Traditional estimation requires data, but without searching them this is impossible. Other theoretical via, in Bayesian statistics, is to give an "a priori" subjective idea of the statistical population. This is not traditional estimation but substitute data for an idea of them in the statistical analysis.