시뮬레이션을 사용한 구조방정식에서 상호작용효과 검증 방법 비교

초록

This study aims to validate the efficiency of the interaction effect verification method in structural equation modeling by comparing various methods under specific conditions to identify the most efficient one. We generated specific condition data using Monte Carlo simulation and analyzed the interaction effects of the generated data using OPI and LMS, based on the conditions. We generated data under 48 different conditions using simulation and analyzed it using OPI and LMS. As the sample size, the level of interaction effect, and the correlation level between latent variables increased, the power also increased. There was no significant difference in the power of the tests for differences in factor loadings across levels. LMS demonstrated relatively higher power compared to the OPI. Based on these findings, I recommend using LMS when researchers validate effects.

키워드