Editorial Reviews. About the Author. Dr. Alan R. Dennis is a professor of internet systems at Systems Analysis and Design with UML, 3rd Edition 3rd Edition, Kindle Edition. by Alan Dennis (Author). Developers and students will learn object-oriented analysis and design. (OOA/D) through three Applying UML and Patterns, Third Edition, is a lucid and practical introduction to thinking and designing with objectsand creating systems that are well The final print images were generated as PDF using. Adobe Acrobat. Systems analysis design, UML version an object oriented approach/Alan .. As in the third edition, the material included within the analysis modeling.
|Language:||English, Spanish, German|
|Genre:||Business & Career|
|ePub File Size:||22.71 MB|
|PDF File Size:||8.50 MB|
|Distribution:||Free* [*Regsitration Required]|
System Analysis Design. UML Version An Object-Oriented Approach. Third Edition. Alan Dennis. Indiana University. Barbara Haley Wixom. University of. Systems analysis & design: an object-oriented approach with UML/Alan .. ful comments on the first, second, third, and fourth editions: Evans Adams, Fort. to a third party. Systems analysis and design /Alan Dennis, Barbara Haley Wixom, Roberta M. Roth.–5th ed. p. cm. through UML , the new standard for object-oriented analysts and design, as .. 2 subiecte.info /iag business analysis benchmark - full subiecte.info; accessed February,.
Real-time control systems Real-time systems are explicitly concerned with the direct control of a system's operations, often physical in nature. For example, the web page used by this customer to order her new refrigerator is her interface with the System A student A family A business Inputs Information Exercises Guidance Money Social standards and norms e. How does it differ from data? Whatever the reason, it's always your fault, even though all you ever tried to do was to give them what they wanted. They also object.
This aims at capturing in one picture or diagram all that is essential to an initial understanding of a system 4. An opposing view is taken by reductionists, who begin with the assumption that See Checkland and Sc:: TIus has a very Imp0r: We believe it is thoroughly misapplied when used to explain complicated human situations.
General is to both, but in this book we are principally interested In the kind. Even some of these are more relevant than others. For example, we will only be interested in system of parlia- mentary democracy If we Intend to develop an Information system that supports it. Two types of system are of particular importance to us.
The first is the human activity system. A key feature of this type is that it centres on a purposeful as we when we look at a business, a club, a hospital, or any orgamzed activity.
There IS often wide disagreement among the participants about what purpose is, and this can be a significant problem for the analysIs of Information system requirements. Information systems are constructed to help people in a human system to achieve their goals, whatever these might happen to be. Human are the context of, and provide the meaning for, the activity of mformatlon systems Anyone who does not understand the meaning and of a system will find that it is impossible to specify, still to build, an mformatlon system that supports it in any useful way.
Each of these information systems is intended, in some way, to help fulfil the purposes of a human activity system. The of systems development can be usefully regarded as a human activ- I! This system transforms various inputs money, skill, staff time, infonna- lion from users how they want the software to work, etc.
Its environment is System Agate a busines. As seen from the perspective of A directOf A copy-writer Another director 1.
Its subsystems include the members of the project team, the methodology they follow and also the various analysis and design models that describe the software. Control is exercised by a team leader or project manager, who receives regular feedback on progress and problems. Suitable feed- forward will help to alert the manager to anticipated problems, as well as those that have already happened.
One advantage of taking a systems view of any activ- ity is that it encourages those involved to think about the sorts of feedback and control that are needed for everything to run smoothly. This applies just as much to software development as to anything else. Inforrnatio" Some computing professionals see the development of information systems as essentially a matter of designing and building computer technology including software that meets a set of clearly understood needs.
While this may be the ideal situation, in practice it is often a simplistic view that misses much of importance. It happens that information technology is now the nonnal technology used to implement an information system, but, as we saw in Section 1. Designing and building technology can be the easy part of the job-at least the easy part to understand-while the hard part is often determining the needs that the technology must serve.
This involves identifying ways that an information system can support the purposes of a human activity system. An understanding of the information that wi ll be useful to the human actors is an important ingredient in this, as is an understanding of how the information can be used effectively. As a result of these many diHerent concerns, Infonnation Systems has become a multi- disciplinlll ' subject that bridges many other fields, in particular computer science and business management, but also psychology, social theory, philosophy and linguistics, among others.
In the following sections, we discuss the relationship between infonnation, infonnation systems and the human activity systems they are intended to assist. Meaning always depends on the perspective of the person who receives a message. We are always surrounded by a vast mass of potential infonnation, but only some of this ever comes to our attention, and only some of that is actually meaningful in our present context.
Checkland and Holwell describe the process by which raw facts become useful i. For an example of how this happens, consider four people watching the evening sky. A plume of smoke is rising in the middle distance. For Alex, the smoke is just part of the view, and she does not even consciously notice it.
Ben sees it, and it evokes a memory of a camping trip long ago. But he is aware that the only connection between past and present smoke is a coincidence of shape and colour, so he moves on to look at something else. Dipti runs to phone the Fire Service before doing whatever else she can to help Chetan save his hOllse.
