Twitter as Public Salience

BY: CHRIS J. VARGO UNC CHAPEL HILL

THE TWITTER opportunity

This study considered Twitter as a source of public opinion (public salience) as it pertains to agenda setting.

AN AGENDA SETTING ANALYSIS

2 ISSUES WERE TRACKED IN PRINT, TV and Twitter

The BP Oil Spill and The Mortgage and Housing Crisis were tracked for a total of 92 days. First and second-level agenda setting variables were coded, and a time series analysis was conducted.

A mild relationship between media salience and public salience was shown.
LITERATURE REVIEW

Agenda setting looks to see if the media control what topics the general public holds salient. (McCombs & Shaw, 1972)

With a high correlation, the study found that there was a relationship between what the news was reporting and peoples most importantly perceived issues.

RESEARCH QUESTIONS H1: A DAYS MAINSTREAM TELEVISION

Were the media simply influenced by the publics agenda, or were the media doing the influencing?

To answer that question of causal order, McCombs and Shaw conducted another study in 1977 that used two different points in time to help establish the order of relationships. That type of lagged correlation helped show that the medias agenda served as a predictor of the publics agenda.

COVERAGE OF SELECTED ISSUES IS A POSITIVE PREDICTOR OF THE VOLUME OF TWEETS PUBLISHED ON THAT SAME DAY.

The second level of agenda setting serves as an enhancement of the media salience variable. (McCombs et al., 1997) (Craft & Wanta, 2004) Newer articles have tackled Internet-driven social media.

H2: A DAYS MAINSTREAM NEWSPAPER

COVERAGE OF SELECTED ISSUES IS A POSITIVE PREDICTOR OF THE VOLUME OF TWEETS PUBLISHED ON THAT SAME DAY.

In a 2010 study of agenda setting and YouTube, Sayre et al. investigated if, when and to what degree videos posted on YouTube may have led or followed traditional news media. Another 2010 study looked at mainstream media and a new measurement, Google Trends, for a possible correlation. That study explored a potential relationship between mainstream media coverage of a particular political rumor and its public salience as measured by online search activity. Both found a same-day relationship between the new media and the traditional media.

H3: MEDIA SALIENCE WILL TRANSFER TO

PUBLIC SALIENCE WITH A ZERO-DAY LAG AND BE THE STRONGEST ON THE SAME DAY.

H&M TWEETS

TV STORIES NEWSPAPERS

BP Oil Spill

HOUSING & MORTGAGE

BP TWEETS

TV STORIES NEWSPAPERS

Method & Analysis

The study collected data from Oct. 12, 2010 to Jan. 11, 2011. Daily Tweet frequency counts were taken for each issue. Daily National Newspaper articles were counted (ProQuest).
Number of Stories Number of Words

SIGNIFICANT PREDICTORS FOR TWEETS

BP OIL!
TV # OF STORIES NEWSPAPER # OF STORIES TV # OF SECONDS TV # OF STORIES

MORTGAGE & HOUSING CRISIS!

2 of 4
IV Predictors

2 of 4
IV Predictors

Daily National Television stories were counted (The Vanderbilt Television News Archive).
Number of Stories Number of Seconds

FINAL THOUGHTS & IMPLICATIONS

Bivariate Correlations were done to determine if significant correlations existed between DVs and IVs. OLS Regressions (Durbin-Watson statistic) were calculated for correlation and regression analysis ARIMA was calculated as a test for the predictor variables.

AD&PR Practitioners should plan campaigns knowing that stories typically break the same day on Twitter, TV and Print. The study shows promise for a relationship between traditional media and Twitter (social media).
Further study will be needed to reinforce these findings.

Due to same-day correlation, relationship order could not be determined.

Future studies should consider increased time intervals to better determine relationships.