Lisa G

CJ 300 Research Methods

September 19, 1999

The Persuasive Press Inference

Part I

The title of the journal used is Communication Research, published six times annually by Sage Publications, Inc., and co-edited by Michael Roloff of Northwestern University and Pamela Shoemaker of Syracuse University.  There are approximately 118 pages

 per issue, with an estimated five articles in each volume.  Communication Research deals with communication issues found in both the mass media and interpersonal relationships.  For example, topics relevant to the mass media include research on whether television violence can be linked to aggressive behavior, and the effects of mass media bias in presidential campaigns.  Interpersonal communication is examined in articles about the relationship between parental mediation and children’s aggression or those with titles like, “Factors that Inhibit Expressing Concern about Another’s Romantic Relationship.”     
Part II
Gunther, Albert C.  “The Persuasive Press Inference: Effects of Mass Media on Perceived Public Opinion.”  Communication Research 25.5 (1998): 486-501.

This study examined the theory that the mass media can influence an individual’s perceptions of what other people are thinking about an issue.  Gunther wanted to figure out how such an influence occurs. He did an experiment to test whether the “persuasive press inference” (the idea that people deduce public opinion from press coverage and come to conclusions about the influential effect it had on others) would hold true. 
His main hypothesis was that people assess public opinion based on their own reading of the news.  As Gunther succinctly put it, “…people assume that what the mass media are saying today must be what the public will be thinking tomorrow” (487).  He named this proposal the persuasive press inference.
The persuasive press inference stem

s from two already developed theories: the third-person effect and the hostile media effect.  The third-person effect states that people think that the mass media affects others more than themselves.  The hostile media effect concludes that people who hold a partisan view of an issue will think that neutral coverage is biased and influences others towards that viewpoint.  The persuasive press inference combines these two effects by claiming that there is a correlation between media content and media influence, and consequent perceptions of public opinion.

To test his persuasive press inference, Gunther set up an experiment where 128 undergraduate students were each given two articles dealing with controversial issues, which they read and then evaluated.  To gauge media content, Gunther focused on giving the stories a particular slant (bias in favor of or against a stance on an issue). He wanted to see if doing so would change the way people perceived public opinion.  He made four sub-hypotheses about the effects of giving out biased stories. In summary, they proposed that presenting an issue favorably would cause the reader to estimate that public opinion on that issue was favorable as well, that this would be true despite reader’s personal opinions, and regardless of whether biased anecdotes appeared in the article.

For his articles, Gunther picked the topics of the Bovine Growth Hormone (BGH) and congressional motions to cut funding for college loans.  He wrote the stories in four ways: a favorable slant with an anecdote, a favorable slant without an anecdote, an unfavorable slant with an anecdote, and an unfavorable slant without an anecdote.  He created different slants by changing key words or adding reinforcing anecdotes.  

He then gave each student one article on each topic, with the slanted articles put in random order ahead of time.  So for example, a student may have received an article on the BGH with a favorable slant to it, but no anecdote, as well as an unfavorable article on college loan funding with an anecdote.  He told the participants that the articles had been published recently in the Milwaukee Journal and Time.  

After the students read the articles, they answered questionnaires designed to measure their personal views and thoughts on the public opinion of the issues.  To gauge personal opinion, Gunther asked students to state on a 7-point scale whether they strongly supported or opposed the issue. They were also examined about their opinion of the coverage, on the same type of scale, whether the slant was favorable, neutral, or unfavorable toward the issue.  Finally, the questionnaire covered the student’s estimate of public opinion on the issue; what they thought the results of a poll would be, whether opinion had changed since the article came out, and in what direction. After the questionnaire was completed the students were informed that the articles were fictional and that they should be mindful of the fact that the articles may have changed their opinions. 
The results of this experiment supported Gunther’s hypothesis of the persuasive press inference; there was a correlation between media content and media influence.  The students were able to judge the slant of the story, and this slant directly influenced the students’ interpretation of public opinion on both issues. Whether students’ gauge of the direction of public opinion was influenced by story slant had mixed results.  For the BGH issue, the slant made an impact, but not on the loan issue. Slant did have some influence on personal opinion, which affected the students’ estimation of public opinion. Finally, the presence of anecdotes and the order the stories were presented in didn’t seem to have any meaningful impact on the students’ personal view of the issue or their gauge of public opinion.      
Gunther operationalized the concepts of his hypothesis using meaning analysis.  At the highest, most abstract level, his independent variable was “personal reading of press coverage,” and his dependent variable was “inference of public opinion.”  He broke these concepts down to the moderate level by altering story slant and measuring the results with a questionnaire. At the lowest, most concrete level, his independent variable was reduced to articles on the 

use of the Bovine Growth Hormone and proposals to cut