THE BCS PROFESSIONAL EXAMINATION
Professional Graduate Diploma
MANAGEMENT INFORMATION SYSTEMS
General comments on candidates' performance
There were many candidates that provided answers of a reasonable standard, having prepared themselves for the examination by reading the required material, including past examiners’ reports. Indeed, the best papers were of a very high standard, and those candidates should be congratulated. However, there were a number of candidates that had clearly entered the examination in an unprepared state and provided very poor answers. These candidates either wrote very little or provided long answers unrelated to the questions set: they often (poorly) attempted all questions on the examination paper.
Candidates should be advised that their performance in the examination is likely to be enhanced if, in addition to reading the recommended material, they consider MIS within their own or other organisations. This will put their theoretical knowledge into context and make the whole subject easier to understand and remember. Candidates should also take time to read the questions carefully, so that they fully understand what is required to answer them. An indication is given below of the expected answer points for this examination. However, marks were given for additional points or for valid alternative answers, if relevant to the question.
SECTION A A1.
a) For EACH of the following business analytics tools, describe its key features and how it can be used to aid knowledge discovery. Support your answers with relevant examples.
i) OLAP tool. (8 marks)
ii) Data Mining tool. (9 marks)
b) Explain the advantages and disadvantages of a manager being the direct user of an OLAP tool rather than providing an intermediary to operate the OLAP tool on behalf of the manager. (8 marks)
Part a) of this question is about the types of business analytics functionality found within MIS.
i) An On Line Analytical Processing (OLAP) tool is software that is typically based on a multi-dimensional database that is extracted from an underlying database or data warehouse, and which enables the ad-hoc and flexible
querying of the multi-dimensional data structure (usually referred to as a data cube) by a user in response to business intelligence requirements. OLAP tools allow the slicing and dicing of the data cube in accordance with BI requirements, and also allow multiple levels of view across the dimensions of a cube using drill down and roll up operations.
Candidates may provide examples of OLAP use, and how this can lead to knowledge discovery for particular organisations. For instance, slicing and dicing the customer data in several ways allows knowledge to be gained about individual customers and their sales, or about customer types and their sales.
(5 marks for sound overview of concept, plus 4 marks of examples/discussion regarding knowledge discovery, to a maximum of 8 marks)
ii) A data mining tool enables the semi-automated or automated searching of data, with the intention of discovering hidden patterns or relationships in data, and through this discovering new knowledge about something. Data mining tools can be statistical in nature, based on techniques such as linear
regression. Neural networks, which are also mathematically-based, are also employed in more sophisticated BI analytics operations.
Candidates may provide additional description regarding one or more examples of the techniques one can find within data mining tools, such as regression analysis, neural networks, decision trees, and even qualitative data mining techniques such as content analysis over the Internet/social media. They should provide examples of data mining techniques being applied, resulting in knowledge discovery within a specific subject domain. Some candidates may comment on the fact that OLAP is sometimes classed as a data mining tool, as it allows the discovery of knowledge, albeit user-directed.
(5 marks for sound overview of concept, plus 5 marks of examples/discussion regarding knowledge discovery, to a maximum of 9 marks)
Part b) Is about intermediary versus manager as direct user of an OLAP tool. Advantages of the manager’s direct operation could be:
S/He knows what s/he wants to gain management information about better than anybody else, and there is no need for (and problem with)
communicating with an intermediary working on his/her behalf. S/He can also react to answers gained straightaway with additional analyses if s/he so wishes, which may not be the case when an intermediary is working on his/her behalf.
There is no delay in the receipt of information in this case, whereas there can be with an intermediary working on his/her behalf.
Practically speaking, there may simply no-one available to operate the OLAP tool within the organisation (additional manpower might be able to be bought in, but this would incur additional costs).
Privacy wise, the less people that have access to corporate information (i.e., the foundation of the organisation’s business intelligence), the better!
Disadvantages could be:
The manager may not be proficient in the OLAP tool, nor with IT in general. It would be disastrous to put such a tool in the hands of someone who cannot operate the system effectively, whereas the intermediary should be able to use the OLAP tool, or be able to pick up its functionality by engaging in a suitable training package, given his/her IT background.
The benefits of an additional intermediary and their knowledge of the subject area is lost when there is no need for such a role. The intermediary may be able to provide additional ideas regarding some analyses that the manager requires.
Having the intermediary operate the OLAP tool means the manager would be free to devote time to his/her other duties, which are the things s/he is
recruited to do. This cannot happen if there is no intermediary.
