Product name and version: Visual Model Editor, M32 Release
Organisers of the test: FH-OOE
Date of the test: October 2008 - January 2009
Date of the report: 05.04.09
Editor: Christoph Richter
Contact name(s): Christoph Richter
While students were quite critical about the overall performance and usability of the Visual Model Editor, the overall idea of the tool was assessed quite heterogeneously. While about one-third said that the VME provides a clear structure and allows to see connections between topics, an equally large group said that the structure is not clear and confusing. This ambiguity is also mirrored in the fact that despite rather poor ratings of the VME’s utility about half of the teams used the VME regularly.
Besides frequent complaints about the overall performance of the SSP, the field trial also revealed some limitations regarding the usability of the VME. Main problems and recommendations derived from students’ reports are the following:
• Lacking overview when trying to access detailed information - Suggestions: In order to ease orientation while being zoomed into a model an outline map might be useful providing and indication of the part of the entire model that is currently displayed. Furthermore, resizable nodes and hideable sidebars would be helpful to display larger parts of a model on a single screen.
• Changing a model is tedious - Suggestion: Users should be able to change the type of a node or link.
• Size and display of node-contents - Suggestions: To provide more flexibility to the user longer titles for nodes should be allowed, given that nodes could be made resizable. Furthermore, the description of a node should be displayed in the information tab once the node is selected.
• Changes are difficult to trace - Suggestions: It is s suggested that users can retrieve a chronological list of changes made to model.
• Complex GUI - Suggestion: To reduce the complexity of the GUI both features should be integrated into a single tab.
Apart from these problems some students reported that they found the use of predefined modelling languages limiting and had sometimes had hard times to find a suitable concept. This complaint backs up the ideas underlying the development of the VM(L)E as it highlights the need for flexible and extendible scaffolds.