Reliability criteria bolster product classification decisions: A reply to Jones

We empirically test the robustness of the Easingwood approach to classify early product life cycle forms for infrequently purchased major products. Our results indicate that the key classification parameters could be unstable to variations in the sample size used for estimation, thereby producing more than one classification for several products. We demonstrate that the problem of multiple classifications can be effectively addressed by using the more rigorous criteria of joint confidence regions (as opposed to point estimates) of classification parameters. The benefit of such rigorous classification is that it increases researcher confidence in Easingwood's classification system.