The sight of the smoke is, on the face of it, a single message available to be received by all, yet its meaning is different in each case. Alex does not even notice it, and Checkland and Holwell call this dala from the Latin for 'given' , meaning a fact that has not been selected for any special attention.
Ben notices the smoke but docs not relate it to any present context. Checkland and Holwell call this capla from the Latin for 'taken' , meaning a fact that has been selected but has no particular importance or meaning. This can be called illformalioll since it has a meaning within the context. The meaning of a fact is always dependent on its relevance to the observer.
We can see this in Dipti 's response to the smoke. She integrates information from many sources: All this information comes together in a single framework of knowledge that is relevant to the contexl In a word, she knows what to do.
For example, Laudon and Laudon describe the main types of information system in business as being: Such a classification is probably less useful today than it once was, since information systems have become more closely integrated with each other, and the boundaries between categories have become blurred.
It is still helpful to present a brief overview of some of the general types of application in organizations, but this really describes the roles that information systems can play, rather than actual types of system. The earliest commercial information systems were operational ones, because routine, repetitive tasks involve little judgement in their execution, and are thus the easiest to automate.
Accounting systems are a good example. Few modern organizations could survive long without a computerized accounting system. Sensible organizations have a 'disaster recovery plan' that details how they intend to cope with an emergency that destroys data or renders computer systems inoperable.
The flow of information through an accounting system is based on thousands, or even millions, of similar transactions, each of which represents an exchange of a quantity of something, usually a money value this is why they are often called 1. For example, when you buy a tube of toothpaste. One records that a tube of tooth- was sold, and the other records the money you paid in exchange for it. As this repeats day after day for each item, customer, checkout and branch, an. Of course, many w'al accounting systems are more compli cated than this, often with subsystems to re.
Other operational systems record orders received from customers, the number of items in stock, orders placed with suppliers, the number of worked by employees, time and cost of mobile telephone calls made by subscnbers, and so on. Management support systems Information systems intended to support management usually work at a much higher level of complexity than This is. However, much of the information used by management to make decisions is derived directly from information stored at the operational level.
Other systems combme elements of the two, meeting a complex set of needs at different levels of the organization. Many of the earliest management support systems were developed Simply by adding a set of programs known as a management information MIS. IO extract data from existing operational systems, and analyse or combme It to give managers information about the part of the organization for which they were responsible. We can easil y sec how this happened with an accounting system.
Once all routine sales transactions were stored on a computer, it was a short step to the realization that if this data were analysed appropriately, it could tell managers at a glance which products were not selling well, which checkout operators took too long dealing with a customer, which store had the lowest volume of trade, and so on. This information is useful to managers because they have a responsibility to maximize the performance of an organizational subsystem.
An important part of this is identifying and resolving problems as they occur. Thus, one cruci. Let us return for a moment to the diagram in Fig. Operational systems either are located in the central box labelled 'what the system does' or, allernatively, they assist its work by supporti ng the flow of inputs or outputs.
Management support systems either arc located in the box in the lower part of the diagram labelled ' how the system is controlled' or, alternatively, they assist its work by supporting the flow of feedback to, or control information from, the control unit. Offiu systrms Office systems automate or assist in the work of office workers, such as clerks, secretaries, typists and receptionists. They also support some aspects of the work of managers, for example, communication word-processing, email, etc.
This might suggest that they are a kind of management support system, except that they are used today by almost every kind of employee, not just by managers. This illustrates the extent to which boundaries between different types of system have become blurred, in part due to the spread of information technology through the modem organization.
It also highlights the fact that the introduction of information systems often changes the way that people work-many middle-ranking and senior staff now type their own letters and reports using a word processor, when once they might have expected a secretary or a typist to do this work for them.
Real-time control systems Real-time systems are explicitly concerned with the direct control of a system's operations, often physical in nature. Examples include lift control systems, aircraft guidance systems, manufacturing systems and the robot forklifts in the McGregor system described in Box 1. For this reason, they are perhaps best considered as a control subsystem of a physical processing system.
Their role is thus very different from both operational and management support systems. Real-time systems usually have human operators to date, few are completely independent of human supervision, though this may become common in the future , but they are generally insulated from the surrounding human activity system. In fact, many authors would not agree that real-time systems are information systems at all We do not regard this as an important issue.
The techniques used for the analysis, design and implementation of real-time systems are very similar to those used for other computer systems, so in practical terms any distinction is artificial. When the human activity system has been understood, the need for an information system has been identified, and the information system requirements have been defined-only then should the emphasis tum to the information technology that will implement it.
This is not how things always happen in the real world. Indeed, it is partly for this reason that so many systems in the past have been unsuccessful. But it is how they should happen, wherever possible.
Section 1. Here we offer only a broad definition of information technology, and this only for the sake of clarity. We understand it to include all the varieties of hardware familiarly known as, based upon or that include within them, a computer or its peripheral devices. For example, obvious things like desktop Pes, pocket elec- tronic organizers, modems, network cabling, file servers, printers and computer- controlled machinery in factories and airliners, and also less obvious things like digital mobile phones, the electronic circuits that calculate fuel consumption in 1.
Due to the unprecedented rate of technical progress, the range of devices that can be described as information technology increases almost daily, and the boundaries between them blur. As digital devices continue to advance in speed and processing power, manufacturers exploit these advances to develop and market new products. For example, consider the mobile phones that combine a digital camera, modem, email software, web browser, diary software, alarm clock, calculator and online gaming.
