Concept Learning is a problem-solving that has been widely in the study of individual cognitive process. The educational psychologists have concorned with the conditionsunder which children learn effectiviy and economically. In designing a typical experiment in this area, experimenter constructed a concept class consisting of all the instances generated from all the possible combinations of a number of attributes(eig., shape, color, etc), each of which has two or more value(eig., triangle or square, red or blue, etc). Generally, two major discussions in concept learning situation have been problems how to present subjects stimulus cards and to give subjects either the first illustrative cards in positive instance or in negative instance. Experimenters have usually called the the former as selection strategy and the latter as reception strategy. The two strategies is called as presentation paradigms. The experimental tasks were composed of three disjunctive concept such as; “circle or yellow,” “border line one or figure number two,” “blue or bordor line two,” in this experiment. Using such disjunctine concept tasks, this study was ready to investigate the effects of independent variables about presentation paradiyms, illustrative methods, tasks under following three major hypothesis conditions; Ⅰ. Subjects will find better performance with the reception pardigm than with selection paradigm. Ⅱ. Subjects will find a significant difference among illustlative methods. 1. Subjects will find better performance with negative illustrative instance than with the others in stances. 2. Subjects will find better performance with positive illustrative instances than with none illustrative instance. Ⅲ. Subjects will find a significant difference among tasks. The 180 subjects were randomly sampled in grade 3 of junior high school. They were randomly assigned to pix experiment groups in each 30 subjects to 2×3×3 factorial design. The points obtained by trials, noninformative hypothesis and redundant errors were statistically computed in repeated measure variance analysis and in Newman-Keuls individual comparisons recommended by winer. The results in this study were following such as; Ⅰ. Subjects found better performance with reception strategy than with selection strategy in disjuctiue cancept learning. Ⅱ. Subjects found a significant difference among illustrative methods; 1. Subjects found better performance with negative illustrative instance than with the other instances. 2. Subjects found no better performance with positive (instance than with none illustrative instance. Ⅲ. Subjects found no significant difference by trials, but found significant difference in noninformative hypothesis and redundant errors.