The average extracted variances of the 2
latent variables, 0.75 and 0.77, were >0.5, indicating
good convergent validity. The correlation coefficient
between the 2 latent variables was as high as 0.87
(SD=0.02) (Figure 1), indicating that the constructs of
‘‘prevention asthma of an episode’’ and ‘‘treatment of
an asthma exacerbation’’ were significantly correlated.
Figure 1 includes the confirmatory factor analysis
(CFA) estimates for the teachers’ asthma care selfefficacy
scale model.