Media Effects, Agenda Setting, Framing and PrimingBack

Winnie Wing-Yi Lee and Stephanie Hiu-Nga Tse
(Members of Public Opinion Programme, the University of Hong Kong)

“Online Engagement and Political Participation: Reception, Expression and Sharing in Facebook Groups and Discussion Forums”


Stephanie Jean Tsang, University of Wisconsin- Madison, USA


This presentation discusses college students’ use of the Internet for political purposes with regards to one of the most controversial issues - the Home Ownership Scheme (HOS) of Hong Kong in 2011. The reception, expression and sharing of the Scheme in Facebook Groups and discussion forums were assessed in relation to political participation.


The presenter conducted surveys in two universities in Hong Kong during March and April 2011, and collected a total of 863 valid samples. Results indicated that “Facebook Group use” relates positively to both “willingness  to participate” and “willingness to protest”, but found no evidence of relationship between forum use and participation.


Concerning the relation between “willingness to participate and protest” and different kinds of specific uses, namely, reception, expression and sharing, of Facebook Group and forums, no significant relationship was found regarding the specific uses of Facebook Group. However, evidence suggested expression in forums negatively relates to one’s willingness to participate and protest, while sharing in forum showed a positively relationship.


Facebook Group differs from discussion forums in being an ideologically homogeneous platform. Specific uses within Facebook Groups are not good predictors of willingness to participate, as people tend to join groups to perform “expressive information sharing”, and lack the motivation and effort needed to be involved in expression and sharing in such Groups.


As for discussion forums, people might gain a sense of empowerment after expressing their opinions online and felt they had fulfilled their job as good citizens. The significant positive relationship between forum sharing and the willingness to participate is not surprising as getting into the forum needs motivation and incentive, and sharing is not as easy as just a click in Facebook Groups.

“Voter Advice Applications in Practice: Answers to Some Key Questions from Turkey”


Thomas Vitiello, Sabanci University, Turkey
Ali Çarkoğlu, Mert Moral, Koç University, Turkey


This presentation investigates the effect of the media events that promoted the Turkish Voter Advice Application (VAA) website ( in the 2011 Turkish Parliamentary elections by running a multiple-interrupted time series (MITS) model.


Since the coming to power of the conservative Justice and Development Party (AKP), the freedom of media has been at stake. “Political parallelism” is common in Turkey, media organizations are associated not with particular parties, but with general political tendencies. Effective partnership with media is critical in reaching a large and well-balanced sample of the electorate.


The project team launched their own “mini-campaign” of the Oypusulasi website as a “double-screen campaign” since it involved both traditional media events and internet-based events. The impact of multiple interruptions caused by various formats of media was evaluated.


The study revealed that VAA website users appeared to have been brought to the tool through multiple sources of information including traditional media sources (e.g. television and newspapers) and internet-based communication (e.g. social networks and online media websites). Even with the lack of national media partners, significant media attention has been generated towards the tool.


In an electoral campaign, the main objective of a VAA is to provide its users with an electoral advice reflecting basic policy preferences. Findings of this study seemed to show that users with political tendencies in line with the media outlet are more likely to connect to the VAA.


The Oypusulasi website was the first VAA ever run during an election campaign in Turkey in response to the growing partisan alignment of the media in the country. It is believed that this political parallelism is also what made the promotion of the VAA more difficult. Yet, the diversity of the media discussing the VAA proved to be essential in order to reach a heterogeneous group of users. This Turkish case can provide an interesting ground for comparison both with data collected through VAAs in Western countries, and with data issued from the VAAs that recently went online in some of the Arab Spring countries, such as Egypt and Tunisia.


“Discussing Occupying Wall Street on Twitter:  Longitudinal Network Analysis of Equality, Emotion, and Stability of Public Discussion”


Chengjun Wang, Pianpian Wang, City University of
Hong Kong, Hong Kong


To evaluate the quality of public discussion on Twitter and to understand the evolution of longitudinal discussion network, the presenters analyze the tweets of Occupying Wall Street for 16 days by investigating the relationships among equality, emotion, and stability of online discussions. Data was retrieved from Twitter on topic about Occupy Wall Street by R-shief (, containing 1,353,413 tweets in 16 days (between Sep 24 and Oct 10, 2011), among which,
88,601 discussions are identified by the symbol of “@”.


The features of public opinion expressed on twitter were appraised and the positive and negative emotions for each tweet were measured with sentiment analysis. The results revealed that: First, participation is highly unequal in both initiating a discussion and being spoken to in the conversation. Second, the stability of discussion for receivers is much higher than that of senders. Third, inequality moderates the stability of online discussion. Fourth, the emotion expressed in online discussion is relatively balanced and stable. No significant relationship was found between emotion and political discussion (e.g. frequency, and standard deviation).


The findings revealed that both emotion extremity and political discussion are power-law distributed, and emotion has a strong linear influence on both frequency and stability of political discussions. The inequality of political discussion appears to have influenced the stability of daily participation of online political discussions. Mentioning other people using @ draws the attention of the target users and mobilize them. People differ greatly in their participation in online political discussions, and the conversations tend to be emotion-free, and individual senders are more likely to keep mentioning other users. Individual receivers’ possibility of receiving a conversational tweet then declines, indicating that there might be more individuals involved in Twitting later on.

“Selectivity in Blogosphere: the Potential for Exposure to Political Information in Non- Political Blogs”


Rebecca Ping Yu, Yu Won Oh, University of Michigan, USA


This study tests the selectivity hypotheses by analyzing the extent to which popular phrases in online spaces during the 2008 U.S. presidential elections are presented in political and non-political blogs. The findings revealed that, while relatively infrequent, political phrases appear in non-political blogs, and cross-ideological citations exist in blogosphere.


The presenters used Google Blog Search to identify blog posts that contained the selected 100 political and non- political phrases. Results showed that while political phrases are more likely than non-political phrases to appear in political blogs, political phrases exist in non-political blogs. These partially support the hypothesis about interest-based selectivity. That is, while those who are not interested in politics can keep themselves away from political blogs, they may happen to encounter political information in other non-political blogs they visit.


The context in which political and non-political phrases appear was explored. The findings revealed that types of phrases are associated with levels of agreement. Overall speaking, when citing political phrases in blog posts, bloggers are slightly more likely to challenge the political phrases rather than supporting the phrases. In contrast to political phrases, non-political phrases are more likely to be engaged in a positive manner in the blog posts.


The levels of disagreement between the blog posts and the phrases are found to be positively and significantly related to cross-ideological citations. In other words, when partisan bloggers cite political phrases across ideological lines, they tend to disagree with the stance of the political phrases. In particular, right-leaning phrases are more likely then left- leaning phrases to be cited in opposite-ideological blogs.


Winnie WY Lee and Stephanie HN Tse (Editors)
Public Opinion Programme,
The University of Hong Kong