Engagement with Counter-Attitudinal Content in Korea and the US

The third speaker in this ICA 2024 conference session is Claire Wonjeong Jo, whose focus is on the effects of cross-cutting exposure – these are seen as including both a better-informed citizenry and greater attitude and affective polarisation, and/or no effects at all. Past research draws largely on survey data, and measure a range of attributes; but perhaps there is a way to observe the actual news use behaviours of participants that provides more direct empirical data.

This study compares such data for South Korea and the United States; it applies social judgment theory, expectation violation theory (which suggests that unexpected messages will be more effective than attitude-congruent messages), and the black sheep effect (which instead suggests that unexpected messages will be less effective). Overall, then, the distinction is between expected and unexpected messages from in-group and out-group media sources. This is further linked to inputs drawn from social identity theory and levels of partisanship.

The South Korean data were drawn from a period of government changeover in March to May 2022; they were obtained from the Naver platform, where users’ past comments were used to determine their underlying political positioning, and partisans of both sides of politics turned out to reject some 91% of all counter-attitudinal content (from a sample of eight articles); conservatives did so more strongly than liberals.

This was then repeated with a larger dataset from the US: here, the study used the Yahoo! News platform, where users’ attitudes were again inferred from their previous commenting behaviours, and discussions by some 2,600 users about the 2024 election were observed. The patterns here were the opposite of the Korean dataset: US liberals were more negative towards counter-attitudinal content from in-group sources.