An experimental comparison of ER and UML class diagrams for data modelling
Empirical Software Engineering
Identifying the weaknesses of UML class diagrams during data model comprehension
Proceedings of the 14th international conference on Model driven engineering languages and systems
| Hi-index | 0.00 |
We present the results of two controlled experiments carried out to compare the support given by the ER and UML class diagrams during the maintenance of data models. The experiments involved Master and Bachelor students performing maintenance tasks on data models represented by ER and UML class diagrams. The results reveal that the two notations give in general the same support. In particular, the correctness level achieved by a subject performing the task on data model represented by an ER diagram are comparable with the correctness level achieved by the same subject performing the task on a different data model represented by an UML class diagram. Moreover, by discriminating the levels of ability (high vs. low) and experience (graduate vs. undergraduate) of subjects we also provide some consideration about the influence of such factors on the correctness level achieved by subjects. In particular, we observe that UML class diagrams better support subjects with high ability than ER diagrams, while no difference can be observed considering subjects with low ability. Regarding the experience factor the results reveal no difference in the correctness level achieved by graduate and undergraduate students.