Evaluation of the Psychometric Properties of Neuropsychological Tools for India's Ageing Population: Insights from the Longitudinal Ageing Study in India (Lasi)

Sumit Kumar Das, All India Institute of Medical Sciences

Population aging, driven by demographic transitions, has become a significant global concern for both the present and future decades. In addition to other challenges, a comprehensive understanding of elderly mental health, facilitated by accurate assessment tools, is essential for evaluating treatment burdens and developing effective policies. Using the first round of nationally representative data from the Longitudinal Ageing Study in India (LASI), this study evaluates the psychometric properties of the neuropsychological tools consisting of 40 questions, utilising a complete sample of 33,668 individuals. This study employed unidimensional graded response model and differential item function (DIF) techniques within the item response theory framework for psychometric evaluation of the scale. About 59% of respondents were from rural areas, with males making up slightly over half the sample, and around 30% identifying Hindi as their primary language. The discrimination parameter ranges between 0.22 to 7.73. While the difficulty parameter ranges between -0.975 to 2.87. Out of 40 items, DIF was detected in 32 items primarily related to place of residence. Acknowledging that some items are expected to have some degree of DIF due to different subgroups, the result suggests the need for refinement of the tool to minimize unintended bias before future use.

Keywords: Population Ageing, Qualitative data/methods/approaches, Older Adults and Intergenerational Relations

See extended abstract.