradigm applies to truth and the development of scientific knowledge, in particular examining how shifts in paradigms influence a scientific community's interpretation of truth. The chapter Vll utilizes four case studies of proposed paradigm shifts in the field of ecology to look at how the scientific understanding of four ecological phenomena has changed over time. Chapter Ill utilizes an artificial dataset to examine how the spatial and statistical interpretation of a hypothetical phenomenon differed between nine different resolutions of "sensor'' used to observe the phenomenon. The artificial dataset had an advantage over real observational data in that the "truth" of the phenomenon was known and could be used as a benchmark against which other interpretations were evaluated. Chapter IV utilizes a real-world seagrass distribution and abundance dataset to examine how the interpretation of seagrass distribution was influenced by sampling strategy. The study utilized two different sampling methods (i.e. spatial arrangement of sampling points) and seven different levels of sampling effort (i.e. number of sample points) to predict the distribution of seagrass across a shallow shoal and examined differences in those predictions. Together, the three components of this thesis reveal two important realizations: that a connection exists between information, uncertainty, truth, and knowledge that must be considered when evaluating a scientific endeavor and; two, that by accepting one interpretation to be more truthful than another, the perception of truth, and the knowledge derived from it, becomes not a deterministic concept, as philosophically proposed, but falls along a continuum of truthfulness. In terms of practical implications, this study reveals that science should be cautious when assuming that a particular method, test, or objective function deemed acceptable by the current paradigm is sufficient for justifying truth and the knowledge derived from it. Such caution is especially prudent when the derived knowledge is used to drive a practical decision. In such cases , consideration should be given to evaluating other perspectives derived from other theories, methods, and standards, and deciding whether or not those other perspectives offer valuable insight into the described truth.