Do people behave in experiments as in the field?—evidence from donations