It is generally recognized that people tends to select more confirmative strategy (seaching for confirming evidences) than falsificative strategy (searching for falsifying evidences) in hypothesis-testing. Such tendency is also assumed to explain the difficulty of the four-card problem (FCP) because hypothesis-testing is explicit in an ordinary FCP problem-type. Therefore problem-type in which hypothesis-testing is implicit will facilitate a selection of falsificative strategy. In a present study two FCP problem-types were set to undergraduates: a proposition-testing type and an example-testing type. The main results were as follows: (1) in the example-testing type, falsificative strategy was selected most; and no confirmative strategy was seen; (2) the problem-types had no effect in the arbitrary propositions (having the arbitrary connection between the antecedent and consequent); and (3) selection of falsificative strategy in the example-testing type was seen essential to its selection in the proposition-testing type. These results suggested that it was premature to deny thematic material effect in FCP, and responses to FCP could be interpreted by the Vygotsky's zone of proximal development.