Psychosocial Correlates of Dual Methods for Contraception and STI Protection in Urban Adolescents

Table 3

Nominal regression analysis showing association of variables with each method dichotomy (final model).

	Model fit			Estimated Odds Ratios
	𝑋^2	𝑃	None/Dual	BC/Dual	CU/Dual
Age	2.01	0.5694	0.924	1.086	0.853
Male gender	1.38	0.7093	1.131	0.946	1.156
Sex feels good for girls	8.15	0.0429	1.033	1.812**	1.027
No boys I know use condom	14.36	0.0024	1.678**	1.564**	0.952
Feels proud about pregnancy	10.81	0.0128	1.791**	1.221	1.210
Likely to get pregnant in next 6 month	10.43	0.0153	1.683*	1.306	1.474**
Had ≥ 2 sexual partners	8.49	0.0369	0.583*	0.931	0.752*
Had sex at least weekly	11.51	0.0093	0.878	1.633**	1.083
Asked partner about use condom	18.56	0.0003	0.139**	0.279**	0.618
Want kids my age to think I am virgin	9.61	0.0221	1.049	0.824	0.701**

(1) Estimated odds ratios were produced by multinomial logistic regression for the nominal dependent variable “contraceptive method”. Each model used “dual method” as the reference group. (2) The variables included in the final model were identified as significant effects on contraceptive method using Maximum Likelihood Estimation. Age and gender were included in each model as control variables. *𝑃 < 0.05, **𝑃 < 0.01.