Inferring abilities after influencing performance

Joshua M. Aronson, Edward E. Jones

Research output: Contribution to journalArticlepeer-review


Two experiments investigated the effects of "perceiver-induced constraint" in the realm of ability attribution. In Experiment 1, subjects tried to influence a target person's performance on a series of anagram problems by offering clues. Half the subjects were motivated to enhance the target person's score whereas the remaining subjects focused their clue-giving efforts at enhancing the target person's ability to solve anagrams. The nature of the target person's performance was also manipulated. The pattern of success was either descending or ascending, but equal in overall performance level. As predicted, subjects trying to improve ability saw the ascending performer as more intelligent whereas subjects trying to enhance performance were more impressed by the descending performer, thus demonstrating the well-established primacy effect in ability attribution. Experiment 2 examined the attributions of observers given goal instructions and performance feedback identical to those in the first experiment. All subjects, regardless of the goal orientation of the clue-giver they observed, saw the descending performer as more intelligent, indicating that the results of the first experiment were not solely the result of the experimental instructions. The results suggest that performance-oriented "inducers" were more concerned with the immediate contingency between their interventions and the target persons response. In contrast ability-enhancing inducers were more focused on the long-term contingency between intervention and performance.

Original languageEnglish (US)
Pages (from-to)277-299
Number of pages23
JournalJournal of Experimental Social Psychology
Issue number3
StatePublished - May 1992

ASJC Scopus subject areas

  • Social Psychology
  • Sociology and Political Science


Dive into the research topics of 'Inferring abilities after influencing performance'. Together they form a unique fingerprint.

Cite this