Interface technologies such as voice-activation may soon make it much easier to interact with computers without needing to press keys or click mouse buttons, while wireless networking makes physical connection unnecessary. Mobile commerce using handheld PDA devices, mobile phones and wireless headsets is changing the way that many people access information and communicate.
For many, it has already removed the physical restriction that requires a user to be in the same place as a bulky PC when he or she wants to access the Internet. On the whole, it appears likely that over the next few years computers will progressively disappear from view, while their effects wi U paradoxically be felt in more and more areas of everyday life.
All the examples of information technology that have just been mentioned are real1y just tools that, like any tool, can be used for many different tasks-and not only those for which they were intended. There is a saying that, if your only tool is a hammer, the whole world looks like a nail. The corollary is also true: So how a tool will actually be used matters much more than how it is meant to be used.
A modem word-processing package provides a skilled user with the facilities to auto- mate many complex tasks by using macro programs, mailing lists and embedded objects like spreadsheets, sound clips and hyperlinks to the Web. Yet many users have no need of aU this and would be just as happy with an electronic typewriter.
The question is, then, if an electronic typewriter is al1 that is required, why install a powerful PC running all the latest software? The underlying assumption is that information systems are only worthwhile if they meet the needs of the organization in which they are installed. In this section, we consider some ways that business needs can be iden- tified at a very high level, suggesting possible application areas for information systems and information technology.
The development of a business strategy essentially begins with the question: In Chapter 12, we will return to this point from the perspective of managing information for its strategic value, both in private and in public-sector organizations. The contents of a strategy i. In the Agate case study, we are told that the strategy is 'to continue to grow slowly and to develop an international market' these are the goals.
The directors also have a view about how to achieve these objectives: These elements have been included in the strategy because the directors are confident, for example, that the technical quality of Agate'S work and the creativity of their staff are both strengths of the company, and wi ll meet the demands placed on them. They probably also believe that their current client base and contact list is cxtensive enough for them to win the kind of work they arc seeking. Many techniques have been developed to help arrive at useful answers, some adapted from their original purpose as tools for developing a busi- ness strategy.
One example is the well-known SWOT approach s. Usually carried but as a group brainstorming exercise, it involves identifying and categorizing everything about the organization's current circumstances that falls into these categories. The resulting strategy is based on finding ways of exploiting the strengths and opportunities, while counteracting the weaknesses and threats.
Another is Porters Value Chain Model , which we consider here. While not the only technique for information system planning, and not necessarily the best for all situations, it is a useful way of structuri ng this discussion because of the systemic view of an organization that it presents. The model is illustrated in Fig. The flow is transformed at each stage, begin- ning with raw inputs that become a product through operations applied to them, then are shipped and sold to a customer who finally receives an after-sales service.
To apply the model, a group of managers and other staff identify the activities and departments that fit the compartments of the model, and identify the value that each adds to the overall product or service. Porter used the chain metaphor to reinforce the point that all primary activities are essential, but any weak link negates the value of work done at every other stage. For example, if a business is good at selling its products, but the products S This stands for Strengths. In a success orgamzatIon, eac I.
They are not: Their role must therefore be tuned to support primary activities. The Value Chain Model can be useful m mfonnatton systems plannmg smce It focuses attention on activities that are critical to a business, either because they are currently a problem, or because they represent a major source of profit or com- petitive edge.
Information systems development projects can then be targeted at assisting those operations that can directly contribute to the success of the organization as a whole. In practice, the basic fonn of the model is often adapted to fit better Wlth the activities of the organization that is being analysed. Note that this makes no distinction between primary value creating and secondary support activities. This means that everything done by Agate's staff contributes to the perceived overall quality of service.
For this reason, all activities potentially add to or diminish the value on offer to its clients. V directly add volue Primary OCfNitic! Ideal ly, value is added at each stage. V Figure 1. Recruit Clients Project Setup Figure 1. Information flows In the diagram are not only in one direction.
In formulating a business stra! Thus the planning cycle IS Iterative. A Slmdar two-way communication takes place between the information information technology strategy planning functions.
The role of the information technology strategy is to enable the successful implementation of syste! The importance of the relationships shown in Fig. For e. One objec- tIVe. If other online retailers aJready taken a significant share of the overall market, they may pose a threat to McGregor's continued survival.
In order to fulfil this business McGregor's managers must identify, define and then develop an appro- pnate set of software systems-this includes the online customer order system the robot warehouse system, the stock control system, the purchase order system,'and so on. Those chosen for development will then become tnteace. Selecbon of the wrong systems or eve opmen systems.
I h' I mean systems that are not compatible with other Vltal systems. IS ouse ,. IS not su IClen y n of the need for the software it is developing to interface Wlth another system. I I" Another critical issue might be poor design of the onhne cata ogue, resu tmg m frustrated customers who buy instead from a competitor's website.. The IT strategy is responsibl e for identifying the hardware components and con- figurations that will allow the software to operate effectively.
