MERCEDES PASSANO

SECTION 8: System Analysis and Design:

Implementing a new computer system

There are four main stages in setting up a new computer system:

1.Analysis - studying the problem

2.Design - designing a solution

3.Implementation - putting the solution into effect and testing it

4.Evaluation - checking that the solution is working as intended

This process is often called systems analysis.

1- Analysis:

The analysis is the first stage you must tackle when developing an information system. Analysis is the process of deciding what (if anything) the information system should do. It is not concerned with how it will be done. During the analysis you should not discuss which computer program you are going to use to produce the system. You do not decide on this until the design stage.

Sometimes it is not clear whether or not it is a good idea to create a new information system to perform a job. If there is an existing system (manual or computerised) doing the same job as the new system will then it may be that a new system would offer few or no benefits over the older system. The analysis should consider whether or not to set up an information system.

The analysis stage of the design of an information system can be broken down into three smaller steps :

a)Feasibility Study

b)Breaking Down The System Into Tasks

c)Requirements Specification

Developing and activating a new computer system is a long process. Because of this, it is important to break the process down into smaller stages, each one requiring a different set of professional skills.

An important part of this process is the first stage analysis. This involves finding out whether the new computer system is needed at all and exactly what it will be used for.

Analysis is all about looking at how a job is done at present and seeing if the job could be done better by computer. To find out, the systems analyst (person whose job it is to investigate the current system and design and implement the new Computer System) might:

observe staff at work

interview staff about their work

send out questionnaires about working practices

inspect documents such as forms, invoices and receipts.

1-a) Feasibility study: (estudio de viabilidad)

The purpose of a feasibility study is to determine whether or not a new information system should be created to solve a problem. The possible advantages of creating a new system must be identified. If there are few advantages then it may not be worth investing in a new system.

Finding Out About Existing Systems

If the job that the information system will carry out is currently being performed by another system then a good starting point for a feasibility study is to examine the existing system. Even if it has already been decided that a new system will be created it is still a good idea to examine any existing system. Any problems with the system can be identified and new ways of performing the job can be considered. A variety of methods can be used to identify how the existing system works. These include :

Method Description:

-Interviews: People who operate the current system can be interviewed to find out what they think of it. Interviews can be in more depth than a questionnaire could be. If an interviewee makes a useful point about the system then the person doing the interview can follow the point up to get further information. This is not possible with a questionnaire.

-Questionnaires: Questionnaires are useful when information about a system has to be gathered from a lot of people. Information can be obtained far more quickly by using questionnaires than carrying out interviews. Because people can remain anonymous they may be more likely to tell the truth than they would in an interview.

-Observation: Observation of how the current system works and what each person does to use it can give an unbiased view of how the system works. Because the person designing the information system is doing the observation himself it is possible to get first hand information about how the system works. The system developer can identify if any of the workers have been lying about their role in the system.

-Documentation: Manuals describing how the current system operates can be collected. Any forms used to collect information for the current system or reports produced by the current system can be looked at.

If there is no existing system then it will be necessary to use other sources of information to decide how a new system should operate. Possible methods of achieveing this include :

-Looking at systems which perform similar tasks elsewhere.

-Talking to the people who the system is being created for to see what they expect the system to do.

Describing An Existing System

Once evidence about the existing system has been collected a description of the system can be produced. In addition to a written description of the system, system flowchart diagrams can be used to show how the existing system operates. Click here to find out more about system flowcharts.

1- b) Breaking Down The System Into Tasks

Usually an information system can be broken down into a number of distinct tasks. For example a system used to manage a car repair garage could be broken down into these tasks :

-Designing an advertisement poster to be displayed around the town the garage is in.

-Booking cars into the garage at appropriate times.

-Producing invoices (bills) for customers showing what work has been done on their cars.

-Keeping track of the spare parts that the garage has in stock.

-Printing pay slips for the garage employees at the end of the week.

Breaking down a job into smaller tasks that can be worked on more easily is known as top down design. The work done in the feasibility study should help identify these tasks. From the garage example you can see that some of the tasks are completely unconnected whereas there may be links between the other tasks :

-The advertisement is not connected to any other task.

-The invoice and stock control system could be linked together. When parts names are written onto an invoice to show that they have been used in an repair then the parts can automatically be taken off the stock list by the stock system. The price of the parts used in the repair can be filled onto the invoice from the stock database.

