Open Conference Systems, ICQQMEAS2013

Font Size: 
DEVELOPMENT AND VALIDATION OF A SCALE TO MEASURE ONLINE ELECTION CAMPAIGN ENGAGEMENT OF GREEK YOUNG VOTERS USING STRUCTURAL
Prodromos Yannas, Georgios Lappa, Amalia Triantafillidou

Last modified: 2015-09-24

Abstract


Political marketers have realized the significance of new media and have recently began incorporating online tools in the elections campaigns they develop. Given the importance of online election campaigns the aim of this paper is to develop and validate a multi-dimensional instrument for the measurement of online election campaign engagement of young voters during the Greek Parliamentary elections in May 2012. Towards this end and following the framework of Churchill (1979) we conducted two studies. Forty six items that captured all possible online campaign activities of voters were generated based on prior literature review (e.g. Gibson et al., 2010). The first study took place two months prior to the first elections of May 2012. A total of 491 responses were obtained from students of a Technological Institute in a Northwestern city of Greece. An exploratory factor analysis resulted in a six factor solution, namely: Twitter engagement, Facebook interaction, e-negative content generation, e-deliberation, e-voluntary expression of interest and e-visiting. Next, a confirmatory factor analysis was conducted using Amos 8.0, leading to a 6-factor model composed of 19 items. Estimates of the Cronbach‟s alpha coefficient for the six factors ranged from 0.78 to 0.85 indicating a good internal reliability. Furthermore, composite reliabilities of all factors exceeded the 0.70 accepted value criterion (Hair et al., 1998). Hence, the model showed good convergent validity. Moreover, discriminant validity was ensured as the AVE value of each factor was larger than the square of the correlation between the examined factor and the rest of the factors. The second study took place one week after the rescheduled Greek Elections of June 2012. In total, 384 students responded to the survey. Results of confirmatory factor analysis indicate that the proposed scale showed good fit [ρ2(133)=593.32, p=0.000, CFI=0.922, TLI=0.900, IFI=0.923, RMR=0.03]. Cronbach‟s alpha values ranged from 0.77 to 0.91 suggesting good internal consistency. Convergent validity of the scale was established as the standardized coefficients for all 19 indicators were significant and ranged from 0.587 to 0.965 while the values of AVE and composite reliability ranged from 0.56 to 0.71 and from 0.79 to 0.90 respectively. All square correlations between the six factors were lower than each factor‟s AVE, demonstrating discriminant validity. The last step in the scale validation process was to establish the criterion validity of the scale by examining its association with other related variables such as distrust in new politicians and use of traditional media for political information. Results indicate that exposure of voters to traditional campaign tools (i.e. political ads) affects significantly and in a positive manner all six online election campaign engagement activities. On the contrary, the level of distrust in new politicians influences negatively voters‟ online campaign engagement except the e-voluntary expression of interest dimension. The study‟s contribution lies in the development and refinement of the 6-factor 19-item model for measuring youth electoral online engagement. The study‟s implication for political marketers and communications consultants is that traditional and new media tactics should be used in tandem in order to gain the most out of electoral campaigns. Specifically, traditional media should be used in order to decrease the levels of political distrust towards new candidates and in turn to increase the online engagement of voters

Full Text: PDF