In this will include specifying the web servers to ensure that the response bme IS always fast enough to satisfy customers. Slow response frustrate customers and result in lost saJes for the company. WlLl not be worked out until later, but the strategy will Identify this as a busmess concern, and will also explicitly relate the various hardware components to the software systems they must run. For many businesses, success depends on finding an appropriate fit between the overall business goals, the information systems that help to fulfil those and the IT on which the infonnation systems run.
This question of strategy ahgnment has even greater importance when the business is in whether this is business-to-customer B2C or business-to-busmess B2B. In either case, for customers, suppliers, partners, collaborators-indeed for any interaction that occurs electronically-the information systems are the company, since the website is really all that can be seen. Moreover, an Internet presence can be seen-and judged-by the whole world.
Either an inappropriate strategy, at any of the three levels, or a poor implementation can bring swift business failure, as a of dot.
In Chapter 2, we explore in more detail the problems that can occur dunng information systems development, and even because of information systems, while in Chapter 3 we introduce some ways that these problems can be resolved.
These include some important ideas rom general systems theory, such as the ideas of control, communication and emergent We have discussed the relationship between information and meamng, the central pomt here being that infonnation only makes any sense within a specific context.
This leads to a necessary set of relationships between the goals or an organization, the steps it must take in order to fu lfil them, the information that its staff must have to carry out their activities, the infonnation systems that provide the information, and finally the information technology that supports the infor- mation systems.
While we recognize that there are many real organizations where this does not happen in the way we describe, we can only wonder how much more could be achieved, and express the hope that readers of this book will start work from firmer foundations.
The idea of information systems is far from new, yet information technology has changed the field almost beyond recognition in the last few years. Review Questions S What is the difference between feedback and feed-forward? How does it differ from data? What is its purpose? A Think of three or four information systems that are not computerized either historical or contemporary. Identify or imagine a computerized equivalent.
For each pair, write a brief description of the boundary and the main inputs and outputs. What are the main differences between the computerized and non-computerized versions? What feedback and system. Identify do ou they would use? Don't be constrained by the desCflp- forWard information y 'magination too. And remember that some control may not tion in Box 1.
What oyou I. Make any. Which do you think are the wea Further Reading They also object. Some authors use the metaphor of a Journey ,to? At the start of a project, many different routes are. Some to the planned destination, while others lead to a desti- nation may surpnse everyone by turning out to be satisfactory, but is never- theless unmtended.
Both are acceptable, although the latter can be nerve-wracking along the way. There are also routes that arrive at the wrong destination, while? We must recognize and avoid these. This IS the exphclt of canying out analysis and design before developing a new sys. IYSls IS a of understanding what needs doing before we begin trying to do It, and deSign IS a way of checking that the planned action or solution really.
An under- of potential problems is an essential precursor to systems analysis and deSign. In this chapter, we look at of project failure from the perspective of of the main players S n 2. We discuss the causes that can lead to fai lure. These are broadly: While not usually a direct re a e a1'd tho I'. Some of these may act as a del. Finally, we mention the costs of the problems, whether or not they result in outright system failure Section 2. But the consequences of failure can be severe. One series of reports on information systems projects in the USA estimates that failed projects accounted for S81bn annually in almost a third of total information systems development expenditure Standish Group, , Many UK public-sector contracts for information systems have been cancelled in recent years or have failed to deliver their intended benefits, although the true costs are often not known with any certainty.
Often poor project management and lack of fi nancial control are blamed for the disasters OGC, But all potential causes of failure are at least to some extent under the control of the developers. A profes- sional must take the possibility of failure seriously, and work hard to avoid it, even if this sometimes is limited to an awareness of the risks, followed by a damage- reduction exercise.
One difficulty with the question What can go wrong? But this is not surprising, as infonnation systems development is a complex activity that always involves people. This is not to say that aU people are difficult-this only applies to a very few! But, in any organiz- ation, people have varying perspectives that influence their view of a situation, and what, if anything, is to be done about it. Although some differences in attitude depend on psychology or personal history, which are beyond the scope of this book, others relate to an individual's position within the organization and their link to an information system project.
It is useful to discriminate between three categories of people with important relationships to a project. The second is the group of managers, here called 'cl ients', who have control or at least influence over the initiation, direction or progress of a project.
The third is the group of professionals responsible for the development of the information system. For the moment, we ignore differences within each group. The examples that follow concentrate on the experiences and frustrations of those who either consume the products of an information system i. I haven't seen a new system' One problem that can be experienced by an end-user is vividly expressed by a term that gained widespread usage in the s.
Vapourware describes a software product that is much talked about, but never released to its intended users. In other words, instead of arriving, it evaporates. While many businesses are reluctant to talk about information system project failures in public, vapourware may be surprisingly common. Some surveys have found that an astounding proportion of information systems development projects fail to deliver any product at the end.
This last example indirectJy hurt patients, as the aim of the system was to help manage hospital resources more effectively throughout the region, and thereby deliver a better, more responsive service.
When a project is not completed, the expected benefits to users and other beneficiaries are not achieved. Systems may fail to meet the criterion of usability in a number of ways, including: Figure 2. Not long ago one of the authors bought a pair of shoes in a local shoe shop, and saw staff struggling to register the sale correctly with a new cash register system. The difficulty arose indirectly from a promotional offer. Any customer buying this style of shoe was also entitled to a free pair of socks.
