Is the Attitude-Attribution Paradigm Suitable for Investigating the Dispositional Bias?

Abstract

The attitude-attribution paradigm requires people to read an essay that opposes or favors some issue under conditions in which the author is said to have had choice or no choice regarding the selection of the essay stance. The no-choice conditions are considered critical for revealing the dispositional bias; in these conditions, the extent to which the readers attribute attitudes to the author in line with the essay direction is taken as an index of people's tendencies to be dispositionally biased. We hypothesized that this paradigm overestimates the dispositional bias by leading attributers to believe that their task is to glean attitude information from the essay in spite of the author's lack of choice. Attributions from 192 subjects were obtained using the traditional or a modified version of the attitude-attribution paradigm. One important feature of the modified version was a statement to attributers that some of the information that they were to receive might be irrelevant to the judgments that they would be asked to make. A three-way interaction emerged indicating that the modified version had no effect in the choice conditions but significantly attenuated the dispositional bias in the no-choice conditions. We argued that the dispositional bias, although genuine, is overestimated by the traditional attitude-attribution paradigm.