September 2014 Sang-Min Park & Moamen Gouda Readme file for the Stata do-file "estimations.do" Lines 1-40 are your basic preliminary stuff. Please note that these estimations require a lot of RAM from your machine, more than 4GB is recommended. Even so, estimations tend to tend a long time to run due to the size of the dataset. In line 44, the output table is generated which will be further used in following calculations. In lines 58-85, the mean marginal effects are calculated. Lines 91-101 are for the descriptives. In lines 107-182, the marginal effect of religiosity is handled. To do this, in lines 117-148, the respective marginal effect and standard error is calculated and saved for different percentiles of religiosity. These saved values are then used to graph the marginal effect in lines 155-180, once with one line per percentile, once with a separate graph for each percentile. In lines 181-237, the same is done for the marginal effect of religiosity conditional on values of dummies male, married and unemployed. In lines 240-330, the same is done for the marginal effect of religiosity conditional on values multiple response variables income, age, financial satisfaction and education. In lines 333-388, the same is done for the marginal effect of religiosity conditional on values of religious denomination.