Spatial Patterns of Heaping in Age Data among Literates, Illiterates, and Numeracy-Literacy Correlates: A Cross-Sectional Analysis of Census 2011, of India.

Abstract

Background: Accurate information on age is an essential prerequisite for demographic and epidemiological studies. This study analyzed the age data quality among the literate and illiterate (combined, rural, urban) population and examined the association between data quality and literacy.

Material and method: Secondary data on age statistics and literacy were obtained from census 2011. We measured age data quality for literates and illiterates (combined, rural, urban) by transforming Whipple's index known as ABCC, abbreviated based on surnames of the developers A'Hearn, Baten, and Crayen (2009). Correlation tests between literacy and ABCC were performed. RStudio (Version 1.3.1073) was used.

Result: Computed ABCC indices in majority states (union territories) for literates (data quality rough) were higher than illiterates (data quality very rough). Urban data among literates and rural data among illiterates were comparatively superior. Correlation between ABCC and literacy rates for (i) literate combined (R = 0.84, P = 3.5e^-10), (ii) literate rural (R = 0.8, P = 1.1e^-08), (iii) literate urban (R = 0.8, P = 1e^-08), (iv) illiterate combined (R = 0.54, P = 9e^-04), (v) illiterate rural (R = 0.48, P = 0.0034), and (vi) illiterate urban (R = 0.73, P = 6.4e^-07) was significant. Age data quality for both literates and illiterates was poor. There was heaping at terminal digits "0" and "5" even among literates, which contradicts the theoretical expectation of quality data among literates.

Conclusion: Correlations between data quality and literacy were significant, with comparatively lower magnitude among illiterates, which indicates the role of literacy in yielding quality data. Awareness, training, ADHAAR-based enumeration, and digitization may be suggested for better age data.