Abstract
Over the past decades, there has been a wide range of empirical research in the e-learning literature. The use of multivariate statistical tools has been a staple of the research stream throughout the decade. Path analysis modeling is part of four related multivariate statistical models, including regression, path analysis, confirmatory factor analysis, and structural equation models. This chapter focuses on path analysis modeling for beginners using LISREL 8.70. Several topics covered in this chapter include foundational concepts, assumptions, and steps of path analysis modeling. The major steps in path analysis modeling explained in this chapter consist of specification, identification, estimation, testing, and modification of models.
TopMain Focus
The use of multivariate statistical techniques has been a staple of the e-learning empirical research stream throughout the decade. This may reflect the fact that the field has drawn in some researchers who have been trained in such analytical techniques which are common in many business disciplines, and those scholars simply brought these techniques with them to design and analyze studies of online learning. Moreover, there is a growing number of studies that have used highly sophisticated statistical techniques such as structural equation models and hierarchal models in recent years (Arbaugh, Hwang, & Pollack, 2010).
This chapter focuses on covariance based path analysis modeling using LISREL 8.70. Structural equation modeling (SEM) is “a comprehensive statistical approach to testing hypotheses about relations among observed and latent variables.”(Hoyle, 1995) SEM methodology is used to test four types of theoretical models: regression, path, confirmatory factor, and structural equation models. LISREL is capable of modeling all four models. All four models can be tested by following the five steps: specification, identification, parameter estimation, testing, and modification. To complement this chapter of path modeling, several other chapters are concerned with path modeling applications, an introduction to SEM using PLS graph, and SEM applications. The remainder of this chapter is organized by the following several sections.