For a small information system the individual tasks can be tackled one at a time. For a large information system different tasks may be worked on by different people. If a task you have identified is still quite large then you may want to break it up into even smaller parts to work on when you produce your design. Continuing to break a task up into smaller parts like this is all part of the top down design process.

From this point on each task should be treated separately. A separate requirements specification, design, implementation, testing and evaluation stage should be produced for each task.

1- C) Requirements Specification

Once it has been decided that the new system should go ahead, the next step is to draw up a requirements specification saying exactly what the new system will do. For example, it will say:

what hardware is needed

what software is needed

what inputs are needed

what processing must take place

what information needs to be output

If it is decided that a new information system should be developed then it will be necessary to decide what the system will do. The requirements specification sets out what the new system will do. It will guide the development of the entire project.

For a commercial project the requirements specification may form part of the contract between the software company and the people the system is being developed for. The software company will not get paid until the system operates as set out in the requirements specification.

A separate requirements specification should be produced for each task that has been identified. Each requirements specification should deal with :

1)Introduction

2)Outputs

3)Inputs

4)Desired Outcomes

5)Alternative Methods

The information you should include in each section is outlined below. A good analysis will discuss why things are being done and justify them.

1) Introduction

For each task you need a separate introduction. Whoever assesses your work will not be familiar with the problem that you are tackling and so you should include as much material as possible here. If the assessor does not understand exactly what the task you are tackling is then you are unlikely to gain good marks for the rest of the task.

2) Outputs

An output is a piece of information produced by an information system. Under the outputs section you should explain what you want the system to tell you. Typical outputs might include :

  • The questions you want a database to answer using queries.
  • The information that you want a spreadsheet to calculate.
  • Any printouts you are going to produce with a DTP/graphics/word processor package.

You should explain what each output is and, if possible, why it is required.

3) Inputs

An input is a piece of information that is typed into an information system. Based upon the outputs that you want to produce you should be able to decide what information you will need to put into the computer system. This could be :

  • Information that will be stored in database records.
  • The information that a spreadsheet will require to perform calculations.
  • Where you will obtain text/pictures from for publications, and what information must be contained in the publications.

4) Desired Outcomes

When you have completed the task you must produce an evaluation of your work. In the evaluation you need to assess the work you have done and compare it to what you hoped to achieve. The results you hope to achieve are known as desired outcomes. To produce a good evaluation you must compare what you have achieved directly to a set of desired outcomes that you produce now. It is therefore vital that you set your desired outcomes out clearly now.

The first desired outcomes will be that the outputs you want to obtain are produced correctly. After this desired outcomes should include any other properties that you would like your system to have. Other possible desired outcomes include :

  • Outputs / inputs should be presented in a particular way. e.g. to include a particular logo or to match a particular layout which you have specified.
  • Validation should be performed on inputs.
  • The task should achieve a particular effect, e.g. attract people to an event.

Desired outcomes should be clear and unambiguous. It is difficult to judge if unclear outcomes have been met.

5) Alternative Methods

This is the first point at which you should consider how you will complete the task. Under this heading you should consider possible methods that could be used to complete the task. You should compare the advantages and disadvantages of :

  • A manual solution : E.g. producing a poster using paints or storing information in a card file system. You should outline a possible manual system if there is not an existing one.
  • A computerised solution : At this stage you do not need to name the specific package you will use but can suggest some possible packages.

You need to explain why the computerised solution is most appropriate (if it isn’t choose a different task!).

Summary of Documentation Required for Analysis Stage

  • Feasibility Study - Is there an existing system to replace ? Is it worth creating a new system ?
  • Breaking Down The System Into Tasks - What separate tasks can the project be divided into ?
  • Requirements Specification - What should each task do ?

Having investigated the present system, the systems analyst will produce a feasibility study. This will look at whether the new system is:

technically feasible - is the new system technically possible to implement in the time available?

economically viable - will the cost of the new system be offset by savings once it is implemented? In other words, will it save the organisation money?

Only if the answer to both these questions is yes is the project likely to continue to the next stage. At this point the decision-makers in the organisation, such as the board of directors in a company, decide whether to go ahead.