Since the socks were a normal stock item, correct records of stock had to be maintained. This meant that the socks had to be 'sold' through the till, even though they were actually being given away for free. A simple way to handle this would have been for the assistant to over-ride the price with zero at the time of sale. The assistant tried this. This took them away from patients for longer I A national motor vehicle insurance company lost most of its digital records of customers' policies due to a system error.
Staff were unable to send renewal notices, but were compelled instead to write to customers asking them to phone in with their policy details 2 Figure 2. The assistant called the manager. After some it appeared that the onJy way to deal with this transaction was to reduce the pnce of the shoes by 1 p, and to sell the socks at a cost of 1 p, thus giving the total. But It will always be an unnecessarily awkward way of handling a routine task.
There many examples of this sort of poor design, and they cause a great deal of frustration and lost time for users. This may be a question of the tasks that should be by. For example, a library catalogue enquiry system would be use If could only retrieve shelving information about a book when prOVided With the title the author's name in fuU and spelt correctly.
Readers often do not know the title of the book for which they are searching. Even if the author's name is known, it may be spell incorrectly.
Another way that a system may fail to. A system may also be of doubtful val ue to its users because it requires them to work in a way that seems nonsensical. One report describes a warehouse ment system, designed partly to increase managers' control over the routine activities of warehouse workers.
The workers found that the new system removed much of their discretion in their work, and this prevented them being able to maximize the use of scarce storage space. One of the authon was a customer at the time. Symons, It IS much more worrying wh ftw.
All 29 people on board were. H'owever, an A House of Lords select committee re recommended that the Ministry of Defence view should be set aside Colk: A client usually has innuence over approval IS gIVen to a.
Some clients but not a 0 ave e power to stop a project once it is under way. A client may also be th so, we can assume that they share the user's perspective on the kind of Ings t at be a problem. They may make only indirect use of the s stem's Insulates from the immediate experience of a badly d: Any project that ovenuns its budget can reach a point an h the total costs outweigh the benefits that will be provided on completion. In other ases the project IS cancelled, either because It IS recogmzed that costs are out of control, or because it becomes apparent the benefits will not be as great as originally promised.
The rationale is summed up in the saying: The rise of e-commerce since the late s has brought new ways for infor- mation systems to cost the organization money, sometimes as a result of what appeared to be no more than routine modifications to the software in use. One example of this caused Barclays Bank serious embarrassment in August when some customers logged onto the onli ne banking service and found that they could view other customers' accounts BBC, For example, a bricks- and-mortar retailer, threatened by rivaJs who sell at a lower price on the Internet, may have little use for an e-commerce website if it is not operational until all the customers have defected and the company has been declared bankrupt.
Many other kinds of project are time-critical. This can be due to new legislation that affects the organization's environment. An example of this was the de- regu1ation of the UK electricity supply market in April This required eiedricity companies to make extensive modifications to their computer systems so that they would be able to handle customers' new freedom to switch between suppliers.
A few years earlier, all local authorities in the UK faced a similar challenge twice in three years, when central government changed the basis for local tax calculations. Each change required hundreds of councils to specify, develop or purchase and successfully install new computer systems that all owed them to produce accurate invoices and record income collected from local tax-payers.
Failure to implement the new systems in time risked a massive cashflow problem at the beginning of the new tax year.
Commercial pressures can also have an effect. This sometimes translates into whether a business succeeds in being the fi rst to market a new product or service, although the advantage is not always permanent.
For some time the continuing success of the Internet bookstore Amazon. Some competitors notably the established US bookseller Barnes and Noble felt obliged to follow Amazon's lead. For the followers, there is not the same need to take risks with new technology. But attracting customers away from a leader may mean differentiating yourself in some way, perhaps offering new services, or perhaps by being even better at what the leader already does well. At the time of writing Spring , Amazon seems to have successfully weathered the many dot.
While Amazon still does not show a sustained profit, it recently joined Forbes. The following scenario is based on a real si tuation observed at first hand by one of the authors. Most saw the mutual benefit of sharing information, and complied. Management claimed that the daily back-up of the LAN drive onto tapes was a further benefit, since there was no longer a need to keep personal back-ups on floppy disks. Then a mC: This erased all data on the LAN drive, and when the engineer tried to restore it rom the tape, it emerged that the tape drive had nOI operated corrC: Uy for several weeks.
AJI tapes recorded over the pre- vious six weeks were useless. Staff were told that all data stored in that time was permanently lost. Re-entering it all took many person-days. The aulty disk and tape drives were replaced, and tapes are now checked after every back-up, but many staff reverted to keeping all important data on their local hard drives.
Most keep personal back-ups on floppy disks too. Perhaps nothing will persuade them to trust the LAN again. The politics with which we are concerned here arc to do with conflicting ideals and ambitions, and the play of power within the organization. There can be disagreement between management and workers, as in the case of the warehouse management system mentioned earlier in this section.
There can also be contention between individual managers, and between groups of managers. One result can be that a manager is sometimes an unwilling client in relation to a project. The following scenari o is based on another real-life situation observed by one of the authors. The head office of a multinational company decided to standardize on a single sales order processing system in all its subsidiaries throughout the world.
