# Faktorska analiza

Srša, Sanja (2015) Faktorska analiza. Diploma thesis, Faculty of Science > Department of Mathematics.

 Preview
PDF
Language: Croatian

In this work, we have introduced the general factor model and presented principal components factor models and maximum likelihood factor models. We have seen that the starting point of the factor analysis is the data covariance (correlation) matrix. Determining the factor model comes down to determining coefficient matrix and the common factors. We have observed that there is a difference between various factor models in additional constraints and the method of determining the coefficient matrix. After presenting the models we elaborate tests for testing if common factors exist and how many common factors there are. For that purpose, we used $\chi^2$ test and Akaike (AIC) and Schwarz (SIC) information criteria. We showed mentioned in the example. After that we have described the factor estimation. It differs depending on whether we consider the factors as random variables or fixed parameters. Finally, we gave an example where we implemented a factor analysis.