Improved Structural Equation Models Using Factor Analysis

Abstract

We develop an agricultural adaptive structural equation model (SEM) that incorporates a large number of factors. These factors simultaneously account for food production while uncompromising food quality and safety. Using the principal component analysis (PCA), we obtain provisional factors, which we rotate using factor analysis, thus leading to reduced number of variables. To decide on the form of the covariance structure in the estimation of the parameters of the regression model, we conduct analysis of covariance. The generated principal components are incorporated into the SEMs where testing of different inter-associations among latent variables (LV) is conducted. For simplicity of the model, we utilise J reskog linear structural equation (LSE) system throughout the investigation process. Using a comprehensive real-life example, we illustrate the concepts and effects of the outcomes. The results show that factors such as energy, transport, labour and fertilizer make a positive contribution in the increase of the quantity and quality food. In addition, we demonstrate how to determine the key factors that influence food production where some factors are not directly measured.