Confronting Data Analysis with Constructivist Philosophy
This paper develops some ideas from the confrontation of data analysis with constructivist philosophy. This epistemology considers reality only dependent of its observers. Objective reality can never be observed. Perceptions are not considered as representations of objective reality, but as a means of the self-organization of humans. In data analysis, this leads to thoughts about the impact of the gathering of data to the reality, the necessity of subjective decisions and their frank discussion, the nature of statistical predictions, and the role of probability models (frequentist and epistemic). An example from market segmentation is discussed.
KeywordsProbability Model Objective Reality Market Segmentation Data Analyst Subjective Decision
Unable to display preview. Download preview PDF.