2- Designing the system

Once the requirements are known, the system can be designed. For the system to work, any or all of the following might have to be designed:

data capture forms ( a form specially designed for collecting data).

output formats

file structures e.g. what fields are needed

what hardware is needed e.g. will it be networked?

what software is needed

flow diagrams to show how the system works (An illustration showing how a system will work).

The design will look at different possible solutions at each stage and give reasons for any decisions made. For example in deciding what hardware to use, the design will consider what type of computer is best and what sort of printer is needed.

The design will also include a test plan to explain exactly how the new system will be tested. The test plan will include the expected outcomes for each test.

The designs are then used by other members of the team during the implementation stage, for example the programmers.

The design phase might included designing new software to be specially-written for the organisation - though more often the software will be bought-in.

Design is the process of deciding how you will set up your information system using the software packages you choose. To produce a good design you need to show how you have arrived at the design you are going to use, and not simply state the design. i.e. you need to justify the choices you have made and perhaps consider alternatives at each step in the design.

The design stage can be broken down into four smaller steps :

a)Selecting Software

b)Selecting Hardware

c)Planning

d)Test Plans

2- a) Selecting Software

Each task could be tackled in a variety of ways. It is vital that you look at a number of possible alternatives solutions before you start work creating your system. If you do not do this then you could do an awful lot of work only to find out that there was a better, easier way to solve the problem.

You will therefore need to decide which piece of software to use. For example, is a spreadsheet or a database more suitable for keeping records of all of the members of a club ? Factors such as ease-of-use, speed, availability of software, your familiarity with the software, staff training, and the features of a software package will influence your choice.

The three important points you must cover when selecting software are :

  • Identify the activities that you will need to carry out to tackle the problem.
  • Investigate and explain how each of these activities could be carried out using two or three different methods.
  • Draw a conclusion about which is the best method to use. Be sure to justify your conclusion.

Below is an example of how to go about selecting a piece of software.

Example Software Selection For Writing An Essay Including a Table.

Having considered the task I have identified a number of different activities that I will need to carry out to write the essay. I have also identified some other activities that would help me write the essay but are not essential. In order of importance these activities are :

  1. Enter and output ordinary text.
  2. Produce tidy tables.
  3. Let me correct errors.
  4. Check my spelling.
  5. Let me use different typestyles to highlight titles.
  6. Number the pages.

I have considered two possible methods of producing the essay. They are (i) using Notepad and (iii) using Microsoft Works. Both of these methods are available for me to use. In the table below I have explained how each of the activities can be carried out using each of the methods.

Activity / Notepad / Microsoft Works
Enter and Output Ordinary Text / Text is typed onto the screen. It can be output by selecting Print from the File menu. / Text is typed onto the screen. It can be output by selecting Print from the File menu.
Produce Neat Tables / Entries in tables can be lined up neatly using the TAB key. / Tables can be created by Selecting Insert Table from the Insert Menu. Entries are automatically aligned. Borders can be put around the table to improve its presentation. Rows and columns can be easily added or deleted.
Correct Errors / Errors can be corrected before printing by deleting the incorrect text using the DELETE key. Extra information can be added at any point in the document. / Errors can be corrected before printing by deleting the incorrect text using the DELETE key. Extra information can be added at any point in the document.
Check Spelling / The essay will have to be read on-screen and any difficult words looked up in the dictionary. If there are any mistakes they can be corrected before the essay is printed. / Spellings can be checked automatically using the Spell Check option. This will identify most mistakes and can suggest the correct spellings.
Use Different Styles for Headings / The font can not be altered. Titles can be highlighted by double spacing letters or typing in capitals. / The font and font size can be changed for headings. They can be underlined or made bold to make them stand out more.
Number the pages. / Page numbers can be included in the footer and will appear on each page. The software works out the number of each page. / Page numbers can be included in the footer and will appear on each page. The software works out the number of each page.
My Familiarity with the Method / I have used Notepad to edit text files on my PC. / I have previously typed a number of small articles on Microsoft Works.

The best option is to use Microsoft Works. It will allow me to carry out all the tasks I want to. I can correct my text on screen before printing and Works will help me find my spelling mistakes - saving a lot of time. Tables can be easily created and edited in a range of attractive styles. A number of different fonts can be used to make titles stand out. Works will also automatically number my pages for me.