Not all managers want to operate IT systems, and are much happier if some other person performed IT-related activities on their behalf. Again, this cannot happen without an intermediary.
(Marks: Each valid positive and/or negative point to maximum of 2 marks * however many valid points made, to a maximum of 8 marks)
(TOTAL QA1 = 8 + 9 + 8 = 25 marks) Examiners’ Comments
This was the most popular question on the examination paper. There were a few candidates who gave outstanding answers and were awarded full marks for these: however, there were also a few candidates that gained zero marks for the question.
For part a), most candidates were able to describe at least some of the basic features of an OLAP tool. However, several candidates incorrectly considered OLAP to be a data
gathering/ETL tool. Descriptions pertaining to a data mining tool were generally not so good: several candidates gave the same answer as for OLAP, and many answers were vague as to how a data mining tool might be employed to search for hidden patterns in data.
Answers to part b) were often satisfactory, although a number of candidates failed to describe the advantages/disadvantages to sufficient detail for the examiner to understand exactly what was being suggested. Some candidates repeated the same answer twice, by giving the negative of a disadvantage of one of the two user roles as an advantage of the other role: for instance, they cited the ‘potential for miscommunication of information
requirements’ as a disadvantage of the intermediary being the OLAP operator, and then ‘no potential for miscommunication of information requirements’ as an advantage of the
manager being the OLAP operator: in this situation, the issue could only be credited once (i.e., no ‘double counting’ allowed).
a) Describe and justify THREE characteristics of quality management
information. (8 marks)
b) For EACH of the following business situations, briefly describe and justify the most suitable MIS subtype that could be implemented:
(i) A food producing company wants to make different decisions regarding the production process wherever these are required.
Specifically, they want to be able to effectively answer questions such as “if one of the production process machines breaks down, what is the effect on the company’s ability to satisfy customer demand for our products?” and “If we add an additional shift to our daily work
schedule on one of our production process machines, what impact would this have on our throughput per day?”.
(6 marks) (ii) A company wants to provide expert help to its technicians when
diagnosing and repairing faults in complex electronic control systems, without having to commission and wait for the human expert to attend
each time. (6 marks)
(iii) The management of a university wants to report percentage attendances in classes for students on a weekly basis, as well as analysing some of the student data on an ad-hoc basis according to current information requirements. (5 marks) Answer Pointers:
Part a) is about the characteristics of quality information (as opposed to poor quality
information or data). It is expected that candidates will choose to list and justify the following three characteristics, but others may be presented (e.g., concise, complete, up-to-date) and each answer will be treated on merit.
Relevant – Information is defined as a message sent to a recipient that is of value to that recipient, working within his/her current context. Given this definition, information has to have some relevancy to be called information. Relevant means that it is meaningful and of value to the recipient within his/her current context. If it was totally irrelevant, then we would have data, not information.
(up to 3 marks for description of the ‘relevant’ information characteristic)
Timely – Good quality information is timely, i.e., on time (neither too late nor too early), as if it were too late, it may not be able to be utilised within decision making (particularly within immutable decision situations in limited
timeframes), and may therefore lose value to the point where it may become totally useless (i.e., just data).
(up to 3 marks for description of the ‘timely’ information characteristic)
Accurate – good quality information must be accurate, as to be inaccurate would mean that the recipient may rely on information that is misleading and, as a consequence, wrong decisions may be made. Accurate information enables appropriate decisions to be taken.
(up to 3 marks for description of the ‘accurate’ information characteristic) (Total QA2 a) = 3 + 3 + 3, to a combined maximum total of 8 marks)
Part b) is about recognising the most suitable MIS subtype for a particular company decision situation. The “clue” to the appropriate MIS lies, in each case, within the wording of the sub-question. For instance, in Part b) i), the MIS solution could only really be a model-oriented DSS, where the model is a simulation of the production process. There is one large model, against which different input is provided to examine the effects of the changes on the production process (outputs). Different decisions around the production process can
therefore be supported. For instance, the model-oriented DSS could help the food producing company answer the question “if one of the production process machines breaks down, what is the effect on the company’s ability to satisfy customer demand for our products?” by changing the input parameters to the model to signify one of the machines has broken down, and then to examine the impacts on output production against the customer demand. The other question “If we add an additional shift to our daily work schedule on one of our
production process machines, what impact would this have on our throughput per day?” can also be answered by changing the input parameters on the model (to signify a new shift) and analysing the effects on the output produced in a day.
