top of page

METHOD

 

During phase one, target respondents – online news readers – were reached by an online news survey posted from December 20, 2013 to February 20, 2014. Respondents were solicited to fill-out the survey through announcements placed on Facebook, @lolmythesis, and Twitter.  Each announcement contained the survey URL for quick and easy access.  A Facebook group page was also created to promote the survey; members were asked to complete the survey.  We promoted a “snowball” method for targeting respondents.  After completing the survey, respondents were encouraged to ask their friends and family members to also complete the survey.  This convenience sample yielded 59 completed surveys, and another 23 incomplete surveys. All incomplete surveys were discarded from analysis.

 

DATA ANALYSIS

 

Scores were calculated for believability, fairness, accuracy, and depth of the online news sites using weighted averages. These scores were then combined to create a credibility index, ranging from 0 to 4 for each source (see Table 1). Regressions were run on the full sample (n=59) to test the strength of the credibility index as a proxy for selective exposure and to predict whether lack of credibility in online news sources is associated with argument-seeking behavior. Chi square tests determined if there is a correlation between ideology or education levels or gender and argument-seeking behavior.

 

The sample was then divided into two subsamples and compared for differences.  Subsample one consisted of argument-seeking respondents (n=13), and subsample two consisted of nonargument-seeking respondents (n=46). Regressions tested the strength of the correlation between formation of opinions and confidence levels and to determine if likelihood of readership is strongly connected to argument-seeking behavior.

 

 

 

 

 
METHOD

 

Phase two was conducted between 22 February and 2 March 2014.  Th interviews targeted a subsample of survey respondents based on their survey responses and their willingness to be interviewed. Interview participants were selected if their survey had indicated they do seek out opposing viewpoints and do enjoy arguing.  Of the 59 completed survey respondents, 13 met the interview criteria, and 4 were interviewed.

 

DATA ANALYSIS

 

Due to logistical constraints, two interviews were over the phone and the other two through email exchanges.  A two-step process was implemented to verify findings: independent coding by the two researchers and participant validation of the analysis. 

 

 

 

METHODOLOGY & DATA ANALYSIS

 

This study used an explanatory sequential mixed methods design.

PHASE ONE

 

PHASE TWO

 

bottom of page