CAJ | 학술논문

在PLS路径建模中，“二级潜变量”没有对应显变量而不能建模的情况。“二级潜变量”通常对应多个“一级潜变量”，采用Patch方法可算出这些“一级潜变量”的第一主成分得分，并且把这些主成分得分作为二级潜变量的显变量，这样便满足了建模的条件。本文基于R软件，以员工满意度为实例展示Patch方法的应用，并对Patch方法作进一步研究。
재PLS로경건모중，“이급잠변량”몰유대응현변량이불능건모적정황。“이급잠변량”통상대응다개“일급잠변량”，채용Patch방법가산출저사“일급잠변량”적제일주성분득분，병차파저사주성분득분작위이급잠변량적현변량，저양편만족료건모적조건。본문기우R연건，이원공만의도위실례전시Patch방법적응용，병대Patch방법작진일보연구。
Models with latent variables of second order, the way they are customarily represented (with no manifest variables), are not allowed in the PLS-PM framework:an LV with no indicators has no place in PLS-PM. The process of the patch approach is a little bit tangled. In the first step we compute latent variable scores of the 1st order constructs with PCA. In the second step, we can run a PLS-PM analysis using the PCA scores of 1st order constructs as indicators of the second order constructs. Through the example of employee satisfaction, the thesis demonstrates the application of Patch approach with the statistical software R.