(1 mark for appropriate solution suggested, plus 5 marks for its Description / illustrated use = 6 marks)
Part b) (ii) could only really be an ES (and this would most likely be rule-based in nature). A rule-based ES is a computer-based system that contains knowledge in the form of IF..THEN rules, which allows it to perform decision support operations (e.g., diagnose problems in complex systems) that would otherwise require human expertise. Candidates may continue to describe the components typically found within an ES; the inference engine, the
knowledge base (which contains the rules in the case of a rule-based ES), the working memory (sometimes considered part of the knowledge base) and the user interface. They may also describe the way in which these components work together to provide a
recommended fault diagnosis (as required in this case), using examples of questions regarding the state of the electronic processes in question, and how the system processes the answers given to those questions. The backward chaining and forward chaining methods of inference may be described/illustrated.
(1 mark for appropriate solution suggested, plus 5 marks for its Description / illustrated use = 6 marks)
The most appropriate MIS subtype for the solution to part (b) (iii) is probably an MRS. An MRS is a computer-based system that provides regular and ad-hoc reports to managers at all levels within an organisation. The emphasis here is on regular, fixed structure reports, and their generation and distribution to those that need them. Here, the University wants both fixed reporting and ad-hoc access to pertinent data (via some form of query language such as SQL). There is limited requirement for complex modelling, and the data is
(1 mark for appropriate solution suggested, plus 4 marks for its Description / illustrated use = 5 marks)
(Total QA2 b) = 6 + 6 + 5 = 17 marks) (TOTAL QA2 = 8 + 17 = 25 Marks) Examiners’ comments
This was also a fairly popular question, but unfortunately there were many that did not understand the requirements of the question, most notably within part b).
Answers to part a) were generally satisfactory when the candidate understood the question to be about the quality of management information; not the quality of an MIS or quality management. Too many candidates read the question to be about the characteristics of a quality MIS, and thus the focus of their answers was on technological matters (such as data security and integration): these answers gained little, if any, marks. Some candidates
provided relevant answers but these were too brief to gain more than a few marks overall. Some candidates described more than three characteristics: in this case, the best three were taken to count towards the final mark for the question.
Answers to part b) were disappointing overall. Many answers recommended an MIS subtype and gave an overview of the theoretical concept, yet were unable to describe how it would be applied to the stated business situation. Answers to b) i) often correctly identified that some form of DSS was required, but did not proceed to propose the eminently suitable simulation model, instead suggesting the use of two separate DSS with very limited and inflexible models (each DSS set up to specifically support the answering of one of the example “what if” requirements in the question). Some candidates felt a GDSS would be the most suitable answer to part b) ii), but failed to realise that this would mean that one or more human experts would still need to be involved in any GDSS-based decision making. A GDSS was also recommended by some candidates as the most suitable MIS subtype for part b) iii), which was surprising given the clear fixed and ad-hoc reporting requirement within the stated business situation.
A3. Explain the meaning of EACH of the following statements regarding effective enterprise-wide BI systems developments:
a) Development of a BI system is better approached by aiming big, and starting small.
(7 marks) b) BI systems development must be business-driven.
(7 marks) c) Scalability is a key issue to consider within BI systems development.
(5 marks) d) The number of data definitions to be added to the BI system data
repository is a better determinant of how long a BI system
development project will take than the number of reporting functions to be implemented.
This question is about BI systems, but covers a wide range of issues concerning their development; strategic alignment, overall philosophy and nature of developments, and hardware/software performance issues (scalability). The question intentionally focuses on larger BI systems that span across the organisation rather than small, more departmental-focussed systems.
Answers to each part may include the following aspects:
a) Development of a BI system is better approached by aiming big, and starting small. This is about seeing a BI systems development project as part of a bigger picture of BI deployment within the organisation, and not as a standalone, one off development. If the latter view was taken, this would lead to BI systems that are incompatible with one another (separate silos of data, with incompatible data definitions and associated ETL processes) and which do not benefit from the integration of data and the knowledge discovery potential therein. However, the statement also recognises that a large BI system environment cannot be built all in one development project, for several reasons (management interest over time can diminish, no
recognisable benefits for potentially several years despite heavy costs, the sheer complexity of such as system even if its functionality can be determined
at the outset, etc.). So, seeing the development of the BI system provision as a series of small projects, but maintaining an integrated view of the provision is the answer.
(For a sound justification and understanding of the statement, 1 mark per salient point to a maximum of 7 marks)
b) BI systems development must be business-driven: This is true of all IS development projects; that they need to be driven by the business and not by the IT department or by the fact that a competitor is doing these
developments. This driven by business has several parts to it;
1. The need must be linked to business strategy (either an opportunity for strategic advantage, or by a threat that makes the BI system a strategic necessity) and existing IS cannot satisfy either a new or indeed an existing strategic requirement.
