It is said that types of rule revision triggered by counterevidence depend on various factors. The current research examined whether activating a metaknowledge, "theories involving fewer principles and non-ad hoc explanation of certain phenomena are better, " promoted revising prior rules into more generalized rules after encountering counterevidence. In Exp.1, it was confirmed that revising prior rules to more generalized knowledge was too difficult, even when a subject had specified knowledge about counterevidence and other instances. In Exp.2, before conducting a rule generation task, above-mentioned metaknowledge was instructed explicitly to experimental group subjects, while not to control group subjects. It was found that subjects of an experimental group, significantly more than a control group, revised to more generalized rules after encountering counterevidence. This finding indicated that using the metaknowledge affected the type of rule revision and promoted revision to more generalized rules. A general discussion found out how this metaknowledge affected processes of rule revision, and its application to education was offered.