Policy conflict and policy change as a function of task characteristics III. The effects of the distribution of the validities of the cues in the conflict task

Abstract.— The effects of the distribution of the validities of the cues in the conflict task on policy conflict, as studied in Hammond's "lens model" interpersonal conflict paradigm, was investigated in a 2 (levels of task predictability) by 2 (levels of cue validity distribution: only one cue valid vs. two equally valid cues) factorial experiment. The results showed that the level of agreement was higher when only one cue was valid than when both cues were valid, but this effect was obtained only when task predictability was high. In the low predictability condition, there were no differences between the two levels of cue validity distribution. In agreement with earlier findings, the results also showed that agreement was a positive function of the predictability of the task. These results show that, to understand policy conflict, it is not sufficient to analyse only the characteristics of the parties to the conflict; it is also necessary to analyse the characteristics of the task facing the parties.