Key points are not available for this paper at this time.
The present research investigated information processing in situations where there is insufficient evidence to make a judgment and no possibility of avoiding a judgment. The research was inspired by the question answering model of Glucksberg and McCloskey (1981), which stated that, if it is important to find an answer to a question, then the failure of a preliminary memory search to find any relevant information leads to additional attempts, that may employ gradually looser criteria of relevance than that used originally. It was hypothesized that the criterion of relevance could finally become loose enough to include a memory representation of a single and not salient instance. The hypothesis was tested in 3 experiments. In Experiment I the experience of a single instance was provided to the subjects by means of subliminally exposed words, and the dependent measure was subjects' choice of better fitting words. In Experiments 2 and 3 the subjects were provided with single instances by means of a quasi-natural interaction with a person who was similar to 1 of 2 other persons the subjects were subsequently exposed to. The dependent measure was subjects' choice of one of those persons as more friendly, and the choice was made again under quasi-natural conditions. All 3 experiments confirmed the model and indicated that subjects based their judgments on the single instance they were exposed to.
Pawel Lewicki (Fri,) studied this question.