Year: 1967
Type: article
Abstract: Computer applications of research findings usually assume either a metric or interval level of measurement.This assumption, not always consciously made, 'follows necessarily as a consequence of the mathematical manipulations which are performed.lAccording to the classical theory of measurement as advanced by Campbell , metric measurement is determined by a process of physical addition.This addition is direct in the case of fundamental measurement and indirect for derived measurement.Physical addition is, in turn, dependent upon the possibility of performing a join, the physical equivalent of addition.Thus, a length, ac, if it can be produced by joining lengths ab and bc, may be thought of as equal to their summation.Similarily, if two masses, x and y, balance a mass z, then their arithmetic sum may be considered equivalent to•z.(The x, y join in this case is completed by placing the two masses, x and y, on the same pan.)The possibilities of performing joins in social science measurement appear limited.Indeed, when the idea of a join is most strictly interpreted, there is apparently only one obviously join-based quantity unique to the social sciences--money.If unable to perform joins to produce quantities, our measurements could be limited, as has been suggested by Coleman,2 to the counting of such objects as people or social relationships, or we could accept the manipulative limits of comparative ordinal measurement.
Source: Social Thought and Research
Author: David Willer
Institution: University of Kansas
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: bronze
OpenAlex ID: W2079074185