Table 3. A multi-level logistic regression was performed using age, gender, education, literacy, and income to predict mobile phone ownership.
Estimate | OR | z value | p-value | Std. Error | |
Age | −0.0836 | 0.92 | −15.3420 | 0.0000 | 0.0055 |
Gender | 0.2551 | 1.29 | 9.0000 | 0.0000 | 0.0283 |
Education | −0.5159 | 0.60 | −40.8650 | 0.0000 | 0.0126 |
Literacy | 0.4522 | 1.57 | 21.4810 | 0.0000 | 0.0211 |
Income | −0.0400 | 0.96 | −21.9030 | 0.0000 | 0.0018 |
The coefficient, odds ratio (OR), standard error, z-value, and p-value for each regressor are shown below. Education was the strongest predictor of mobile phone ownership, whereas income and age had little predictive ability.