Theory, Measurement, and Replication in the Social Sciences

A special type of problem is encountered in attempting to associate theoretically defined concepts with operational definitions whenever we deal with properties that have been defined in terms of their causal implications. The problem is illustrated by the concepts "mass," "power," and "discrimination." To measure these variables, we must make a number of theoretical assumptions about the operation of other variables. If replication is possible, these assumptions may be simple and plausible; where replication is difficult, problems of measurement become so entangled in theory that it is difficult to separate the two.