This thesis addresses two large blocks of objectives: the analysis of the factors of choice public university in Valencia and modeling of these processes of choice.These aims are divided, in turn, in research areas: analysis of the underlying motivation to undertake university studies, analysis of the factors influencing the choice of university and degree, the sources of information used by the and students and the influence of third parties in this process, as well as the reasons why the student ultimately chooses the university and the degree in which he was admitted;and choice modeling using binomial logistic regression models and artificial neural networks, in particular multilayer perceptron models (MLP) and radial basis function (RBF). To achieve the goals defined, this thesis consists of a total of six empirical studies that have been organized in three trials in a thesis that is divided into three parts and consists of seven chapters. In terms of methodological design, the thesis develops a phase of qualitative research and two phases of quantitative research.The first quantitative research work took place in FormaEmple@ 2010 developing 401 enterviews to 4th year secondary school students. The second quantitative research was developed in 2013, using a sample composed by students admitted in Valencian Public Universities, respecting quotas by university, and within each institution, quotas of campus, branch knowledge and gender. The results of this thesis are consistent with research conducted at the international level, and their contributions can be grouped into three main categories: academic, methodological and management.From an academic perspective it addresses the college selection process from a marketing perspective educational approaches aligned with establishing a parallel with the election processes in the context of consumer behavior, addressing the different stages of the student's choice. From a methodological perspective, this thesis employs techniques of univariate and multivariate analysis, and analysis techniques such as advanced data binomial logistic regressions and artificial neural networks. From a management perspective, the implications of this thesis can be extended both the universities and the education system as a whole, as policy-makers.
2016-02-26
