Negative attitude toward HIV was not only a cognitive process that occurred to individual, but also a socio-cultural phenomenon in community setting. That social phenomenon especially found in developing countries in which the familial cultures are still held firmly [52]. Several forming factors of stigma are cultural construction and religious misunderstanding and they are present in every community [53] and [54]. Kujiper [55] emphasized that attitude and behavioral change need a long term process and strong commitment. Religion and culture strongly affect the forming of faith, life values, and individual as well as social behavior. The result of Shaluhiyah's research [56] emphasized that socio-cultural and socio-sexual factors affect teenagers’ sexual behaviors.
Nyblade et al. [57] explained that midwives treating HIV-infected patients also face similar stigma from the society. Even if the midwife experienced exposure and then was contracted with HIV, then she would blame herself and internalized the stigma. Midwives and health workers were afraid to lose social status and considered to be immoral if they were infected by HIV. That fear was reflected in midwife Sm's story on several years worth of her experience.
At that time there was a woman who was almost giving birth and Sm was in charge in administration office. Several moments later, Sm found out that the woman died and was declared as HIV positive. Knowing that, Sm was really afraid that she’d contract the virus – even though she had only done intake interview and checked for blood pressure – and she prayed that she would not contract the virus. Sm did not do VCT since she did not know where to go and whom to talk about it with. Even if she had known where to do VCT, she would not have gone since she was not ready to find out about the result and people would alienate her.
The result above could be explained based on Planned Behavior Theory [27], which is when a person's behavior can be predicted from his/her behavior intention. Meanwhile, behavior intention could be affected by attitude, subjective norm, and behavioral control, in which subjective norm and external behavior control are socio-cultural constructed. Observation result showed that there was no midwife who socialized HIV to patients at HMCP even when negative attitude towards HIV had been decreased. Socialization behavior did not show due to negative socio-cultural construction towards HIV and AIDS.
The result was in accordance to the finding of Adekeye and Adeusi [17] that had been described earlier, in which knowledge and attitude altogether were only 5% affect the reduction of discriminative behavior of health workers towards people living with HIV. In the study, knowledge on HIV of participants in the control group increased significantly but their negative attitude did not decrease significantly. Meanwhile, changes in ECC occurred due to the fact that the four midwives who were observed during ECC had already had quite vast working experiences (more than 15 years experiences) so that the skill to do ECC had been mastered well. It implied to the small improvement of basic listening skill and attentive skill of midwives participants in the experiment group.
On the other hand, quasi experiment with small number and non randomized subject has a few limitations on validity [49]. First is statistical conclusion validity which is threated by extraneous variance in the experimental setting, in example by room's high temperature and inaccurate effect size estimation with an example is dichotomous HIV knowledge's score making effect size become underestimate. Second is internal validity threated by selection used non randomized, history of subject's experience between meeting, and maturation reflected from age range from 22 years old until 54 years old. That differences are might be affected subject's cognition when absorbing and memorizing information. Third is construct validity threated by the weakness of check manipulation because only used qualitative description from two observer. The last, external validity threated by interaction of causal relationship with settings because the possibility of social-cultural norm in rural area is different with urban area.