Skip to main content

Table 4 Associations between exposure, mediator and outcome variables, Ethiopian Demographic and Health Survey, 2016

From: The mediation effect of contraceptive use and women’s autonomy on the relationship between intimate partner violence and unintended pregnancy in Ethiopia

Models Associations between βa p-value
Sexual IPVb Sexual IPV and contraception use 0.033 0.872
Sexual IPV and women’s autonomy −0.351 0.053
Sexual IPV and unintended pregnancy 0.539 0.004
Emotional IPV Emotional IPV and contraception use −0.404 0.007
Emotional IPV and women’s autonomy −0.380 0.006
Contraception use and unintended pregnancyc −0.269 0.103
Women’s autonomy and unintended pregnancye −0.539 <0.001
Women’s autonomy and contraception used 0.057 0.699
Emotional IPV and unintended pregnancy 0.339 0.037
Effect of emotional IPV on unintended pregnancy when controlling for contraception use and women’s autonomy 0.281 0.094
IPV (physical, sexual, emotional) IPV and contraception use −0.003 0.982
IPV and women’s autonomy −0.307 0.009
Contraception use and unintended pregnancyc −0.290 0.079
Women’s autonomy and unintended pregnancye −0.537 <0.001
Women’s autonomy and contraception used 0.087 0.553
IPV and unintended pregnancy 0.332 0.021
Effect of IPV on unintended pregnancy when controlling for contraception use and women’s autonomy 0.300 0.040
Multiple partner control behaviours Control (multiple) and contraception use −0.391 0.006
Control (multiple) and women’s autonomy −0.661 <0.001
Contraception use and unintended pregnancyc −0.259 0.113
Women’s autonomy and unintended pregnancye −0.519 <0.001
Women’s autonomy and contraception used 0.048 0.749
Control (multiple) and unintended pregnancy 0.453 0.004
Effect of multiple partner controlling acts on unintended pregnancy when adjusting for contraception use and women’s autonomy 0.365 0.022
  1. aThe models were adjusted for age of women, education, rurality (urban/rural), religion, region of residence, number of children ever born, and wealth; IPV, Intimate Partner Violence; b the remaining path coefficients of sexual IPV were not estimated because the first model to estimate path a were insignificant for both mediators; c,d,eThe associations were different in the succeeding models because the models were adjusted for different forms of IPV