Information Search in the Testing of Quantified Hypotheses: How “All,” “Most,” “Some,” “Few,” and “None” Hypotheses Are Tested

Abstract

A series of experiments examined how the search for information in the testing of hypotheses is affected by quantifiers that specify the frequency of the test relation. The experiments showed that a positive search is used more in the testing of propositions in which some instances are predicted to have the test relation than in the testing of propositions in which most or few instances are predicted to have the test relation. In addition, the studies demonstrated that a negative or disconfirmatory search is more likely to characterize tests of absolute hypotheses than tests of hypotheses of the majority or minority form. Finally, the findings indicated that a negative search also is more common in tests of hypotheses predicting the absence rather than the presence of a phenomenon. Thus, as expected, participants generally engaged in a diagnostic search rather than the confirmatory search for information that has been so ubiquitous in previous research.