19
This study addressed the application of the knowledge gap hypothesis to the health
domain, assessing the knowledge differentials on breast cancer among US women aged
50 and above. A secondary analysis on the 1992 NHIS data was conducted to test the
hypotheses. As expected, education, perceived risk and mammography experience were
all found to contribute significantly to the resulting knowledge gap on breast cancer.
Hypothesized interaction effects between education and the two motivational variables
failed to reach statistical significance.
First, the results clearly supported the classical knowledge gap hypothesis that with the
infusion of media information, there will be a significant knowledge gap between
different SES groups (operationalized as formal education in this study). Older women
with college education obviously knew a lot more about breast cancer than those without.
Moreover, education was significantly correlated with all other independent variables
examined in this study, which further proved its validity as a knowledge gap indicator (p
<. 05).
Results also supported the existence of knowledge gap, albeit small, between groups with
different risk perception. The gap between women with low perceived risk and those with
medium to high perceived risk was comparatively more evident. Two-tailed correlations
showed that perceived risk was positively related with education (p< .05), indicating that
the better-educated women were more aware of their risks on breast cancer. Perceived
risk was also negatively correlated with age, race and marital status (p< .01), all of which
were highly correlated with SES. This again proved the low-SES women to be most
Convention