Abstract

PIP: A multivariate analysis and logistic regression of data from the 1983 Bangladesh contraceptive Prevalence Survey was conducted to determine factors associated with contraceptive knowledge and family planning services. The analysis was limited to currently married women. When the researcher controlled for other independent variables, the factors which had a significant positive relationship with knowledge of modern contraceptives were as expected: education (unstandardized coefficient = 1.39), urban residence (1.15), current use (1.14), employment (0.45), and land ownership (0.23) (p .001 for all variables). Unexpectedly, women visited by family planning workers knew fewer modern contraceptive methods than those not visited by them (-0.16; p .001). Muslim women knew more modern contraceptive methods than did non-Muslim women (0.15; p .05). Standarized coefficients indicated that a change in educational status would effect the greatest positive change in number of methods known by source (0.27) followed by urban residence (0.2). Religion (0.02) would least effect change. The computed probability of a women being visited by a family planning worker was close to the actual percentage that were visited (about 30%). The logistic regression showed that age, land tenure, education, religion, employment status, and number of deceased children did not significantly affect the probability of being visited by a family planning worker. The significant variables affecting probability of being visited by a family planning worker were urban residence (coefficient = 0.53), number of methods spontaneously mentioned or knowledge of greater number of methods with source (0.52 and 0.2, respectively), current contraceptive use (-0.21), and family size (-.005). these findings suggest that education and urban-rural residence variables require more research than other variables which have been accorded much research (e.g., employment status, religion, land tenure).