Warren C Sanderson
Type: Journal article
Publication date: October 2016
Published in: Journal of Aging and Health Vol. 28, No. 7, pp. 1178-1193
To provide an example of a new methodology for using multiple characteristics in the study of population aging and to assess its usefulness.
Using the English Longitudinal Study of Aging (ELSA), we investigate three characteristics of each person 60 to 85 years old, by level of education, hand-grip strength in 2004 (measured in kilos), chair rise speed in 2004 (measured in rises per minute), and whether the person survived from 2004 to 2012. Because the three characteristics are measured in different units, we convert them into a common metric, called alpha-ages.
We find that the average of the alpha-age differentials in the measures of upper body and lower body strength predicts educational differentials in subsequent survival better than either physical measure alone.
This result demonstrates the benefit of combining characteristics, using alpha-ages to convert incommensurate observations into a common metric.