Keywords: New sources of information on public opinion and the use of social media to voice public opinion

Background
With the multitude of political news stories that are circulated among a myriad of social media sites, it would seem that users would be knowledgeable and informed about the latest events (Eveland, Marton, & Seo, 2004). But exposure does not necessarily mean that news is ‘learned’- cognitive reasoning is also needed to stimulate political learning. The Cognitive Mediation Model (CMM) (Eveland, 2001) and the Orientations-Stimuli-Reasoning-Orientations-Response (O-S-R-O-R) model (Cho et al., 2009) propose news elaboration as making cognitive connections to past experience and prior knowledge and deriving new meanings from news content. These models, however, do not consider how cognitive reasoning takes place while using social media.
In today’s digital world, consumers reconstruct, reformulate, repurpose, reframe and share political news through social media. This study conceptualizes this behavior as curatorial news use, which plays a vital role in ‘learning’ through intense involvement with the news (e.g., Brightwave, 2014; Mihailidis, & Cohen, 2013).
This study conceptually differentiates curatorial news use from news exposure or news elaboration, which are operationalized as precursors to curatorial news use. Using the CMM and O-S-R-O-R model this study investigates (1) the direct influence of exposure to political news on social media on levels of political knowledge; (2) the direct influence of social media news elaboration on levels of political knowledge; (3) the direct influence of curating social media news on levels of political knowledge; and (4) whether news elaboration and curatorial news use mediate the influence of exposure to political news on social media on political knowledge.

Method
Data were derived from a representative survey of 1,135 South Korean adults. Data were collected in South Korea between November 3 and November 30, 2016. A survey research firm constructed a sampling frame of 5,000 Korean voters. Stratified quota sampling based on age and gender data of registered voters assured representativeness of the population.
The direct and differential effects were tested using hierarchical regressions. To test the mediating effects, news elaboration and curatorial news use were entered as part of a multiple mediator model using ordinary least squares (OLS) regression procedures (Hayes, 2013).

Results
The findings reveal that exposure to social media for political information is directly related to learning political issues. Second, the results indicate both a direct and indirect influence of news elaboration on political knowledge. Third, political learning is an outcome of curatorial news use. Lastly, curatorial news use on social media mediates the association between exposure to political information and political knowledge. In other words, curatorial news use is a mediating variable that connects social media news exposure to political learning.
That curatorial news use is more strongly associated with political knowledge than news elaboration adds two theoretical implications; curatorial news use is conceptually different from news elaboration, and curatorial news use inherently arises from news elaboration.
This study demonstrates that CMM and the O–S–R–O–R model could be adapted to fit the current social media environment.