On: An erroneous paper on religion and generosity is retracted

[im3w1l]

What likely happens is that since the country variable is basically random, you will get a model where the country has negligible effect on the prediction.