An Examination of the Responses to a Multivalue Logic Test

Summary

Test responses to a standardized reasoning test (the Watson-Glaser Critical Thinking Appraisal, Subtest 1) were examined. Ss read narrative material and evaluated proposed conclusions as true (T), probably true (PT), insufficient data (ID), probably false (PF), or false (F). Logically, different standards should be used for definite (T and F) and probabalistic (PT, ID, and PF) conclusions. However, Ss responses to probable items (e.g., PF) indicate that they treated the probable items as midpoints between insufficient data items (ID) and definite items (e.g., F). Study 1 was based on archival data, the test results of 82 male executives (mean age 36.5). Study 2 was based on an experiment examining test responses of 41 male and female university students (mean age 19.4) given different types of instructions. The second study was conducted to ensure the generality of Study 1 results.