But the Hong Kong office already had information systems that linked electronically with customers in Singapore, Taiwan and other places in South East Asia. It became apparent that the existing links would not work with the new system. For the Hong Kong managcment, this meant extra costs to make the new system work, and disruption to established rclationships which, in their view, al ready worked smoothly and did not need 10 be changed.
They Iherefore had little desire to see the project succeed in their region, but felt they had no other choice. Had they been less scrupulous, they might have tried to find ways of sabotaging its progress, either in the hope thai it would be abandoned altogether, or at least that they might be exempted from the global rule.
Requirements can change for many reasons. This does not apply to new currently under Systems that have been in operation for time may affected. From a client's perspective, the motivation is usually to make infonnation system fit better with the business, and therefore provide better support for business activities.
This is because the developer adopts the role of 'supplier' to the 'customer' i. For this reason, when problems occur the developer may feel forced into a defensive position of justifying the approach taken during the project. Since at this stage we are discussing only prob. However, for a developer, gi ven the responsibility for building a system to meet those requirements, they can be a real headache.
If we were able to distil the essence of how many developers feel about this, it would read something like the following. No matter how skilled you are, you can't achieve anything until the users, clients, etc. Eventually, with skill and perseverance, you produce a specification with which everyone is reasonably happy. You work for months to produce a system that meets the specification, and you install il In no time at all, users complain that it doesn't do what they need it to do.
You check the software against thc specification, and you find that it does exactly what it was supposed to do. The problem is that the users have changed their minds.
Thcy just don't realize that it's not possible to change your mind late in a project. By then, everythi ng you have done depends on everything else, and to change anything you would al most have to start all over again. Or it turns out that they didn't understand the specifi cation when they accepted it Or there is some ambiguity about what it meant, and you've interpreted it differently from them. Whatever the reason, it's always your fault, even though all you ever tried to do was to give them what they wanted.
In reality, of course, analysts, programmers, etc. But this doesn' t always make it less frustrating when it happens.
There may also be an externally imposed deadline for example, a project to develop a student enrolment system that must be ready by the start of the academic year. Another external pressure results from the impatience of users and clients to see tangible results. This, too, is often understandable, since they are not SO much concerned with the information system itself, as wi th the benefits it can bring them-an easier way to do a tedious job, a quicker way to get vital information, and so on.
But it can be very counter-productive if it becomes a pressure within the project team to cut short the analysis and get on with building something any- thing! The result of haste in these circumstances is usually a poor product that meets few of the needs of its users.
Developers know this, but they don' t always have the power to resist the pressure when it is applied. These may include the use of techniques such as object-ori ented analysis , knowledge of methodologies such as the Unified Software Development Process , skill in pro- gramming languages such as Ct or Java or detailed knowledge of hardware performance such as networking devices. There must be a complementary set of skill s within the team for a project to succeed. For example, the analysts must all use the same or related techniques otherwise the results will not be coherent e.
Similarly, all programmers must be able to interpret the designs or else they will not know what the class operations arc meant to do. Problems occur when the available staff do not possess adequate expertise in the particular skills required for a project.
Many highly skilled and experienced systems analysts may know SSADM very well a structured methodology used widely in the UK , but have little experience of object-oriented analysis.
Many excellent programmers know Visual Basic inside out, but have little experience in Java or Smalltalk. As a conse- quence, some projects with highly skilled staff are still carried out rather poorly, because the staff arc inexperienced with the particular techniques they must use. Anyone who has ever tried to repair anythlng electronic or mechanical, such as a motor vehicle, washing machine or VCR, will know that much of the time is spent trying to understand what the various parts do, and how they interact with each other, This is true even when a maintenance manual is to hand.
The situation is no different for computer software. While software may be more intangible in fonn than a VCR, it is no less mechanisti c in its operation. At times this mealls that a project is forced on an unwilling 2.
But if opposing views prevail , the team may find itself the time ' h" Id b d I h' ' tted to trying to achieve w at It said cou not e one. On the other hand, they believe that the design and xecution of the software is not open to serious question. Many of these tech- are undoubtedly very talented, but this view is patently absurd since it assumes that the answer to a problem is known before the si tuation has even been investigated.
In a word, it is a prejudice. We will say no more about it, other than to comment that anyone who hopes to learn the truth about a situation must also be prepared to examine critically their own preconceptions. Type of failure Qualtty problems Productlrity problems Reason for failure The wrong problem is addressed Wider influences are neglected AnalysisJ is carried out incorrectly Project undertaken for wrong reason Usen change their minds External events change the environment Implementation is not feasible Poor project control Comment System conflicts with business strategy Organization culture may be ignored Team is poor1y ski lled, or inadequately resourced Technology pull or political push New legislation May not be known until the project has started Inexperienced project manager Figure 2.
To tm, eategory, we would add design and implementation. Even when analysis is carried out correctly, this is IIlilI no tllal ti,e ,. In Flynn's view, pro- jects. These categories are what are sometimes called 'ideal types'.
This means that they are intended to help explain what is found in reality, but that does not imply that any real example, when examined in all its detail, will precisely match anyone category. Real projects are complex, and their problems can seldom be reduced to one single cause.
