A decision-making analysis of persuasive argumentation and the choice-shift effect

Amiram Vinokur, Yaacov Trope, Eugene Burnstein

Research output: Contribution to journalArticle

Abstract

A subjective expected utility (SEU) decision-making analysis was performed on the content of arguments generated by subjects privately or during group discussion in response to choice-dilemmas shown to shift toward risk and caution. It was demonstrated that arguments generated in private as well as those emitted in discussion were concerned with evaluating outcomes given in the choice-dilemma, that is, with assessing utilities. Relatively few focused on the probability of achieving these outcomes or on utilities for risk and caution per se. Furthermore, the distributions of private and discussion arguments over content categories were very similar. This suggests discussion does not elicit new arguments (e.g., about the value of risk or caution per se) which have not already been considered privately before discussion. This finding is contrary to value theory and consistent with the persuasive-arguments formulation. Additional analyses disclosed both sets of arguments were directed toward changing utilities so as to make (a) risk (low probabilities) highly attractive on items shifting toward risk, and (b) caution (high probabilities) highly attractive on items shifting toward caution. Moreover, the import of the arguments, their direction of influence, was correlated with the actual shifts in choice at the group level. Thus, our results provide strong support for an SEU interpretation of argumentation effects. Implications concerning the "risky shift" as an instance of attitude change produced by informational influence (argumentation) were discussed.

Original languageEnglish (US)
Pages (from-to)127-148
Number of pages22
JournalJournal of Experimental Social Psychology
Volume11
Issue number2
DOIs
StatePublished - 1975

    Fingerprint

ASJC Scopus subject areas

  • Social Psychology
  • Sociology and Political Science

Cite this