Naslov
The Attempts at Identifying ghe Group of "Likely Voters" in Pre-Election Polls in Croatia

Autori
Lamza Posavec, Vesna

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Izvornik
Psychology in Dialogue with Related Disciplines, Abstract Book, 8th European Congress of Psychology, 6-11 July 2003, Vienna, Austria / Edlinger, Sabine ; Mehta, Gerda - Beč : Facultas Buchhandels- und Verlags AG, 2003, 220-220

Skup
8th European Congress of Psychology

Mjesto i datum
Beč, Austrija, 06.07.2003. - 11.07.2003

Vrsta sudjelovanja
Poster

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
vjerojatni glasači; perryjev model; prognostička valjanost
(Likely Voters; Perry's Model; Prognostic Validity)

Sažetak
Background and aim: Low electoral participation generates the need for a coherent predictive model for distinguishing "likely voters" from non-voters. Long research tradition focused on this problem rarely offers universal answers, primarily because of the strong contextual determination of public opinion and political participation. One of the most popular models of identification of "likely voters" is Perry's model, based on nine predictors related to voting intention, previous electoral participation, formal voter's registration, interest in forthcoming elections and politics in general, and the degree of certainty in voting intentions. The aim of this study was to test the applicability of Perry's model of "likely voters" identification and it's affect on predictive validity of pre-election surveys in a specific social and political context in Croatia. Method: Data were collected through pre-election surveys conducted on representative samples of adult population in 1992 (parliament and presidential elections ; N=3D2114), in 1993 (county and local elections ; N=3D3484), and in 2001 (local elections ; N=3D485). Results/conclusion: Results have shown that some of Perry's predictors, (especially the intention to vote, interest in politics in general, and interest in forthcoming elections), can be used as good criteria for identification of sub-samples of "likely voters" and non-voters, and also for weightening the results of the whole sample. Further analysis indicates that, in Croatian context, the main selection of "likely voters" occurs spontaneously through survey acceptance or refusal, and through non-response on questions regarding political choice. This auto-selection of respondents on both levels is highly significant, leaving only a moderate space for improvement of prognostic validity. Political abstinence is also correlated with perception of general social and political situation, and with evaluation of political actors. Therefore, the predictive model of election participation in Croatia should, beside Perry's predictors, include contextual predictors closely related to evaluation of concrete political and election situation.

Izvorni jezik
Engleski

Znanstvena područja
Sociologija

POVEZANOST RADA