Resumo

TREVINO CANTU, Jesús Antonio. Standardization alternatives for space-time composite indices. The case of educational backwardness in the states of Mexico from 2000 to 2020. Invest. Geog [online]. 2022, n.109, e60615.  Epub 29-Maio-2023. ISSN 2448-7279.  https://doi.org/10.14350/rig.60615.

Researchers of composite indexes have detected and attempted to eliminate the implicit weighting of data because it leads to biased indexes, with little success so far. This bias is usually attributed to variables having different ranges of values, unequal maximum and minimum values between variables, or the asymmetry of each variable. The procedure for addressing this triple challenge is data standardization. No procedure is currently available that simultaneously solves this triple issue and, with it, implicit weighting. Equal maximum and minimum values between variables solve the unequal-range issue but do not correct the asymmetry of each variable. An equal range between variables does not imply equal maximum and mínimum values between them, nor does it guarantee the symmetry of each variable. Finally, correcting the asymmetry of each variable does not imply equal maximum and minimum values or equal ranges between variables. The objective of this study is to suggest and illustrate a variable standardization procedure that eliminates the implicit weighting of each variable. The standardizations reviewed in this paper to elaborate composite indexes solve only partially the implicit weighting issue. Thus, this paper proposes a standardization procedure that simultaneously solves the triple challenge imposed by unequal ranges, unequal maxima and minima, and asymmetry. The propose new procedure, balanced standardization, solves the issue of implicit weighting in time and space in the cross-sectional and longitudinal aggregation of variables of the different observation units. This work illustrates the proposed standardization procedure with variables related to the educational lag in the states of Mexico. The methodology proposed herein is applicable to any topic requiring spatio-temporal standardization.

Palavras-chave : standardization procedures; implicit weighting; unequal ranges; different minimum and maximum values; asymmetry.