Same Dimensions, Half the Score: A Composite Index and the Components Published Beside It
DOI:
https://doi.org/10.69019/jelp.v1i2.86Keywords:
composite index, literacy, dimensionality, measurement validity, educational indicatorsAbstract
Jambi and Jawa Timur record almost identical scores on both published dimensions of Indonesia's community literacy development index — 0.5650 against 0.5670 on the first and 0.5920 against 0.5610 on the second — and composite scores of 25.53 and 56.29, differing by a factor of 2.2. Using Central Statistics Agency figures for 2025 across the thirty-three provinces with data, this article examines the relationship between the composite and the two dimensions published alongside it. The two dimensions correlate with each other at 0.9462 on Pearson and 0.8807 on Spearman, so they carry very nearly the same information; their coefficients of variation are 0.3236 and 0.3285. The composite correlates with the first dimension at 0.8568 and the second at 0.8222, but with the simple average of the two at only 0.8506 on Pearson and 0.7574 on Spearman — less than with either dimension taken alone. Provinces move an average of 4.79 rank places of thirty-three between the composite and the average of its dimensions, and Jambi moves twenty-four. The article concludes that the composite is not a function of the components published with it, that the two dimensions are close to a single dimension, and that a user cannot reconstruct the index from the table in which it appears.Downloads
Published
Issue
Section
License
Copyright (c) 2026 Yusrina Nur Dianati, Dewi Susanawati, Dewi Kurnia Sari (Author)

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International Licence (CC BY-SA 4.0).