Στατιστικές τεχνικές για ανάλυση ψηφοφοριών.
Statistical techiques for voting analysis
Σταμπέλου, Θεοδώρα - Παρασκευή
In these days of economic crisis, rapid political changes and global political fluidity, the final outcome of every voting, cannot be considered known beforehand. This fact constitutes the biggest challenge for Public Opinion Research companies, which, before every general election invest huge amounts of money on carrying out exit polls, aiming at a considerably precise and reliable prediction of the result. An important tool to achieve that, is a number of statistical models used by analysts in order to predict the outcome. The main objective of this thesis is to describe the process of building such prediction models, as well as the statistical techniques of analyzing the exit poll’s results which, except for predicting the outcome, provide more information about the social characteristics of the electoral body. Moreover, we will illustrate a prediction method different to that of the exit polls and less widespread in the public opinion research companies, the method of simulation.