APPLICATION OF ADVANCED MODELING IN SOCIAL WORK RESEARCH: A DEMONSTRATION OF MULTILEVEL PATH MODELING ANALYSIS BY MPLUS

Social work research has a crucial role in steering development, implementation and intervention of quality services and programs, in which application of more advanced statistical modeling analyses may redress the inaccuracy of findings and Type I error, that are mostly likely derived from conventional statistical procedures at individual-level analyses, such as ANOVA and OLS Regression. A multilevel path modeling analysis demonstrated by Mplus, an easy-to-use statistical program for advanced modeling, is suggested as a preliminary step to open up the window for other more sophisticated and useful statistical modeling procedures that are imperative for us to comprehend influences of multi-layer social situations on human behaviors and outcomes. Results showed that, in addition to knowledge on effects from a higher level, the multilevel path model increased explanatory power and rectify some augmented path parameters at individual level. Procedures and construction of model command syntaxes for the usage of Mplus are also reported in detail.