Skip to the page
Year 10 · ICT · Spring 2 · The systems life cycle; presentations

Development, testing and implementation

TheoryPresent ▶

What this lesson covers

7.3 Development and testing

Cambridge IGCSE ICT 0417 · Topic 7: The systems life cycle · Paper 1

7.4 Implementation

Cambridge IGCSE ICT 0417 · Topic 7: The systems life cycle · Paper 1

You will learn to

  • Explain why a system must be tested before it is used.
  • Choose normal, abnormal and extreme test data and give the expected outcome.
  • Describe direct changeover, parallel running, pilot running and phased implementation.
  • Choose an implementation method for a scenario and justify it.

Success criteria

  • I can complete a test plan row with test data and the expected outcome.
  • I can give normal, abnormal and extreme test data for a mark from 0 to 100.
  • I can explain why a hospital might choose parallel running.
The syllabus points for these sections
  • Explain why a system must be tested before it is used.
  • Describe testing of modules, functions and the whole system.
  • Complete a test plan with test data, expected outcome, actual outcome and remedial action.
  • Choose normal, abnormal and extreme test data.
  • Explain the use of live data.
  • Describe direct changeover, parallel running, pilot running and phased implementation.
  • Give advantages and disadvantages of each method.
  • Choose the most suitable method for a given scenario and justify it.

Key words

Say each word, then read what it means. Every one is in the glossary too.

Systems life cycle
The stages used to create a new system: analysis, design, development and testing, implementation, documentation, evaluation.
Systems analyst
The person who studies the current system and plans the new one.
System specification
A list of the hardware and software the new system needs.
Field
One item of data in a record, such as Surname.
Validation
An automatic computer check that data is sensible and allowed.
Check digit
An extra digit calculated from the other digits, used to detect input errors.
Data capture form
A paper or on-screen form used to collect data.
Test plan
A table listing each test, its data, the expected and actual outcomes and any remedial action.
Normal data
Sensible data that should be accepted.
Abnormal data
Data that should be rejected, such as out of range or the wrong type.
Extreme data
Data at the limits of the allowed range, which should be accepted.
Live data
Real data from the organisation, used to test the system with known results.
Remedial action
What was done to fix a test that failed.
Direct changeover
The old system stops and the new one starts immediately.
Parallel running
Old and new systems run together until the new one is trusted.
Pilot running
The new system is tried in one branch before the rest.
Phased implementation
The new system is introduced one part at a time across the organisation.

The explanation

Read it here, or open the study page below to go one step at a time with things to try.

7.3 Development and testing

Why test, and the test strategy

Development means building the system: creating the file structures, forms, validation and reports.

Testing checks that the system works as the design says, and that it handles mistakes without crashing.

A test strategy decides what is tested and in what order: each module (a separate part), then each function (one task, such as 'record a loan'), then the whole system working together.

e.g.Example
Library system modules: member records, book records, loans, reports. Test the loans module alone first. Then test the function 'return a book'. Then test the whole system: a new member borrows a book, returns it late, and appears on the overdue report.
?Why?

Testing small parts first makes errors easy to find. If the whole system fails at once, the fault could be anywhere.

Types of test data

Normal data is sensible data that the system should accept.

Abnormal data is data that should be rejected: outside the range or the wrong type. Some books call it erroneous data.

Extreme data is data at the very limits of what is allowed. It should be accepted.

Computer Science 0478 also uses boundary data: the limit value and the value immediately beyond it. ICT 0417 asks for normal, abnormal and extreme.

Field ruleNormalExtremeAbnormal
Exam mark 0 to 10056, 730, 100101, −5, 'sixty'
Year group 7 to 139, 117, 136, 14, 'ten'
Loan length 1 to 21 days141, 210, 22, 3.5
e.g.Example
Try the toy with the range 1 to 100. Classify 50, 1, 100, 0, 101 and 'abc'. Notice that 1 and 100 are extreme and accepted. 0 and 101 are rejected. ICT calls them abnormal. The toy labels them boundary, the Computer Science term for the value just outside the range.
!Watch out

Extreme data is accepted. Students often think extreme means 'too big'. Too big is abnormal.

There is something to try for this step in the IGCSE library.

The test plan

A test plan is a table written before testing. Each row is one test.

The first columns are: what is being tested, the test data and the type of data.

Then come the expected outcome (what should happen), the actual outcome (what did happen) and the remedial action (what was done to fix a failure).

If the actual outcome matches the expected outcome, the test passes. If not, the system is fixed and the test is run again.

