ABSTRACT
The preceding quotation from Einstein’s reply to Robert Thornton, a young philosopher of science who began teaching physics at the university level in 1944, encapsulates succinctly the importance of examining the methodology, history, and philosophical foundations of different scientific fields to avoid missing the forest for the trees. The field of interest in the discussion that follows is modern econometrics, whose roots can be traced back to the early 20th century. The problem of induction, in the sense of justifying an inference from particular instances to realizations yet to be observed, has been bedeviling the philosophy of science since Hume’s discourse on the problem. Modern statistical inference, as a form of induction, is based on data that exhibit inherent chance regularity patterns. Model-based statistical induction differs from other forms of induction, such as induction by enumeration, in three crucial respects.
