Evaluation of Two Textual Programming Notations for Children

Many researchers have developed many programming environments for children. Typically each of these environments contains its own programming notation ranging from computer code to animated virtual 3D robots and in some case the notation consists of physical objects. While some of these notations were created by examining how children naturally describe computer programs, little research has examined how children understand programs written using these notations. Even less research has examined how children understand programs written using multiple notations. This paper describes an evaluation that compares how children can understand computer programs written using different programming notations: conventional code, English, or a combination of the two. The children were about eleven years old and we measured speed in answering questions about computer programs and the accuracy of their answers. We found that children reading computer programs written in a conventional-style notation were more efficient (faster with no reliable difference in accuracy) than children reading programs written in English. Children with access to a combination of both notations performed between the two other conditions.