# Why Analytics Must Ultimately Be Subjective By:: [[Ross Jackson]] 2024-04-23 A great deal of attention is given to the objectivity of analysis. Analytics, as popularly understood, is expected to be objective. This typically means that the analysis is free from bias. However, as a human, evaluative endeavor, it is difficult to imagine how analysis could be objective. In _Philosophy and Real Politics_, Raymond Geuss explained that “there is no obvious single dimension along which we can distinguish the good, the bad, the better, the worse, the best” (p. 39). There is no objective basis for selecting the dimension(s) along which one makes an assessment. Any assessment. All assessments. Determining what to measure is a value judgment. Is efficiency more important than equity? Are environmental impacts more critical than concerns of profitability? The list of competing demands is extensive. There are no objective determinations related to the relative prioritization of these assessment dimensions. The degree to which they matter is contextual and idiosyncratic. One will undoubtedly develop different prioritizations of alternative courses of action depending on which dimensions are selected and how the specified dimensions are weighted. For this reason, analytics must ultimately be subjective. The selection and prioritization of evaluative dimensions is always subjective, and the analytic results will be contingent upon that subjective determination. Analytics is as much figuring out _what_ matters to _whom_ as applying advanced statistical techniques. Within the framework selected, analysts have professional obligations as to how to treat data, conduct analysis, and interpret results. It is within that somewhat limited space only that analytics can be objective. Analytic procedures can be objective. Outside those procedures, analysis is subjective. #### Related Items [[Analytics]] [[Subjective]] [[Objective]] [[Assessments]] [[Value]]