Document Type
Unpublished Work
Publication Date
2021
Subjects
Older people -- Health risk assessment, System analysis, Race, Ethnicity
Abstract
This paper demonstrates the use of reconstructability analysis to examine race/ethnic differences in physical functioning status among older Americans. Reconstructability analysis is an information theory and graph theory method which has been successfully used in multivariate data modeling. The health and retirement study data from 2004 are used for this demonstration. Out of eighteen independent variables included, seven variables appeared in the best models selected. Uncertainty reductions in physical functioning status were substantial across the models but the magnitudes and predictive variables varied slightly by race/ethnicity. Examination of the model probability distributions indicated that composite independent variable states of overlapping variables are related to physical functioning status in a similar fashion across groups with a few differences. Risk increasing and risk reducing states of overlapping predictive variables revealed some differences among race/ethnic groups but these differences were small. However, small differences among groups may have nontrivial differential effect over time.
Keywords: reconstructability analysis; exploratory analysis; physical health status; physical functioning; race/ethnicity; health disparity
Rights
© The Authors
Persistent Identifier
https://archives.pdx.edu/ds/psu/42837
Citation Details
Melekin, Amanuel and Zwick, Martin (2021). “Reconstructability Analysis of the Predictors of Physical Functioning Status in Older Americans by Race and Ethnicity.” Unpublished work.
Included in
Life Sciences Commons, Medicine and Health Sciences Commons, Physical Sciences and Mathematics Commons, Social and Behavioral Sciences Commons