In this study, an interactive programming learning environment was built with two types of error prompt functions: 1) the key prompt and 2) step-by-step prompt. A quasi-experimental study was conducted for five weeks, in which 75 sixth grade students from disadvantaged learning environments in Taipei, Taiwan, were divided into three groups: 1) the key prompt group, 2) the step-by-step prompt group, and 3) the group in which teachers explained the errors. The aim was to investigate the effects of different interactive methods on students' learning in programming. The results showed that the step-by-step prompt group had better learning effectiveness, self-efficacy, and problem-solving ability than the key prompt group and the group with the teacher explaining the errors. Because the students from the disadvantaged learning environment were new to the graphical program, they had a low level of understanding of the program. The interactive learning environment with step-by-step prompt, as opposed to the key prompt, made it easier for disadvantaged students to understand the problems and think about their solutions.