This experiment was undertaken in an effort to discover whether the different types of linguistic responses involved in the process of inductive inference could be investigated accurately by purely objective methods. The material used consisted of lists of Chinese characters which sixty subjects were asked to memorize. The subjects understood the experiment to be a test in memory, and were not aware that their spontaneous and incidental linguistic responses regarding the presence of common radicals in the various Chinese characters and the relation of these radicals to the meanings of the compound characters were being noted and used by the experimenter as data for the study of induction. On the basis of the results of the experiment the author classified the process of inductive inference into five aspects, namely: (1) reaction to relationship; (2) number of evidences; (3) nature of the hypothesis; (4) the detection of negative instances; and (5) the attitude toward the negative instances. Every aspect showed various subdivisions in accordance with the different modes of response in the experiment. The author is convinced that the results of the investigation demonstrate the possibility of studying the so-called higher mental processes without using the method of introspection. From Psych Bulletin 21:03:00230. (PsycINFO Database Record (c) 2006 APA, all rights reserved)