Salience, importance and evaluation in judgements about people

Three experiments are described which examine the mutual operation of different weighting criteria. The criteria are evaluative direction, and what are termed ‘general salience’ and ‘importance-in-context’. The first experiment establishes a paradigm case where general salience is varied but importance-in-context is controlled. Later experiments extend this paradigm case to stimulus compounds where both general salience and importance-in-context are controlled and to compounds where both are varied. Each experiment employs different stimulus material: trait adjectives, tape-recorded messages and newspaper articles. The three weighting criteria are shown to operate in an apparently non-additive way, in that the influence of one depends upon the state of the others. This is viewed as consistent with a sequential judgement process in which a perceiver searches for a significant starting point and then adjusts his provisional estimate in the light of the other information available to him.