8.5 Overall Influencer Effects
The analysis thus far has considered the effects of influencers on each behavior. Here we examine the overall effect of each influencer on the consistency of self-reported responses with biomedical recommendations. We do this by fitting the same model described above but omitting the categorical variable for each type of behavior, and hence removing the interactions between behavior and influencer as well.
The result shows that the ASHA has the strongest positive effect of all the influencers considered with a roughly six fold increase in odds of following biomedical advice relative to being influence by no one, followed by the government doctor, then the ANM, and then the private clinics. The only influencer with an overall negative effect was the RMP. Higher parity and older ages are progressively less likely to comply with biomedical advice than the youngest ages and lowest parity. The odds ratios for education consistently increase which is noteworthy even if none of the confidence intervals exclude 1.0.
Figure 8.8: Plot of odds ratios showing overall effect of each influencer on biomedically recommended behaviors.
From the above Figure we can calculate the overall influence of each influencer on the odds that women are adopting the biomedically-consistent practice. If we rank them by magnitude we have the following order:
| Influencer | expEst | Rank |
|---|---|---|
| Iasha | 6.03 | 1 |
| Igovdoc | 4.84 | 2 |
| Ianm | 4.81 | 3 |
| Iprivclinic | 2.60 | 4 |
| IOTHER | 2.27 | 5 |
| Ifam | 2.09 | 6 |
| Imedia | 2.05 | 7 |
| Ifriendrelnei | 1.30 | 8 |
| Idai | 1.13 | 9 |
| Irmp | 0.49 | 10 |
All of the positive effects are associated with the medical system or the media but the ASHA impact is the largest in magnitude.