How Small Sample Size and Replication Can Increase Accuracy in Experiments: Lessons that Marketing May Learn from Agricultural Scientific Method

Abstract

This paper examines the use of small sample sizes and replication in marketing experimentation, including full factorials, fractional factorials, Latin squares and their derivatives such as conjoint analysis. It is well understood within agricultural research that the sample size used within these experiments should be kept to a minimum if maximum reliability is to be achieved. This understanding, which underlies the massive success of agricultural research in the last century, does not appear to have been transferred to marketing. This article explains the logic behind this counterintuitive claim. It then discusses the links between the use of small sample size and replication in experimental research. It concludes that the current very low level of replication in marketing can be related to a very basic mismatch between academic marketing's theoretical expectations of replication outcomes and the degree to which these expectations can be meaningfully achieved by replication within any living environment.