Abstract (Translated):
In the present study, we propose a standardization of indicators based on the Linear Scaling Technique (LST) that is called "interval LST". Traditional methods have certain limitations in terms of robustness, capability to analyze time series, loss of original variability for data and capability to generate ordinal scales for representing levels - low, medium, and high - of the social phenomenon to be studied. This technique manages to overcome these limitations through two stages: 1) establishing intermediate levels of correspondence through an inter-subjective consensus among specialists, and 2) the linear standardization among the reference values. The proposed method represents a double advance. On the one hand, it allows us to build partial indices (one for each dimension of a concept) that meet all the required conditions. On the other hand, the method can generate a comparable ordinal scale between the various components that are available to be added by any other particular method that preserves such a condition. To illustrate these advantages, we present an exercise of modification which is applied to three social well-being indicators using data from Argentina. However, the method can be applied to the standardization of other theoretical concepts.