Correspondence Analysis and Logistic Modeling: Complementary Use in the Analysis of Age at Onset of and Causes of Mental Disability in India

Dilip C. Nath, Gauhati University
Prasenjit Sinha, Gauhati University

Logistic or other modeling approaches are often appropriate for studying mental disability data. However, health surveys may be more complex. With numerous variables, there is a need for exploratory analysis. Parsimonious description of the data is also a useful complement to modeling. Correspondence analysis may be useful in such exploratory phases. The complementary use of the two approaches is presented in the context of a disability survey among Indian population, which focused on the relationship between age at onset/causes of mental disability and different socio-demographic factors. The data for this study is from the survey of disabled persons in India conducted nationwide by the National Sample Survey Organisation, India during the year 2002. According to this study of the dataset, age at onset and causes of mental disability in India varies with the factors like geographical location, caste, education, marital status, place of residency (urban and rural).