Real or Artifact? Shedding light on how and when repeated expression can result in polarization
Norris, Meghan Elizabeth
MetadataShow full item record
Researchers have long noted that repeated expression of a judgment can cause that judgment to become more extreme. Three perspectives were explored for why this effect occurs. The first perspective implied that repeated expression results in changes in the evaluation such that it becomes more extreme. The next implied that individuals misinterpret ambiguous response scales and respond based on confidence which results in extremity. The final perspective implied that participants understand response scales but are sometimes incapable of accomplishing the judgment, and invoke confidence as a heuristic to guide responding. Experiment One tested these ideas in a colour judgment paradigm in a 3 (Level of Frequency: 3 vs. 5 vs. 8) x 2 (Question Order: Paired Judgments vs. Separated Judgments) mixed-model design. Repeated expression resulted in greater extremity, and greater confidence in judgments. Confidence was found to mediate the relationship between repeated expression and increased extremity regardless of question order. Experiment Two further disentangled the relationship between repeated expression and extremity by directly manipulating task difficulty, and by manipulating the nature of the response scale labels. A 3 (Level of frequency: 3 vs. 5 vs. 8) x 2 (Type of scale: numerical rating scale vs. colour shade scale) x 2 (Task Difficulty: 144 judgments versus 80 judgments) mixed-design experiment was conducted. Repeated expression again led to increased confidence in judgments. Results showed that repeated expression led to increased extremity when participants responded to the numeric rating scale that was considered ambiguous, but not when using the less ambiguous colour shade scale. Confidence fully mediated the effect of repeated expression on extremity in the numeric rating scale condition. In the colour shade scale condition, the mediation of confidence was offset by a direct negative effect of frequency on extremity. Overall, evidence was found for both the response mapping ambiguity perspective and the evaluative change perspective. Evidence did not support the ability perspective.