Model of Latent Profile Factor Analysis for Ordered Categorical Data

In the literature factor analysis is admittedly a well-known and effective multivariate method in the reduction of extensive and broad data, e.g., in the analysis of too many variables. It is also known for the process of unidimensional or multidimensional scale/s construction. Typically, in many studies (especially those pertaining to market research area) a common factor analysis solution is used (based on continuous data). However, there are rarely ever undertaken studies pertaining to latent variable models where other type of data is used based on discrete variables. One of these models might be called Latent Profile Factor Analysis - LPFA. In this article author’s main objective is to propose and discuss its (LPFA) main assumptions. In order to prove the model’s functionality in practice of market research, a brief example of LPFA model for ordered categorical data (based on one-factorial solution) in reference to hedonic consumption data is given at the end of the paper.