An Investigation into Design Diagrams and Their Implementations
Alan Hayes, Thursday 3rd June, 10:45 am - 11:00 am, Student Presentations 1, JLB Meeting Room 10
Cloud created by:
28 May 2010
Part-time student in his 5th year.
Research question: link between the student-submitted design and its implementation to be assessed in such a way to get formative feedback from a system.
Showing some diagrams – i.e. a student produced diagram of the employment and payroll and a diagram produced from student code
Errors from example
- Missing method from the payroll classs
- The relationships between the classes have been incorrectly represented
Feedback to be generated: identifications of relationships, missing methods and attributes from the employee and payroll class
Taxonomy of student errors: erroneous or missing data in the design, incorrect relationships
How to represent this:
Multiple views of the same component: such as the student diagram, tutor diagram, student code and tutor code
His system looks at providing automotive feedback to the student – which includes the good practice of the student
Convergence of literature: looking at the assessment of diagrams, assessment of implementation (and something else)
One assessment structure is:
Student design -> Forward engineering -> Inferred code
Suggested for the student diagram there might even have apply tagging so that the student diagram could be tagged. A reverse engineering could also be done and make a comparison between the two.
Has a diagram on the inputs of the systems. He indicates that he has collected 6 data sets – and he needs to anonymise the data and has divided into two parts: one part for experimentation and the second for evaluation.
Some students’ diagrams are in different formats which make it difficult to analyse and hence sometimes had to hand-produce the tagged structures. Some labels are quite similar and need to determine when they are the same.
Are you hoping the students would evaluate the tool? He is going to take academics and ask them to generate comments from their own and also from the tools and compare these two to see the efficacy.
How are you approaching the problem of comparison (diagnosing, formative feedback) – it is a very challenging problem? The tagging structure that the students’ use – he loads it up in a java-based programme – and then goes through and match things – and from that he develops a heuristic score. The heuristic is developed from the literature. The generation of feedback is a consequence of the comparison (hope I got that right!).
09:25 on 3 June 2010