STATISTICAL METHODS IN PSYCHOLOGY

Author: CONCEJERO CEREZO PEDRO.
Year: 2004.
University: COMPLUTENSE DE MADRID [www.ucm.es].
Place of defense: FACULTAD DE PSICOLOGÍA.
Place of preparation: FACULTAD PSICOLOGÍA U.C.M..

Summary: ROC is the acronym of English Receiver Operating Characteristic.Una ROC curve is a graphic representation of the success rate (probability that the system detects a signal when it is present) versus the false alarm rate (probability say that the system detects when in fact it is absent) detection yes / no a signal that may be present or ausente.Las curves ROC has been used in various fields of medical diagnosis, for its validity to compare both the diagnostic accuracy of different tests simple laboratory as complete sets of tests, protocols, etc.. including subjective and objective data (for example, used especially in the imaging, which often depends on the opinion of an expert who examined radiography). This argument raises the implementation of this set of techniques for evaluating the effectiveness of certain statistical models usual in market research and in particular pair predicting abandonment of a program fidelización.Los statistical models that are used by the prediction and understanding this phenomenon are from the traditional regression models to data mining (Data Mining). The objective is therefore to use the ROC curves as a tool for measuring the diagnostic accuracy of these techniques in this context particular.En practice, the ROC curves can be obtained through two procedures: parametric or non parametricos.Las first applied in environments experimental and require the assumption of normal distributions, which is why they are known as "curves ROC binormales." latter do not have this requirement, so that its implementation is more prevalent in many fields aplicados.En thesis used both approaches and compares its viability in the event of failure of the assumptions in the case of the parametric approach. The ultimate aim of this thesis is therefore the application, comparison and preparation of a proposal for the implementation of the different methodologies based on ROC curves to predición and diagnostic context-specific market research, such as loyalty.

Author: TIRADO GONZÁLEZ SONIA.
Year: 2005.
University: MIGUEL HERNÁNDEZ DE ELCHE [www.umh.es].
Place of defense: FACULTAD DE CIENCIAS SOCIALES Y JURIDICAS DE ELCHE.
Place of preparation: FACULTAD DE MEDICINA.

Summary: The significant development that have taken Spanish universities in recent decades and development, not least, which is still to come, derived from the integration process and the new articulation of the European Higher Education Area, has placed quality as a the pillars on which to build the school for universities places in this new stage with guarantees of success. From this perspective, recognition of the quality brings the urgent need to not only operationalize the construct to be measured and to provide valid and reliable instruments for this purpose, but also to create mathematical models that enable us explanation of the phenomenon. The overall objective of this work is to construct a mathematical model of the quality of teaching at the University Miguel Hernández de Elche. For this reason, the theoretical framework is the concept of quality checks, procedures for evaluating the quality present in the specific field of universities, introducing the quality system of UMH, he examines the design and construction of mathematical models in the field of psychology and is reviewed and discussed the dispute referred to the concept of satisfaction. To achieve the goal, in the empirical study presents the development and validation estimate of a mathematical model of student satisfaction with teaching in the UMH including, as a preliminary step, the construction and validation of a questionnaire, based on the Theory Classical Tests and supported in the model structure covariances. The creation of a formal model thus enables us to explain what it means quality at the university and what are the most appropriate to take decisions to maintain or improve it.