SEARCH

SEARCH BY CITATION

This study examined the convergent, discriminant, and predictive validity of several variables proposed to augment the theory of reasoned action (TRA), using both principal components analysis (PCA)/multiple regression and confirmatory factor analysis (CFA)/structural equation modeling (SEM) among a sample of the UK population regarding their intention to have a child. PCA revealed good convergent and discriminant validity for attitude vs. anticipated regret, subjective norm vs. moral norm vs. social relations, but not for intention vs. desire or perceived behavioral control. Multiple regression analyses showed that the additional variables predicted a significant increment in the variance in intention. CFA, however, showed moderate convergent validity and poor discriminant validity and the structural model comprised the 2 predictors from the TRA only.