Qualifying the importance of findings

Abstract Scott Armstrong's article points to problems with the process of scientific publication in marketing, due largely, he believes, to the failure of peer review as a mechanism for identifying studies whose findings are important. This commentary maintains that peer review does, by and large, serve as a valid screen for publishing high-quality research and that his “publish-all-studies” recommendation is not a good alternative. His criteria for “important” findings also warrant qualifications. Replication is not essential for theory testing, but is useful for establishing effect sizes. Validity as traditionally viewed and as adopted by Armstrong is a false pursuit that does research great harm. The true validity issues are to specify sound conceptual definitions of constructs and then achieve correct measurement. Usefulness, his third criterion, can be theoretical and not just practical. While more research focused on developing and testing practical managerial principles is desirable, it is not the only necessary type of research and in any event is not likely to be attempted by any but senior academics.