Your browser does not support JavaScript!

Home    Search  

Results - Details

Search command : Author="Δαφέρμος"  And Author="Βασίλης"

Current Record: 1 of 11

Back to Results Previous page
Next page
Add to Basket
[Add to Basket]
Identifier 000460791
Title Η πρόβλεψη του εκλογικού αποτελέσματος στο πλαίσιο της διαδικασίας Multinomial Logistic Regression
Alternative Title The prediction of the electoral result in the context of the process Multinomial Logistic Regression process
Author Γκολέμη, Ζαχαρένια
Thesis advisor Δαφέρμος Βασίλης
Reviewer Παπαδάκης Νίκος,
Αλεξόπουλος Νεκτάριος
Χατζηπαντελής Θεόδωρος
Αναστασιάδου Σοφία
Γαγάνης Βαλάντης
Τσαγκαράκης Κώστας
Abstract In this thesis, a Stratified Random Sampling of telephone interviews from the telephone directory of OTE in the 13 regions of the country was carried out on a sample of one thousand people, at the Laboratory of Social Statistics and Political Research of the University of Crete, All citizens with the right to vote had the right to participate in the surveys. Firstly, a brief presentation of electoral behavior and its models is carried out, and we continue with the statistical analysis, in which nine models have heen developed with the aim of identifying the best models, and the synthesis and presentation of the research conclusions. In the model 1, we define as dependent the variable PARTY (parly choice), as Independent variable, SEX (gender) and as reference category, i.e. as comparative measure, the SYRIZA party, with the intention to compare it with all parties regarding the gender and then by applying the Model Fit indices we conclude that we couldn’t consider it as a good model. In model 2 we define as dependent, the variable PARTY (party choice), as independent variables, SEX (gender), CAT_AGE (age categories) and the interaction between them and as reference category, i.e. as a comparative measure, the SYRIZA party, with our intention to compare iL with all parties regarding gender, age groups and the interaction between them. Eased on the above model, we conclude that there is no interaction between the independent variables of our model. In model 3 we define as dependent the variable PARTY (party choice), as independent variables SEX (gender) and CAT_AGE (age categories) and as reference category, i.e. as comparative measure, the SYRIZA party, with our intention to compare it with all the parties regarding gender and age categories and then applying the Model Fit Indices, we lind that the fit of the particular model is good. In model 4 we define as dependent the variable PARTY (party choice), as independent variables, EDUC (Education), CAT_AGE (age categories) and the interaction between them and as a reference category, i.e., as a comparative measure, the SYRIZA party, with our intention to compare it with all parlies regarding education, age groups and the interaction between them. Based on the above model, we conclude that there is no interaction between the independent variables of our model. Tn model 5 we define as dependent ihe variable PARTY (party choice), as independent variables, EDUC (Education), SEX (gender) and the interaction between them and as a reference category, i.e., as a comparative measure, the SYRIZA parly, with our intention to compare it with all parlies regarding education, gender, age groups and the interaction between them. Applying then the Model Fit Indices, we conclude that our model cannot be accepted on the one hand because we have no guarantee of a good fit and on the other hand because it cannot be easily interpreted. Tn model 6 we define as dependent ihe variable PARTY (party choice), as independent variables, EDUC (Education), SEX (gender), CAT_AGE (age categories) and the interaction between them and as a reference category, i.e. as a comparative measure, the party SYRIZA, with our intention to compare it with ail parties regarding education, gender, age groups and the interaction between them. Applying then the Model bit Indices, we conclude that our model cannot be accepted on the one hand because we have no guarantee of a good fit anti on the other hand because it cannot be easily interpreted. Tn model 7 we define as dependent the variable PARTY (party choice), as independent variables, EDUC (Education)» SEX (sex), CAT_AGE (age categories) and questions Q2, Ql 1, Q12, Q13, Q14, Q15 and Q16 and as a reference category, i.e. as a comparative measure, the SYRIZA party, with our intention to compare it with all parties regarding the independent variables. Applying ihen the Model Fit Indices, we conclude thal it can be accepted. Tn model 8 we define as dependent ihe variable PARTY (party choice), as independent variables, EDUC (Education), SEX (Sex), AGE (Age) and questions Q2, Q11, Q12, Q13, Q14, Q1S and Q1 ή and as reference category, i.e. as a comparative measure, the SYRIZA party, with our intention to compare it with all parties regarding the independent variables. Applying then the Model Fit Indices, we conclude that the model can be accepted, i.e. it is interpret able. In model 9 we define as dependent the variable PARTY (party choice), as independent variables, EDUC (Education), SEX (Gender), AGE (Age), the interaction between variables Gender - Education and questions Q2, Q11, Q12, Q13, Q14, Q15 and Q16 and as a reference category, i.e. as a comparative measure, the SYRIZA party, with our intention to compare it with all parlies regarding the independent variables. Applying then the model tit indices, we come to the conclusion that the model can be accepted, i.e., it is interpretahie. We conclude that the estimated by the statistical analysis percentages are very close to the percentages of the Parties that took part in the elections of the 25th January 2015. Also, in the contest ND - SYRIZA, the variables, the ideological reasons, the salvation of the country, the undertaken obligations, the behavior of the citizens concerning the memorandum and the system and linally the Chief.
Language Greek
Subject Electotal behavior
Εκλογική συμπεριφορά
Issue date 2023-07-04
Collection   School/Department--School of Social Sciences--Department of Political Sciences--Doctoral theses
  Type of Work--Doctoral theses
Permanent Link https://elocus.lib.uoc.gr//dlib/d/6/7/metadata-dlib-1701670701-416694-5282.tkl Bookmark and Share
Views 591

Digital Documents
No preview available

Download document
View document
Views : 0