A Study on the Suitability of Visual Languages for Non-Expert Robot Programmers

Abstract

A visual programming language allows users and developers to create programs by manipulating program elements graphically. Several studies have shown the bene ts of visual languages for learning purposes and their applicability to robot programming. However, at present, there are not enough comparative studies on the suitability of textual and visual languages for this purpose. In this paper, we study if, as with a textual language, the use of a visual language could also be suitable in the context of robot programming and, if so, what the main advantages of using a visual language would be. For our experiments, we selected a sample of 60 individuals among students with adequate knowledge of procedural programming, that was divided into three groups. For the rst group of 20 students, a learning scenario based on a textual objectoriented language was used for programming a speci c commercial robotic ball with sensing, wireless communication, and output capabilities, whereas for the second and the third group, two learning scenarios based on visual languages were used for programming the robot. After taking a course for programming the robot in the corresponding learning scenario, each group was evaluated by completing three programming exercises related to the robot features (i.e. motion, lighting, and collision detection). Our results show that the students that worked with visual languages perceived a higher clarity level in their understanding of the course exposition, and a higher enjoyment level in the use of the programming environment. Moreover, they also achieved an overall better mark.