on Student Performance in Object-Oriented Effects of Natural Language Complexity
Ke Li, Jenny Coady, Rob Pooley, Rick G. Dewar · 2009
This chapter examines whether there is more to teaching students how to identify object-oriented (OO) concepts in domain analysis than has perhaps previously been appreciated. Understanding domain descriptions, expressed in natural language, in the early stages of software development is crucial to the suc- cess of a project. A study is presented which measured the performance of senior undergraduate computing students in terms of the types of errors they make as they tried to understand and identify important domain concepts. It concludes that the form of natural language used has an influence on the accuracy of a person's interpretation. In particular, students have a tendency to be confused by complex sentences and domain irrelevant information. In addition, students fail to notice important concepts indicated by infrequently appearing words.