Why these inconsistencies??
Currently, I am inspecting requirements specifications, UML and ARIS models created by my students. And I ask: Why these inconsistencies?? I have told them several times that requirements must be consistent between documents. We even had a session which was dedicated to creating consistency between these models. At least, I told them to use this session for this purpose.
What happened? In one review, I had to them: "You invested double effort for creating confusion!! This is a mistake also made in practice. Nevertheless, it is a grave fault."
Honestly, I do not understand why people do this at all. It could be so simple: You just sit down and write the use cases or the data model once. Seriously. Then, you can reuse this model for all documents you create.
What good reason do people have for re-creating another data model instead of using the one they created last week? Why do people prefer creating two bad, incomplete, sluggish models instead of one good one? Don´t they trust in their own result from last week? Do they simply forget that there is already a data model? Don´t they believe that life can be simple? Do competing subgroups/ persons within the "team" create competing models because they can not agree on one model or do not talk to each other at all? Do they want to punish the person who made the document template and specification guidelines by filling redundant chapters with rubbish pseudo content?
What happened? In one review, I had to them: "You invested double effort for creating confusion!! This is a mistake also made in practice. Nevertheless, it is a grave fault."
Honestly, I do not understand why people do this at all. It could be so simple: You just sit down and write the use cases or the data model once. Seriously. Then, you can reuse this model for all documents you create.
What good reason do people have for re-creating another data model instead of using the one they created last week? Why do people prefer creating two bad, incomplete, sluggish models instead of one good one? Don´t they trust in their own result from last week? Do they simply forget that there is already a data model? Don´t they believe that life can be simple? Do competing subgroups/ persons within the "team" create competing models because they can not agree on one model or do not talk to each other at all? Do they want to punish the person who made the document template and specification guidelines by filling redundant chapters with rubbish pseudo content?
AndreaHerrmann - 2. Jan, 15:14
1) many people don't understand what a model is. It always amazes me when I ask my students about "a" definition of a model. Therefore it is difficult to design a good model, if you don't know what you should describe and what you should leave out.
2) When they look at something from a different side (e.g. UML versus ARIS) they see different importances. Each type of description concentrates on different aspects - even if the model should be the same. You could write an ARIS-model and try to translate it into a sequence diagram but of course there are two main sequence diagrams. Therefore it is already difficult to decide for the one that fits better. (Unless you are really firm in both disciplines.)
3) Forgetting is always a primary explanation that can be used for an explanation. Yes, they do forget, because they are not thinking in the model domain. They are thinking in their own descriptions - and those are not models.
4) They want to make it even simpler:)
5) I often tell the story that the 3 amigos used to discuss about the symbol for use cases for 9 months. (That is what I destilled from various sources about UML.) And the original use case symbol was one that could not be drafted on the computers of that time, unless special programs were used.
6) Well the last question you must answer yourself. Maybe you can explain your new scoring system: "good points score +1, superfluous points score -1. If the total sum is negative, the student has flunked.
-
Difficult however, to carry it through:)
cheers
you are right with everything you write. :-)
Therefore, my course contains a short philosophical excursion about what a model is. You know, the stories like: "What happens if the landscape and the map differ from each other?"
Yes, each model has a different focus. So, it is OK or even good that one model goes more into the details in the right top corner and the other in the left bottom part. But real contradictions are warning signal.
For the next course, I plan to discuss this more, supported by an example. The poor students evidently need more help to understand what they do!