An experimental evaluation of data type conventions

Abstract

The language in which programs are written can have a substantial effect on the reliability of the resulting programs. This paper discusses an experiment that compares the programming reliability of subjects using a statically typed language and a “typeless” language. Analysis of the number of errors and the number of runs containing errors shows that, at least in one environment, the use of a statically typed language can increase programming reliability. Detailed analysis of the errors made by the subjects in programming solutions to reasonably small problems shows that the subjects had difficulty manipulating the representation of data.