Path Analysis

Path analysis is a statistical procedure for testing the causal relationship between observed variables. In a path analysis model, this cause–effect relationship is not discovered via data analysis but instead is formulated based on the researcher’s knowledge or on previous studies. Path analysis was initially developed by Sewall Wright in 1921 for examining the direct and indirect effects of variables on other variables; a century later, it continues to be a popular statistical procedure. Since the rapid development starting in the 1970s of a more comprehensive family of statistical techniques known as structural equation modeling (SEM), path analysis has been viewed as a special type of SEM in which only observed variables are involved in the analysis.