Table 5. Bivariate analysis of predictors of client and staff perceived quality care with marginal effects.
| Model 1 | Dependent variable: Overall client-perceived quality | |||
| Independent variables | Coef. | Std. Err | Marginal Effect + | (95%Conf. Int.) |
| Technical quality | -0.018* | 0.005 | 0.0002 | -0.028 -0.007 |
| Age (mean = 45 years) | -0.006* | 0.003 | 0.0000 | -0.011 -0.000 |
| Females | 1.0 | 1.0 | 1.0 | 1.0 |
| Males | 0.014 | 0.097 | -0.0002 | -0.176 0.204 |
| Not married | 1.0 | 1.0 | 1.0 | 1.0 |
| Married | -0.138 | 0.092 | -0.0019 | -0.317 0.042 |
| Other religions | 1.0 | 1.0 | 1.0 | 1.0 |
| Christian religion | -0.141 | 0.127 | 0.0020 | -0.391 0.109 |
| Educated | 1.0 | 1.0 | 1.0 | 1.0 |
| No formal education | -0.056 | 0.125 | 0.0008 | -0.301 0.189 |
| Public facility | 1.0 | 1.0 | 1.0 | 1.0 |
| Private facility | -0.353** | 0.100 | 0.0049 | -0.549 -0.158 |
| Urban location | 1.0 | 1.0 | 1.0 | 1.0 |
| Rural location | 0.684** | 0.980 | -0.0095 | 0.492 0.876 |
| Wealth quintile 1 | 1.0 | 1.0 | 1.0 | 1.0 |
| Wealth quintile 2 | 0.420** | 0.118 | -0.0078 | 0.188 0.651 |
| Wealth quintile 3 | 0.496** | 0.127 | -0.0089 | 0.247 0.746 |
| Wealth quintile 4 | 0.980** | 0.130 | -0.0143 | 0.724 1.236 |
| Wealth quintile 5 | 0.915** | 0.134 | -0.0137 | 0.652 1.177 |
| Obs. | 1,903 | |||
| Pseudo R2 | 0.0105 | |||
| Log Likelihood | -7496.86 | |||
| Prob > chi2 | 0.0000 | |||
| Model 2 | Dependent variable: Staff-perceived quality | |||
| Independent variables | Coef. | Std. Err | Marginal Effect + | (95%Conf. Int.) |
| Technical quality | 0.11** | 0.013 | -0.0002 | 0.085 0.137 |
| Age (mean = 39 years) | -0.00 | 0.008 | -0.0000 | -0.016 0.015 |
| Females | 1.0 | 1.0 | 1.0 | 1.0 |
| Males | 0.19 | 0.222 | -0.0005 | -0.248 0.624 |
| Not married | 1.0 | 1.0 | 1.0 | 1.0 |
| Married health staff | -0.29 | 0.224 | -0.3706 | -0.725 0.151 |
| Other religions | 1.0 | 1.0 | 1.0 | 1.0 |
| Christian religion | -0.30 | 0.499 | -0.0023 | -1.278 0.677 |
| Other qualifications | 1.0 | 1.0 | 1.0 | 1.0 |
| Tertiary education | 0.34 | 0.209 | -0.0002 | -0.071 0.748 |
| Clinical staff | 1.0 | 1.0 | 1.0 | 1.0 |
| Non-clinical staff | -0.22 | 0.285 | 0.0006 | -0.776 0.340 |
| Urban location | 1.0 | 1.0 | 1.0 | 1.0 |
| Rural location | -0.97** | 0.225 | 0.0004 | -1.412 -0.529 |
| Public health facility | 1.0 | 1.0 | 1.0 | 1.0 |
| Private health facility | 0.57* | 0.226 | -0.0025 | 0.123 1.010 |
| Obs. | 324 | |||
| Pseudo R2 | 0.0276 | |||
| Log Likelihood | -938.32 | |||
| Prob > chi2 | 0.0000 | |||
Source: WOTRO-COHEiSION Project Household and Health Facility Surveys (March, 2012)
*p<0.05
**p<0.0001
+Conditional marginal effects (Model CE: OIM): Marginal effects represent the in probability when the respective predictor/independent variables increase by one unit (i.e. 0 to 1 for binary variables and instantaneous change for continuous variables)