Many of the examples in the following sections show some of the characteristics of more than one category of cause. Galm, In order to apply this to the quality of a computer system, clearly it is necessary to know i for what purpose the system is intended and ii how to measure its fitness.
As we shall see, both parts of this can be problematic at times. The wrong problem Here the emphasis is on the purpose for which a new system is intended. The argu- ment goes like this: At worst, it may do real harm if the objectives of an information system are in direct conflict with the organization's business strategy. The difficulty is in knowing the right viewpoint to take when defining the aims of a project.
If the aims of the organization as a whole are unclear, or are not communicated to those responsible for planning information system projects, there is always a risk that a project will fall into this error. It may then be regarded as a complete success by its developers and users, yet appear a failure when seen in a wider frame of reference. A primary cause of system fai lures in general is that some projects are started with no clear idea about exactly what are the nature and goals of the client organization.
This means that failure, or at least lack of success, is almost inevitable. If an organization itself is not understood, then it is extremely difficult to identify and develop information systems that support it in fulfilling its aims. Neglect of the context This emphasizes the fitness of an information system to fulfil its purpose. This can take the form of a system that is too difficult to use, since the designers have taken insufficient account of the environment in which its users work, or the way that they like to work.
Some examples given earlier in this chapter can be interpreted in this way, depending on assumptions about the situation. For example, in one case cited earlier Section 2. Managers believed they needed to control the activities of workers more closely.
Yet the system designed to do this had also the side effect of obstructing the workers from carrying out their work in an efficient way, to the detriment of the whole company. Even if the aims are clear at the outset, purpo pitfalls lie along the route, particularly if the development team does not many. This may be because they lrectJy affect the external design of the system e.
To give a recent example, over the last few years many organizations seem to have rushed into some sort of e-commerce activity. While this is clearly a great success for some, others have derived little or no benefit. During the year there seemed to be so many 'doLcom' crashes that the Guardian newspaper's website ran a column called ' DoLcom deathwatch', which featured only stories about troubled and failing Internet companies.
It has been pointed out that many organizations did not think carefully enough about some key questions McBride, The sensible conclusion must be that there is a great deal more to successful trading on the Internet than just writing a few web pages in HTML and placing them on a web server where surfers can find them. Some organizations, in moving onto the World Wide Web, are simply following a trend. They do so not because they under- stand what it can do for their business, but precisely because they do not under- stand, and therefore fear the consequences of being left out of something good.
Two possible underlying reasons explain why this can happen. One of these is political push within the organization. For example, a powerful group of managers may wish the business to look modem, even when no clear benefit has been identified. The other is the pull of new technology. An organization is very vulner- able to this if the most senior managers have little understanding of information technology, and therefore no rational basis for evaluating the exaggerated claims that vendors are prone to make about their newest products.
In practice these two reasons often combine into a single force that can be irresistible. Other companies, typically Internet start-up businesses with no history of trading in the physical world, seem simply to have believed that the Internet represented a 'new economy' where establi shed business rules no longer appli ed. It is still just as important to plan and design with care, to pay attention to costs and income and to ensure that proper controls are in place.
Much of the na'ive and sloppy thinking that led to the dot. If quality is the Hrst concern of users and clients, then productivity is their other vital concern.
The questions that are likely to be asked about productivity are as follows. Requirements drift Requirements almost always drift if they are allowed to do so. In the simplest case, this just means that users' requests change over time. In principle, it would be unreasonable to prevent this from happening, but, in extreme cases, change requests can bedevil a project and even prevent its completion.
The longer a project proceeds, the more complex both its products and its documentation become. To compound this, each part of the final system probably depends on many others, and the inter-dependencies grow more numerous and complex. Jt becomes progressively more difficult over time to make changes to a system under development, because any change to one part requires changes to many other parts so that they will continue to work together.
A limit is reached when a project is stalled by an escalating workJoad of making all the changes required as a conse- quence of other changes that were requested by the users. At this point, manage- ment have onJy two choices. They can cancel the project and write off the money spent so far this happened at the London Stock Exchange in , Alternatively, an effort can be made to bring the project back on track.
This is almost always both difficult and expensive, and requires highly skilled management. External events This is one cause of failure that is nonnally beyond the control of both project team and higher management.
Depending on the environment in which the organiz- ation operates, decisive external events can even be impossi ble to anticipate.
Nevertheless, it is prudent on any project at least to assess the vulnerability of the project to external events, since some are much more at risk than others. For example, a project to build a distributed infonnation system that is to operate on new, state-of-the-art computers communicating over public telephone circuits may be sensitive to external factors such as the reliability of the telephone network and call pricing. By contrast, a project to build an infonnation system that will operate on existing, tried and tested hardware within one building can safely ignore these factors.
POO P f project is ultimately responsibl e for Its successful completion, The be argued that any project failure is also a failure of the and It cou ment To some extent this is true, but there are also some cases ro,'ect manage. P the only identifiable cause of failure overall IS a fadure 0 management. S where I ys due to either poor planning at the start, or a lack of care In.
