Published online by Cambridge University Press: 31 July 2008
An attempt has been made to determine which of several social, demographic and economic variables best account for the variance in contraceptive use in a low-income, urban Colombian population. The results indicated that of the independent variables included, occurrence or non-occurrence of intercourse in the past month, education, employment status, marital status and age were the best predictors, in that order. Altogether these factors accounted for 17% of the variance in contraceptive use and 87% of the variance within the model. The remaining variables had relatively little predictive utility; and this was particularly true for interspousal communication, race and religious commitment.