2. Support is needed for the development by top management, as without this support, the financial investment, the sponsorship of the project across the organisation and the help during problematic project times, will not be there. A project that is initiated without such support is very likely to fail. (This support is aided by the strategic alignment described in point 1 above.)
3. If a business need is clear, all employees are more likely to see the benefits of the proposed BI system, and less likely to resist.
(For a sound justification and understanding of the statement, 1 mark per salient point to a maximum of 7 marks)
c) Scalability is a key issue to consider within BI systems development:. This is a more specific question, where candidates need to understand scalability (to do with performance not degrading – can be associated with both hardware and software (e.g., the DBMS within which the BI system data repository is housed)). Essentially, BI system performance should not degrade with the increase in aspects such as the amount of data held within the BI data repository, the number of users accessing reports and the number of reports and/or software tools access the data in the repository DBMS. Candidates should recognise that over time, BI systems can become extremely large VLDBs (data repositories housing terabytes of data) and that scalability under these circumstances can become a really big issue for organisations to achieve (need to consider parallel processing, partitioning of data, etc.).
(For a sound justification and understanding of the statement, 1 mark per salient point to a maximum of 5 marks)
d) The number of data definitions to be added to the BI system data
repository is a better determinant of how long a BI system development project will take than the number of reporting functions to be implemented. This statement is reflecting the fact that the majority of the complexity in the development work on a BI system is to do with the data that forms the basis of the BI system data repository. The development of reports off a given data repository is typically relatively straightforward to achieve (anything complex like an AI-based analytics capability will typically be developed via a suitable
business analytics software package rather than developing the software from scratch). It is the preparing and loading of the data into the data repository that is typically the lengthier process – ETL has to be designed and programmed, which may involve substantial (lengthy) cleansing and transformation of source data sets. The more data items defined to add to the data repository, the more items will have to be sourced and hence more ETL programs written/revised. Data requirements are therefore a far better indicator of project timescales than the reporting function development requirements.
(For a sound justification and understanding of the statement, 1 mark per salient point to a maximum of 6 marks)
(TOTAL QA3 = 7 + 7 + 5 + 6 = 25 marks) Examiners’ comments
This question was the least popular on the paper, and overall performance on the question was very poor. There were very few candidates that had at least a basic understanding of all aspects covered by the sub-questions.
Several candidates thought part a) was either all about IT strategic alignment or prototyping, and therefore provided lengthy theoretical descriptions of the chosen concept which were largely irrelevant. Most answers to part b) articulated the fact that BI proposals must come from the needs of business, but did not really explain what it means to be “business driven” and why this is important. Many candidates did not understand ‘scalability’ and therefore provided either blank or totally incorrect answers to part c). Answers to part d) were disappointingly poor, with very few candidates noting the complexity and hence time involved in implementing effective ETL operations within their argument. Indeed, several candidates tried to argue the complete opposite of what was required by the question, i.e., that the quantity of report functions are better determinants of project timescales than the quantity of data definitions!!
You are the Systems Development Manager for a large organisation.
A friend has emailed you for help and advice regarding her new role as lead user representative for an Enterprise System. This system is of strategic importance for her organisation, as it has failed to make profits for the last two years.
Prepare an email to your friend which:
a) Describes the fundamental purpose AND likely duties of a user representative.
(10 marks) b) Recommends and justifies THREE areas of Management Information
Systems knowledge that would be useful in her new role.
(15 marks) Answer pointers
a) Fundamental purpose and duties
There are many variations of answer. Some points that could be included are:
Because a user representative is the link between the developers and the organisation, she will have to make decisions on behalf of the organisation or be the conduit for
A user representative should have an in-depth knowledge of the business, which can be passed on to the developers.
As she will need to formally sign-off various stages of the development, she needs to have the appropriate authority to do this on behalf on the organisation.
She will need to resource and organise other users for testing/parallel run activities. The user representative may also need to control and administer the development budget and approve expenditure for any changes.
(One/two marks per point, total 10 marks) b) Three areas of knowledge (for expansion)
There are many different areas of knowledge, including:
Knowledge of the development methodology used. This is essential, as the user representative will be a major player and needs to understand all aspects of the role. For example, if a linear method is used, the representative will need to know when the
documentation from each stage is available to ensure sufficient time is allowed for
agreement/sign-off. If an iterative method is used, the user representative will need to be available on a regular basis and would not be available for other activities.