As a result, the manager a ows t e prOJec 0 momtonng pr V'. II d erm its its costs to grow In an uncontro e way. IS IS partlcu ar y e eVldent unu a. The problems of testmg and e uggmg a new sys em steadi ly more complex as attention is focused on larger and larger sub- t Sometimes the task of interfacing several large, complex software systems, sys ems.
An example of the kind of complex installation tackled. A further difficul ty occurs when each supplier that their product is perfonning exactly as specified and any problem must he elsewhere.
It can then be very hard to locate the exact origins of the problem. Techni cal problems with the implementation do not ah.. A practical difficulty that encountered In Its use was with the deli berately high-tech design that used a digital The digital map interface was very unpopular with its users, who It difficult to pinpoint an ambulance with sufficient accuracy and thus regarded It as.
Barker, This was a contributory factor to the system's to diS- patch ambulances speedily to emergencies, resulting from the appitcallon of a new technology to a critical task that was not sufficiently.
An implementation problem led to the crash of the onhne sportswear Boo. The software for their website was much delayed m development, but proved a disaster even when delivered. It out few home pes were sufficiently advanced to run the sophisticated 3D vlsuahz- ations without crashing. Even when the software ran without crashing, most images were very slow to download, adding to the users' frustra. Begmmng With Chapter 12, we consider later what the software designer can do t avoid costly and damaging faiJures.
In practical terms, thinking about ethics normally means that we are trying to establish a way of judging the effects that one person's behaviour has on other people. Given that all computer-based information systems have a direct effect on someone's life, it is hard indeed to think of one that does not have a significant ethical dimension to its design, construction or use.
As wit h any system of rules it is necessary to consider the conseque. It is not the case to say that a breach of the ethical rules that apply to any aspect of an infor- mation system project is net: That is why it is important for you to attend both the class and lab sessions see attendance policy discussion. Your prior MIS classes have emphasized database development and programming, so I assume that you already have those skills. Since that is what you already know, most of you will be eager to start programming as soon as you find a client.
In other words, this class provides you with: Your portfolio is a collection of all the projects completed as part of your MIS degree that illustrates your accomplishments and skills as an MIS professional.
The portfolio is an excellent tool to share with prospective employers since it shows the work you are capable of doing. Software We will use MS Project for some assignments. You may use MS Visio or any other drawing tool for other assignments. Free student versions can be downloaded at: Assignment and Grading Policies Our textbook covers the most current object-oriented analysis and design techniques used by systems development professionals worldwide.
The slides will introduce new concepts, especially about project management, as well as address material from the textbook. Late work will not be accepted. Submit your work on time.
Part of project management is to anticipate the unexpected and build appropriate contingency into the work plan. Do likewise when creating personal work plans to complete assignments. If you "have" to have a particular grade in this class, "earn" it. Don't tell me you need an "A" the last week of class when you have earned a low "C" all semester. It is your responsibility to earn whatever grade you need. Grading Scale Grades will be assigned as follows and posted on Blackboard.
It cannot have anything stapled, glued, taped or otherwise affixed to it. Tests cover material covered in the textbook, PowerPoints, lectures, class discussions, and material presented by guest speakers, including student presentations. They will emphasize interpretation and application of course material, not rote memorization. As much as possible, each test will cover material in the portion of the class that precedes it.
However, knowledge is cumulative, and successful completion of a test may require mastery of material covered earlier in the semester. Take tests during the scheduled time. If, due to emergency or illness, you know you will miss a scheduled test, it is your responsibility to let me know ahead of time or worst case, within 24 hours of the missed test.
Make up examinations may be oral, essay, or another format, as determined by the instructor. Project Heavy emphasis is placed on the project. You need to find a real company to work with.
Your stakeholder has to be present for the final presentation. The content of the project is up to you, but it has to involve techniques and software development tools you learned in your MIS and CS classes. Everything you learn in this class should be applied to the project. Include as many project management tools as possible; for example, one of the first things to create for your project will be a Gantt chart. Specific guidelines for projects will be discussed in class and posted on Blackboard.
Project ideas from the past: I expect everyone to be active contributors to the project, in terms of quality and quantity of input and by helping other team members to do their best.
Unprofessional performance and free-riding will be reflected in a student earning a lower grade for the project and its associated deliverables, and may mean that one or more members of the team will receive a failing grade. Special Topic Presentation Two people who are in different project teams will select a topic, conduct research about the topic, and make a presentation to the class. If appropriate to your topic, include an interactive element to your presentation.
In addition, you must create two questions about the material covered in your presentation that are appropriate for inclusion on one of the two tests. Presentations should be minutes long. PowerPoint slides and questions must be sent to me no later than 9 a. I will post your presentation to Blackboard so that it is available to your classmates. Executive Presence Much of your learning will occur as you prepare for and participate in class discussions.
Most people in business are evaluated on what they say, how they say it, and how they present themselves. The classroom gives you the opportunity to hone your discussion, debating, and impression management skills. Your participation will be evaluated based on what you contribute, not just what you know. Effective participation has much more to do with quality than quantity.
In other words, those who dominate air time without contributing to the advancement of the discussion will not be rewarded. Executive presence also includes not engaging in activities that show disrespect to me or to your fellow students, including talking on cell phones, texting, or browsing the Internet during class.
I realize that individuals come to this class with different backgrounds in technology topics.