Industrial Purchasing Behavior: Some Final Comments

We thank Professor Levitt for the kind tone of his response. Moreover, he and Professor Vatier have generously provided further insight into his study b the use of regression analysis and its Bayesian interpretation. Nonetheless, we believe that some comments on his reply are in order. We believe that this exchange has been productive in defining more precisely the conclusions and limitations of a pioneering marketing study. It is our contention, however, that our original assertions have been substantiated and that evidence for the source effect is, at best, very weak.