An Experimental Method Study of User Error Classification in Human-computer Interface

In the paper, we proposed an experimental analysis model of user error and established the error-cognition corresponding relationship according to the error collection-classification method. Through the design of error experiment, various errors in operation interface as well as reasonable and sufficient data were obtained. Based on the repeated experiments and user interviews, we classified errors to extract typical errors according corresponding frequency. According to the cognitive processing, five error types were identified. In the later secondary experiment, we verified the rationality of the error types. The obtained experimental results validated the three basic assumptions proposed before the experiment. Each error can be classified as the identified corresponding error type. In the statistics data, perception confusion is the dominant error type, indicating that operation interface contains understanding deviation. With time limit, error rate was significantly increased. We identified error types in operation interface according to the user error classification experiment method.