On the Use of Item Response Models in the SEM Perspective Anna Simonetto 1 and Maurizio Carpita 1 Abstract For the analysis of complex models for latent constructs measured with several items, the Structural Equation Models (SEM) are being widely disseminated. In this study, our intent is to show how to include in a SEM framework an Item Response Model (IRM), in order to preserve the important characteristics that distinguish this type of approach, such as the possibility of calculate the measurement scales and the reduction of the complexity of the model. We compare these results with a standard SEM. 1 Categorical Data Analysis and Measurement In literature, many models have been proposed for the analysis of latent constructs measured by several categorical indicators. Based on their specific needs, scientists focused on the properties of the constructed measures or on existing relationships between the different latent factors. Obviously, according to the different approaches and different theoretical contexts from which several authors have drawn inspiration, the models differ in assumptions and notation. Focusing on the two well-known models Structural Equation Model (SEM) and Item Response Model (IRM), we wanted to check the possibility to intersect the two environments theorists to the peculiarities of each. Our approach derives from an idea of Gibbons et al. [1] and it is based on the identification of two main approaches to analyzing multivariate latent aspects, taking into account the categorical nature of the observed variables [2]: the Underlying Variable Approach (UVA) and the Item Response Theory (IRT). The UVA assumes that the observed categorical outcomes are incomplete observations of unobserved continuous variables: underlying each of the categorically observed variable there is a continuous variable which is actually measuring the 1 Department of Quantitative Methods, University of Brescia, {simonett,carpita}@eco.unibs.it.