The information quality is widely used in many applications. However, the existing information quality can only deal with the probability distribution. Compared with probability distribution, the basic probability assignment (BPA) in evidence theory is more efficient to handle uncertainty. As a result, it is necessary to generalize the existing information quality. In this paper, a new expression for information quality is proposed to measure the information quality of BPA. When the BPA degenerates into a probability distribution, the proposed generalized expression for information quality in this paper is degenerated into the information quality proposed by Yager. Numerical examples are used to demonstrate the effectiveness of the generalized expression for information quality. In addition, a weighted average combination rule based on the new expression for information quality is presented. A numerical example in target recognition is illustrated to show its validity in combining conflicting evidence.