A structural
equation model was a valid measure for investigating the causal
relationships among the variables, because: (1) several variables
were correlated; (2) physical health was measured by the
combination of three items: the number of diagnosed chronic
diseases, number of illness symptoms, and SPH; and (3) when the
correlated measurement errors were taken into account, the model
was able to specify the relationships among the observed variables
and health status (Lin and Wan, 2001).