The frequency of listening to the radio did not significantly affect the likelihood that single women would support wife-beating.
The exogenous variables used as exclusion restrictions in the bivariate probit model are several proxies of exposure to mass media collected from the GDHS: frequency of watching television, frequency of listening to radio and frequency of reading newspapers or magazines.
Education, place of residence, wealth index, ethnic affiliation, religious affiliation, women's autonomy in household decision-making, and frequency of listening to the radio were significantly related to acceptance of wife-beating over the years.
However, when the data are not separated by ever-married and never-married status, women's autonomy, frequency of listening to the radio, and marital status did not significantly affect the likelihood that a woman would support wife-beating.
Among traditional mass media, only the frequency of listening to radio was associated with knowledge about TB transmission.
Among the traditional mass media, the frequency of listening to radio was associated with correct knowledge about tuberculosis transmission.
Frequency of reading news paper/magazines and watching television were not significant; frequency of listening to radio (aOR 1.08, 95% CIs 1.04, 1.13) was associated with correct knowledge.
People belonging to the poorest wealth quintiles were more likely to have TB while the frequency of listening to radio was associated with knowledge about TB transmission.
A media exposure index was derived using a Principal Components (factor) Analysis (PCA) ) [ 33] and was based on responses to questions asked on the frequency of watching television, the frequency of listening to radio, and the frequency of reading newspapers.
Correct knowledge about TB transmission was associated with education and income, but among mass media, only the frequency of listening to radio was associated with correct knowledge which was similar to the results reported from Bangladesh [ 24].
