MSc Software Engineering

Cohort: MSE/06/PT

Examinations for 2006 - 2007 / Semester 2

MODULE: SOFTWARE QUALITY ASSURANCE

MODULE CODE: SEM5201

Duration: 2 Hours

Instructions to Candidates:

1. Answer all questions.

2. Questions may be answered in any order but your answers must
show the question number clearly.
3. Always start a new question on a fresh page.
4. All questions do not carry equal marks.
5. Total marks 70.

This question paper contains 3 questions and 4 pages.

Page 1 of 4
SB/102
 ANSWER ALL QUESTIONS

QUESTION 1: (20 MARKS)

(a) A programmer claims that because only a small proportion of software errors
turn into software failures, it is unnecessary to make substantial investments
in the prevention and elimination of software errors.
i) Do you agree? Justify your answer. (2 marks)
ii) Discuss the outcome of accepting these views. (4 marks)

(b) John is an exceptional programmer. Testing his software modules reveals (4 marks)
very few errors, far fewer than the team’s average. He keeps his schedule
promptly, and only rarely is he late in completing his task. He always finds
original ways to solve programming difficulties, and uses an original,
individual version of the coding style. He dislikes preparing the required
documentation, and rarely does he do it according to the team’s templates. A
day after completing a challenging task, on time, he was called to the office of
the department’s chief software engineer. Instead of being praised for his
accomplishments (as he expected), he was warned by the company’s chief
software engineer that he would be fired unless he began to fully comply with
team’s coding and documentation instructions. The position taken by the
department’s chief software engineer is fully justified, could you suggest why
his/her position was so decisive?

(c) Most of the people claim that testability and verifiability are actually different (4 marks)
names for the same factor. Could you explain why?

(d) Some professionals claim that increased software usability necessarily (4 marks)
involves decreased efficiency. Others claim no dependence between
software efficiency and usability. Do you agree with the first or the second
group? Discuss your answer.

(e) It is clear that a software development project carried out by a software (2 marks)
house for a specific customer is carried out under content and timetable
obligations, and is subject to the customer–supplier relationship. Discuss
whether a customer–supplier relationship is expected when the software
developed is to be sold to the public as a software package.

Page 2 of 4
SB/102
QUESTION 2: (20 MARKS)
(a) The XYZ Council has decided to develop a software package that will serve
the youth clubs operated by the city. The software’s main tasks will be:
- Follow up of monthly payments of the members.
- Preparing lists of participants in the various courses offered by the clubs.
- Production of reminder notices to course participants who fail to appear
regularly.
- Statistical reports about membership and participation in club activities.
The city already implements the following software packages:
- Tax collection.
- Public library.
- School follow up and achievements control.
- Water consumption billing.
The City Council has asked the Information Technology Unit to report to
the council about the possibilities for reuse of the city software packages
already available to the city in the youth club software package.

i) Could you suggest which modules of the existing city software packages
could be reused in the new software? Present the reuse modules in a
tabular form. Give assumptions about the contents of the existing
(4 marks)
software packages and the required new software.

ii) Could you grade the reused modules suggested in (i) according to the
scope of adaptation efforts required to apply the reused module in the
youth club software package. (4 marks)
(b) Performing a contract review raises many difficulties.
i) List the “built-in” difficulties to carrying out a large-scale contract review. (2 marks)
ii) List the steps that should be taken to make a large-scale contract review
feasible. (2 marks)
(c) It is said that failure to meet the interoperability requirements can negatively
affect the correctness level of the software system, and even can cause
nonconformance with correctness requirements. Elaborate on the above
statement and explain the mentioned interconnections between factors. (3 marks)
(d) An SQA professional claims: “I find all the reasons given for a proposal draft
review to be justified. I also believe that a review contributes to the quality of
the proposal, especially in clarifying and precisely defining requirements, and
in preparing more realistic estimates, among other issues. However, once
the proposal has been presented to the customer, there is no need for a
contract draft review. The task of reviewing the final negotiations results and
the final version of the contract should be left to the legal department and to
management.” Do you agree with the above statement? Justify your answer. (5 marks)

Page 3 of 4
SB/102
QUESTION 3: (30 MARKS)
(a) One can expect that in many cases, participants in an inspection session are
able to suggest solutions for a detected defect or, at least, point out possible
directions for its solution. While it is clear that these suggestions are crucial
for the development team, it is commonly recommended to avoid any
discussion about solutions during the inspection session.
i) Give one arguments in favour of this recommendation. (1 mark)
ii) What other kinds of cooperation between the moderator and the
review team would you prefer to observe in a session? (3 marks)

(b) Not a few software industry professionals maintain that the main goal of
software testing is “to prove that the software package is ready”. Explain in
your own words why this is not a suitable goal for software testing. (2 marks)

(c) By making use of the following tables, calculate the cost unit (cu) for each
phase for:
i) The standard SQA plan. (12 marks)
ii) The Comprehensive SQA plan. (12 marks)
Software development Average % of defects originating in
phase phase
Requirement specification 15%
Design 35%
Unit coding 30%
Integration coding 10%
Documentation 10%
System testing -----
Operation -----

Quality assurance activity Defects removal Defects removal

effectiveness for effectiveness for
standard SQA comprehensive
plan SQA plan
Specification requirement 50% 60%
review
Design inspection ----- 70%
Design review 50% 60%
Code inspection ----- 70%
Unit test 50% 40%
Integration tests 50% 60%
Documentation review 50% 60%
System test 50% 60%
Operation phase detection 100% 100%
***END OF QUESTION PAPER***
Page 4 of 4
SB/102