Nonlinear analyses in sales research: theoretical bases and analytical considerations for polynomial models

Abstract

Despite theoretical bases that may support nonlinear relationships, there is a relative paucity of studies testing these effects in organisational and behavioural research. Theories hypothesising linear associations among variables in sales research found to have low predictive validity are prime candidates for re-conceptualisation with theories predicting nonlinear relationships. In the sales domain, while nonlinear-oriented theories have been advanced and many nonlinear associations have been explored, myriad additional variables may possess theoretically-based nonlinear relationships with key sales-related outcomes. Further, nonlinear integration can be useful in improving linear analyses involving moderation. The purpose of this article is to expound upon and promote the utilisation of theoretically-driven nonlinear analyses in sales research. In this pursuit, the article delineates the rationale for nonlinear examinations and provides an extensive review of nonlinear research conducted in the extant sales literature. Based on this review, theoretical bases used to support nonlinear hypotheses are advanced and types of polynomial tests are identified. Additionally, guidelines in conducting and interpreting main effect and moderated polynomial analyses are provided along with future research opportunities for nonlinear analyses in sales research.