BCS Higher Education Qualification

Diploma

October 2022

EXAMINERS’ REPORT

Object Oriented Programming

General comments1

Questions Report:

A1 Syllabus Content: 1.2; 2.3

This question was answered by over 50% of candidates, but had a low pass rate.
Candidates were able to score good marks on Part A, where they were able to show
they understood what protected visibility in a class meant. Weaker answers
described the three types of visibility, but then failed to give an example of where
protected could be used in a real-world scenario. Most marks were lost where
answers implied that protected meant information security, which was not
appropriate in this context.
Part B was often not attempted, or the answers lacked sufficient detail to score a
high mark. A number of candidates provided general constructors only, which
showed no evidence of any conversion. Some examples showed type conversion, few
provided answers that converted an object passed as a parameter
to the type of the conversion constructor’s class. A small minority of candidates
mixed up conversion constructors with overloading.
A2 Syllabus Content: 2.2, 2.3

This question was answers by over 64% of candidates and most candidates scored
well in part A. Better answers compared the advantages and disadvantages of manual
and automatic memory management, with relevant examples given from object-
oriented programming. Weaker answers listed a set of bullet points without going
into any depth, such as stating a disadvantage was “high cost”, without explaining
what they meant by this, which was needed for a higher mark. The advantages given
were often just the opposite to the disadvantage, such as “low cost”.
A substantial number of candidates did not attempt Part B of the question. The more
able candidates were able to provide code to illustrate hybrid inheritance, often
including a class diagram to explain how the classes demonstrated this. Weaker
answers mixed up hybrid inheritance with multiple inheritance and some included
code that only demonstrated one type, typically single inheritance. Some candidates
struggled to suggest a real-world example, though credit was given if they could
outline what hybrid inheritance would look like using Class A, B, C type examples.
A3 Syllabus Content: 3.4; 4.2

This question was attempted by over 50% of the candidates and most scored good
marks on Part A of this question. The majority of candidates were able to describe
what a singleton class was, though some did not get beyond stating it was a single
instance of a class. To obtain a high mark, an appropriate real-world example was
needed, which some candidates were not able to provide.
In Part B, most candidates could explain what an is-a inter-class relationship was and
provided appropriate code to demonstrate this. Less able candidates were unable to
provide a suitable has-a inter-class relationship, or mixed the two types up. A good
answer needed to clearly show inheritance for an is-a relationship and composition
for a has-a relationship, sometimes, the answer showed two classes, but not any
code to show how the inter-class relationship would work.
B4 Syllabus Content: 3.2; 4.4

Question B4 was attempted by almost 80% of candidates, making it the most popular
question on this year’s paper. The quality of answers varied enormously.
In part (a) the most common mistake was not concentrating on structural constraints,
but (in extreme cases) simply listing all of the class names, along with the operations
and variables they contain. Many of the answers were excessively long and
unfocussed. Many candidates did correctly identify the multiplicity of the inter-class
relationships, although other structural constraints were quite vaguely or confusingly
expressed. In part (b), many candidates simply listed, with brief explanations, one or
two testing methods (such as black box, white box, or unit testing), ignoring the part
of the question that asked that the tests could be used on the class diagram from part
(a). Some marks were awarded if some testing methods were accurately described.
Higher marks were only given if the candidate linked the testing method to object
oriented programming and the class diagram specifically.
B5 Syllabus Content: 3.2; 3.4

This question was attempted by approaching 65% of candidates, but was one of the
two least well answered questions on this year’s paper.

Part (a) asked that candidates identify an example design pattern in one of three
categories (creational, structural, behavioural), and then go on to describe it, including
the problem it addresses and the basis of its offered. In many cases, candidates
proposed design patterns that did not match the categories (e.g., an iterator being
identified as creational). Other issues included not providing any description of the
design patterns identified, or not clearly stating the problem they are intended to
address. In part (b), candidates were asked about the use of object interaction and
object state transition diagrams. Example diagrams were not requested, but most
candidates attempting this question provided them, perhaps to help answer the part
of the question asking for a real-world scenario. It was a common mistake to
provide/describe an object diagram instead of an object interaction diagram. In some
cases, the proposed states in the state transition diagram examples were not actually
states (for instance, some conveyed actions rather than states).
B6 Syllabus Content: 2.3, 3.3; 4.4

Question B6 was the second most popular question on this year’s paper, with nearly
80% of candidates attempting it. Approximately half of those attempting the question
gave a good answer.
Part (a) asked about the key elements of object constraint language (OCL), for which
most candidates were able to speak about pre and post conditions and context (with
some also mentioning invariants and guards), and the role of OCL to formalise aspects
of a UML diagram that are inherently lacking in detail. In part (b), most candidates
seemed to largely follow the OCL statement, but descriptions were rarely completely
clear, and several misinterpreted the statement (e.g., believing that the mark
argument was added to the grade variable rather than replacing it, or that the mark
had to be >100). Part (c) was generally quite poorly answered, with most candidates
rephrasing the same point over and over again rather than truly identifying distinct and
meaningful advantages/disadvantages of getters/setters.