Informations and abstract
Keywords: quantitative social research, regression function, regression model, prediction, association, causal effects
In this article it is argued that many quantitative sociologists routinely misunderstand and, therefore, misuse regression analysis. To illustrate the main sources of such a bad habit and contribute to its amendment, in the first place the main features and applications of regression are briefly spelled out. Then, a real-world example is used to show how the very essence of regression is typically ignored or poorly understood in the everyday practice of quantitative social research. Finally, five general rules are suggested for making proper use of regression and exploiting its full potential.