多變項統計分析

臺北科技大學 115-1 學期課程，課號 366679。

課程大綱

This course offers an in-depth introduction to Partial Least Squares Structural Equation Modeling (PLS-SEM), a powerful and flexible technique for analyzing complex relationships among observed and latent variables in applied research. The course is grounded in the use of the SEMinR package in R, providing students with a hands-on, code-based learning experience that emphasizes practical application over mathematical derivation. Throughout the semester, students will develop a strong conceptual and analytical foundation in PLS-SEM, beginning with the distinction between reflective and formative measurement models, and advancing through the evaluation of structural models and the interpretation of model outputs. Particular emphasis will be placed on the following topics: model specification, measurement validation, structural path estimation, bootstrapping for significance testing, predictive relevance assessment, and advanced modeling techniques such as mediation, moderation, and higher-order constructs.

課程進度

Week 1: Introduction to PLS-SEM and Review Hayes's Conditional Process Analysis Week 2: Getting Started with R, RStudio, and SEMinR Week 3: Building Basic Path Models in SEMinR Week 4: Reflective Measurement Models: Specification Week 5: Reflective Measurement Models: Assessment Week 6: Formative Measurement Models: Specification Week 7: Formative Measurement Models: Assessment Week 8: Self-Study (No Physical Class) Week 9: Midterm Week 10: Structural Model Evaluation & Bootstrapping and Significance Testing Week 11: Predictive Power and PLSpredict Week 12: Mediation Analysis Week 13: Moderation Analysis Week 14: Moderated Mediation (Conditional Process Modeling) Week 15: Higher-Order Constructs Week 16: Multi-Group Analysis and Measurement Invariance

評量方式

Attendance 20% Mid-term presentation 40% Final exam 40%

教材

【遵守智慧財產權觀念，請使用正版教科書，不得使用非法影印教科書】 使用外文原文書：是 Course materials will be prepared by the instructor. Optional: Hair, J. F., Jr., Hult, G. T. M., Ringle, C. M., Sarstedt, M., Danks, N. P., & Ray, S. (2021). Partial least squares structural equation modeling (PLS-SEM) using R: A workbook. Springer. https://doi.org/10.1007/978-3-030-80519-7