Knowledge of data and process modelling. The user representative is likely to be presented with models to understand or might be expected to develop models herself. Project management methodologies. Often, a user representative is expected to be the project leader or project sponsor. Knowledge of the appropriate methodology, PRINCE 2 or similar, would be beneficial.
(Five marks per area of knowledge, including justification, total 15 marks) (Total marks = 10 + 15 = 25 marks)
For part a) many candidates concentrated on user representatives’ personal qualities and gained no marks for these. Other candidates provided answers that were too brief or failed to provide the required description. The better candidates clearly considered the activities of user representatives, possibly drawing from their own experiences, and provided answers that mirrored the developer’s activities and areas where collaboration with users is essential. Part b) gave the candidates considerable scope to demonstrate their knowledge of important areas of MIS that would be of benefit to any user. Unfortunately, many candidates provided answers of insufficient depth or which showed a lack of knowledge or understanding. B5.
You have been recruited as the Management Information Systems manager for a life insurance company where the Chief Executive Officer (CEO) has two concerns: first, the capabilities of the company's policy administration legacy system need to be improved and advice is needed regarding the range of options available; secondly, since the introduction of a new network, there has been a steady growth in the number of end user systems and the CEO feels that some guidance is essential.
Write a memorandum to the CEO to:
a) Describe SIX possible options the company has for improving the capabilities of its legacy systems. Include in your answer the advantages and
disadvantages of each option.
(15 marks) b) Discuss the prerequisites of successful end-user computing and comment on
how the success of such systems could be measured.
a) Options for improving legacy systems
Restructure (or equivalent wording). Remove redundant code and develop and/or simplify interfaces with other systems.
Advantages: inexpensive, retain functionality; risk averse in the short term; change is minimised.
Disadvantages: short term fix, fixed structures of legacy systems are retained.
Re-engineer (or equivalent wording). Not BPR, but placing legacy system on a new hardware/software platform whilst retaining existing business functions.
Advantages: removes threat of hardware/software unavailability; business processes unchanged.
Disadvantages: business processes unchanged; fixed structures of legacy systems are retained.
Refurbish(or equivalent wording). If legacy system is maintainable, add extensions such as a web front end or provide extracts for EUC.
Advantages: can give impression of changed business processes without underlying changes; inexpensive.
Disadvantages: fixed structures of legacy systems are retained.
Rejuvenate (or equivalent wording). Transfer data to a modern DBMS, re-engineer with BPR, include data warehousing/mining.
Advantages: provides for the medium term; releases organisation from the fixed structures of legacy systems.
Disadvantages: can be almost as expensive as a new system; user time required for testing.
Replace with a package(or equivalent wording). Use a package with the same functionality as the legacy system.
Advantages: the advantages of using a package - preset business solution; support and maintenance, etc.
Disadvantages: the disadvantages of using a package - need to modify; limits individual business processes; vulnerability to supplier, etc.
Replace with a new bespoke system(or equivalent wording). Write a new system for present and future.
Advantages: opportunity for complete strategic review; use of modern technology. Disadvantages: high cost; high risk.
(Two marks for each option identified and described, maximum 12 marks, plus one mark for presenting the answer in memorandum format, plus two marks for quality of argument and approach, total 15 marks)
Part b) – End User Computing
Five expected prerequisites are listed below.
Agreed boundaries between end-user and corporate MIS computing/any agreed general requirements.
Use of common software/desktop setup. The need for the whole organisation to have a common computing platform.
Shared resources. The need for shared programs, data, files, standards, naming conventions, networks, etc.
Data ownership and availability to be agreed by all participants. This should include a possible need for an initial data analysis to include the identification of both standing and transactional data.
(One mark for each prerequisite, maximum five marks) Five measures of the success of end-user MIS
User, management and customer surveys Degree of data duplication
Consistency and timeliness of data
Productivity improvements (Business and MIS) Costs versus benefits measurements
(One mark for each measurement, maximum five marks) (Prerequisites + measurements 5 + 5 = 10 marks) (Total marks = 15 + 10 = 25 marks)
This was a popular question but there was a tendency for many candidates to write pages of unnecessary introduction. Although no marks were lost, this approach meant they lost time and many then subsequently failed to answer the second part of b) (see comments below).
Part a) The “six Rs” answer to this question can be found in MIS textbooks and those candidates who knew the various alternatives gained very high marks. Regrettably, a significant number of candidates wrote about development methodologies or project stages and no marks could be given.
Part b) This was another opportunity for candidates to display their textbook knowledge and some high marks were gained. However, a large number failed to suggest any way in which the success of end-user computing could be measured and therefore missed out on at least some of the available five marks.