Pooling noisy data sets (Q4356796)

This article proposes and illustrates an econometric method for pooling two noisy data sets. The pooled estimates depend on parameters measuring doubt about the prior, experimental contamination and structural similarity. Implicit or explicit reference must be made to these three concepts regardless of the method that is used to pool data. The explicit mathematical reference to these concepts in this paper is intended to facilitate the discussion about the assumptions that underlie a pooling exercise. The method is illustrated by a study of the determinants of the growth rates of developed and developing countries. These two data subsets yield estimates that are very similar, which makes the results not very sensitive to the form of pooling and which points toward low experimental bias and a high degree of similarity. For the entire collection see [Zbl 0844.00014].