TestTest dataTypeExpected outcomeActual outcomeRemedial action
Mark field range 0 to 10056NormalAcceptedAcceptedNone needed
Mark field range 0 to 1000ExtremeAcceptedAcceptedNone needed
Mark field range 0 to 100100ExtremeAcceptedRejected with errorChanged rule from <100 to <=100, retested: accepted
Mark field range 0 to 100101AbnormalRejected, error message 'Mark must be 0 to 100'Rejected, message shownNone needed
Mark field range 0 to 100sixtyAbnormalRejected, error message 'Enter a number'Accepted and savedAdded type check, retested: rejected
Surname presence check(left blank)AbnormalRejected, 'Surname is required'Rejected, message shownNone needed
?Why?

Writing the expected outcome before testing stops the tester accepting whatever happens. The test plan also becomes evidence in the technical documentation.

Live data

Live data is real data from the organisation, such as last month's actual library loans.

After testing with made-up test data, the system is tested with live data whose results are already known from the old system.

If the new system gives the same results as the old one, it is working correctly with real-world data.

e.g.Example
The library enters last October's 1,240 real loans. The old card system recorded 63 overdue books. The new system's overdue report should also show 63.
!Watch out

Live data contains real personal data, so it must be kept secure during testing.

7.4 Implementation

Four ways to change over

Implementation means putting the new system into use. Staff are trained and data from the old system is transferred.

Direct changeover: the old system stops and the new one starts at once, for example over a weekend.

Parallel running: the old and new systems both run at the same time, until the new one is trusted.

Pilot running: the new system is used in one branch or department first. Once it works, it goes to the rest.

Phased implementation: the new system is brought in one part at a time, across the whole organisation.

e.g.Example
A chain of 40 supermarkets gets a new till system. Direct: all 40 shops switch on Monday morning. Parallel: every shop runs old and new tills side by side for a month. Pilot: one shop in Getafe uses it first, then the other 39. Phased: all shops get the new payments part first, then stock control a month later.
!Watch out

Pilot and phased are easy to mix up. Pilot = one place, whole system. Phased = whole organisation, one part at a time.

Advantages and disadvantages

Each method trades cost and speed against risk.

The key question: what happens if the new system fails?

MethodAdvantagesDisadvantages
DirectCheapest: only one system runs and staff do the work once. Benefits are immediate. Fastest.Highest risk: if the new system fails, there is no backup and work stops. Data could be lost. Staff must be fully trained before the switch.
ParallelLowest risk: if the new system fails, the old one is still working. Results can be compared to check the new system.Most expensive: staff do every job twice, so extra staff or overtime are needed. Takes longest. Data entered twice can lead to errors.
PilotIf it fails, only one branch is affected. Staff in the pilot branch can train others. Cheaper than parallel.Slow to reach the whole organisation. If the pilot branch fails, it suffers on its own. The pilot branch may not be typical of the others.
PhasedIf one part fails, only that part is affected and it can be fixed. Staff learn one part at a time.Slow: takes a long time to fully implement. Parts of old and new systems must work together, which is hard. Costly to evaluate each phase.
?Why?

There is no best method. The right one depends on how bad a failure would be, how much money there is, and whether the system can be split into branches or parts.

Choosing the right method

Use direct when the system is small, simple, or the old system has already stopped working. Also when two systems cannot run side by side.

Use parallel when failure would be serious, such as payroll, banking or hospital records.

Use pilot when the organisation has many similar branches, such as shops or banks.

Use phased when the system splits into separate parts, such as a school system with attendance, marks and timetable modules.

e.g.Example
Try the toy. For each scenario, pick a method and read why. Scenario: a hospital moves patient records to a new system. Best answer: parallel running. A failure could put patients at risk, so the old system must keep working until the new one is proven.
!Watch out

In a 'choose and justify' question, the justification must link to the scenario. 'Parallel is safest' is weak. 'Parallel, because if the new payroll fails, 300 staff still get paid by the old system' is strong.

There is something to try for this step in the IGCSE library.

Study it

Open the study page and go one step at a time.

ICT topic 7

The systems life cycle

Step by step, with exam questions and a quiz.

ICT topic 19

Presentations

Step by step, with exam questions and a quiz.

ICT topic 15

Proofing

Step by step, with exam questions and a quiz.

Topic 14 · Year 9

Programming further

Real programs keep data in lists, take text apart as strings, split work into subroutines, and get tested on purpose.

Check yourself

One question at a time. The reason shows after each answer.

Saved on this device.