Table 3

Predictors of responses favouring the circumcision of a male child in Botswana

CharacteristicProportion (%) favouring circumcision* among those with characteristicProportion (%) favouring circumcision* among those without characteristicUnivariate analysis p valueMultivariate analysis adjusted odds ratio (95% CI)
*“Favouring circumcision” defined as those responding that they would definitely or probably circumcise a male child, as compared to those who would definitely or probably not circumcise a male child.
†Univariate analysis represents χ2 for differences among all ethnic groups, by the Kruskall-Wallis test. Four individual ethnic groups with p<0.2 were included in the multivariate model as potential confounders.
‡Univariate analysis represents χ2 for differences among all locations, by the Kruskall-Wallis test. Location was entered into the model as a class.
Male sex230/271 (85)178/223 (80)0.14NS
Married111/127 (87)297/367 (81)0.10NS
Has children286/331 (86)122/163 (75)0.0011.8 (1.0 to 3.4)
Secondary school education or higher257/320 (80)151/174 (87)0.07NS
Christian religion269/321 (84)139/173 (80)0.33
Employed310/373 (83)98/121 (81)0.59
Ethnic group†0.28NS
Location (all sites)‡0.00020.03
Age (median)0.002NS
    Favour circumcision: 30 years
    Do not favour circumcision: 26 years