As shown in the above results, the statistically signif-icant differences in health behaviors found accord-ing to age, education, diagnosis, smoking status, and experience of receiving patient education are all factors that should be controlled for in order to examine effects of cardiac knowledge and cardiac self-efficacy on health behaviors. This was done using hierarchical multiple regression (Table 6).The overall model significantly explained 48% of variance in health behaviors (F=14.52, p