G51SWT Feedback 2003
Attendance at the weekly lab sessions was drastically lower than last
year (when there was an assessed coursework each week) and the
average marks for coursework were also well below those last year
(possibly as a result).
- 1st Coursework
- Most people who attempted this managed OK and the average mark was
around 50%.
- 2nd Coursework
- The average mark for this was around 60% partly because an element
of groupwork was involved and most groups managed to get
nearly full marks for that aspect.
- 3rd Coursework
- The average mark was around 40% which was disappointing. This was
partly because the 3rd exercise required the use of recursion
which prevented several people from progressing onto the 4th
exercise (which was worth 1/2 the marks) but another
contributory factor was clearly that for many students this
was the first time they had attempted to write a perl program
despite having been set weekly non-assessed exercises
throughout the module and being warned that this coursework
would be difficult if the assessed exercises were not attempted.
The Exam
Performance in the exam was disappointing and suggested among other
things a basic lack of exam technique with many students unable to
repeat basic definitions given in the lectures and module notes.
Roughly equal numbers of students attempted questions 2, 3 and 4
with slightly fewer (around 70 as opposed to around 100)
attempting question 5.
- Question 1
- This was compulsory with an average mark of around 37%. The
students had been warned in advance that this would be a comprehension
question based on the sample solution of one of the
non-assessed courseworks. As a result it was disappointing
that only about 3 students managed to get full marks for the
first part (describe in your own words what this program
does).
- Question 2
- This was a question on structural testing. The average mark was
around 35%. The biggest problem was that many students did
not know the terminology involved and so were unable to
produce test sets for statement and edge coverage of the code
involved. There were also a number of students who did not
know how to represent loops in control-flow graphs or who were
confused by the layout of the code into thinking it started
with some sort of three-way condition.
- Question 3
- A design question. This was probably the hardest on the paper and
certainly the one with the least clearly defined answers. The
average mark was again around 35%. Again most students were
let down by lack of revision and didn't know what a Blackboard
architecture was nor were able to suggest using a
client-server architecture for a standard web application both
of which were covered in the notes.
- Question 4
- eXtreme Programming. This was the least technical question and
this is reflected perhaps in a slightly higher average of 40%
for the question. Again the biggest problem was a lack of
knowledge of the basic terms and concepts involved in XP.
- Question 5
- This was the most technical question and attempted by
correspondingly fewer students. However in general it was the
more able students who tackled it resulting in an average of
60%. A basic perl programming exercise with, on the whole,
pleasing answers.
The exam revealed a need for clearer notes/guidance on program
architectures and a greater emphasis on textbook exercises in
this area.
However it also revealed a fundamental lack of knowledge of the
basic terminology which prevented possibly otherwise able
students from demonstrating that they could use the techniques
and methodologies involved. It is hard to know how this can be
redressed.
Louise Dennis
Last modified: Thu Jun 5 10:12:52 BST 2003