A New Classification Framework, Theoretical Interpretation and Application of Curvilinear Relationships

Abstract

In many study fields, scholars have extensively explored the linear relationships between variables. But some recent studies have shown that the linear models sometimes fail to reflect true relationships between variables and often misguide practice. Therefore, scholars strongly advocate exploring the curvilinear relationships between variables. Drawing on previous studies and the principles of mathematical statistics, this paper combines the linear relationship and the curvilinear relationship and divides the curvilinear relationship into four categories according to the quadratic coefficient of the quadratic curve and the coefficient of the linear model. Meanwhile, according to their characteristics in the economic and management practice, this study names these four types of curvilinear relationships as Gradually Getting Better effect, Too Much of a Good Thing effect, A Little Bad Thing effect or Catfish effect, and Prolonged Illness Makes the Patient a Good Doctor effect, respectively. Then these four types of curvilinear effects are discussed from four aspects in this paper: data analysis methods, theoretical connotation, application examples and theoretical explanation. Finally, the application of these curvilinear effects in future research is outlooked.

Keywords

Curvilinear effects, Gradually Getting Better effect, Too Much of a Good Thing effect, Catfish effect, Prolonged Illness Makes the Patient a Good Doctor effect

RIGHTS & PERMISSIONS

2022 Higher Education Press