Quantitative measure of evaluative labeling in news reports: Psychology of communication bias studied by content analysis and semantic differential

Abstract

Two studies examined partisan word-connotations in news reports. They focused on what, if any, normative judgements were conveyed through the choice of referent terms of key agents and examined if the usage of terms differed systematically in emotional connotations according to which agents they were applied to. Study 1 used content analysis of every article posted on the Norwegian state news media’s webpage in a one year period from 16th February 2011 to 16th February 2012 (N = 689) on the topic of Libya. Study 2 used semantic differential measures of the most frequently occurring referent terms accumulated from Norwegian subjects (N = 316). This made possible quantitative comparisons of the reference terms depending on the emotional connotations of these words and on what agent they were applied to. The research found evidence for considerable bias in word-connotations. It is suggested that biased word associations are used to convey normative judgement towards news report agents and that these associations can influence readers’ attitudes towards these agents.