Fine tuning selection semantics in a structure editor based programming environment: some experimental results

Abstract

Structure editing holds much promise for improving the quality of introductory programming education. However early structure editors have often been clumsy and counter intuitive to use. This study reports the results of a laboratory experiment in which the user interface of a structure editing environment was modified in several ways to make its semantics more closely resemble what students seem naturally to expect. Analysis suggests that it is possible to improve students' editing performance, without getting bogged down in unnecessary details of language and environmental semantics.