본문 바로가기
  • Home

Moderated Regression Analysis in Korean Tourism and Hospitality Research: Is 3-step Hierarchical Procedure Necessary?

Kim, Jin-hoo 1

1세종대학교

Excellent Accredited

ABSTRACT

Moderation analysis is conducted frequently because it helps to better understand human behavior and social phenomena. Korean tourism and hospitality research investigating moderation effects often uses a three-step hierarchical variable entry procedure, which incorporates an independent variable first, moderator variable next, and the interaction term last in a regression model. Reviewing the articles published in the Journal of Tourism Sciences for 2016-2018, this study examined how often this procedure was used, what problems it had, and how it should be changed to test moderation properly. Results include: (1) the three-step hierarchical procedure was widely used; (2) the procedure is based on the premise that the independent variable and the moderator variable must have a significant effect on the dependent variable for a significant moderating effect to exist; and (3) the three-step procedure has no theoretical basis and increases the probability of missing moderation when in fact it exists. Based on the results, this study urges researchers to discard the idea that the three-step procedure is necessary in testing moderation. Using the procedure itself may not be wrong, but whether the independent or moderator variable has a significant effect on the dependent variables in the first- and the second-step models should not affect determining moderation effects between them.

Citation status

* References for papers published after 2023 are currently being built.