How is a system tested, and what test data proves it works?
Levels of testing (unit, integration, system and acceptance), the purpose of a test plan, and the use of normal, boundary and erroneous test data.
A CCEA A-Level Digital Technology answer on testing: the levels of testing (unit, integration, system and acceptance), the purpose of a test plan, and the use of normal, boundary and erroneous test data with expected results.
Reviewed by: AI editorial process; not yet individually human-reviewed
Have a quick question? Jump to the Q&A page
Jump to a section
What this dot point is asking
CCEA wants you to describe the levels of testing a system passes through (unit, integration, system and acceptance), to explain the purpose of a test plan, and to choose normal, boundary and erroneous test data with expected results. Testing is the life-cycle stage that proves the system meets its requirements before it goes live.
The levels of testing
Testing builds upwards: prove each small part works (unit), then prove the parts work together (integration), then prove the whole meets the requirements including non-functional ones such as speed (system), then let the customer confirm it does what they need (acceptance). A fault is cheaper to find and fix at the unit level than after the whole system is assembled.
The test plan
A good test plan is written from the design and requirements, so each requirement has at least one test that proves it is met. Recording the expected result before running the test is what makes the test meaningful, the actual result is then compared against it.
Normal, boundary and erroneous test data
Testing only with normal data is the classic mistake: it confirms the happy path but misses the edges and the invalid inputs where most defects live. A field accepting 0 to 100 must be tested at 0 and 100 (accept), just outside at -1 and 101 (reject), a typical 57 (accept), and with rubbish such as "abc" (reject).
Why testing matters
Untested or lightly tested software fails on real data, often at the boundaries or on invalid input, which damages trust and can corrupt data. Systematic testing with a plan and well-chosen data gives evidence that the system meets its requirements, which is exactly what acceptance testing then confirms to the customer.
Try this
Q1. State the level of testing that confirms the customer's needs are met before sign-off. [1 mark]
- Cue. Acceptance testing.
Q2. For an age field accepting 18 to 65, give one boundary value and one erroneous value. [2 marks]
- Cue. Boundary: 18 or 65 (or just outside, 17 or 66). Erroneous: a non-number such as "twenty", or a clearly invalid age such as -3.
Q3. Explain why a test plan records the expected result before the test is run. [2 marks]
- Cue. Comparing the actual result to a pre-stated expected result makes the pass or fail objective and repeatable, rather than judged after the fact.
Exam-style practice questions
Practice questions written in the style of CCEA exam questions on this dot point, with worked answer explainers. The year tag is the paper they imitate, not the source.
CCEA AS 16 marksDescribe the different levels of testing carried out on a system as it is developed.Show worked answer →
Work up the levels in order, defining each by what it tests.
Unit testing checks a single module or function in isolation, confirming it behaves correctly on its own. Integration testing checks that separately tested modules work correctly together once combined, focusing on the interfaces between them. System testing checks the whole, complete system against the requirements specification, including non-functional requirements such as performance. Acceptance testing is carried out with or by the customer to confirm the system meets their needs and is fit for purpose before sign-off.
Markers award marks for each level correctly named and described, and for showing the progression from small parts to the whole system to customer sign-off. Confusing integration testing (modules together) with system testing (whole system against requirements) is the common error.
CCEA AS 14 marksA field accepts a percentage mark from 0 to 100. Give one example each of normal, boundary and erroneous test data, and state the expected result.Show worked answer →
Give one value of each type with the result the test should produce.
Normal data: a typical valid value within range, for example 57. Expected result: accepted. Boundary data: a value on the edge of the valid range, for example 0 or 100 (and just outside, such as 101, to test the limit). Expected result: 0 and 100 accepted, 101 rejected. Erroneous data: data of the wrong type or clearly invalid, for example the text "abc" or the value -5. Expected result: rejected with an error message.
Markers reward one correct value of each type and the matching expected result. Choosing 50 for both normal and boundary, or giving values without an expected result, loses marks. The point of boundary testing is to probe the exact limits where off-by-one errors hide.
Related dot points
- The stages of the systems development life cycle, the activities carried out in each stage and the documented outputs each stage hands on to the next.
A CCEA A-Level Digital Technology answer on the systems development life cycle: its stages from feasibility through analysis, design, implementation, testing and maintenance, the activities in each, and the documents each stage produces.
- Designing the inputs, outputs, processing, data storage and user interface of a system, and the design tools used: data flow diagrams, system flowcharts, entity relationship diagrams and the data dictionary.
A CCEA A-Level Digital Technology answer on the design stage: specifying inputs, outputs, processing, data storage and the user interface, and the design tools used including data flow diagrams, system flowcharts, entity relationship diagrams and the data dictionary.
- Implementation and installation of a system, the changeover methods (direct, parallel, phased and pilot), their advantages, disadvantages and suitability, and data conversion and user training.
A CCEA A-Level Digital Technology answer on implementing and installing a system: the direct, parallel, phased and pilot changeover methods, their advantages, disadvantages and suitability, and data conversion and user training.
- The purpose and contents of technical documentation and user documentation, and the audience each serves during development, use and maintenance.
A CCEA A-Level Digital Technology answer on system documentation: the purpose, contents and audience of technical documentation and user documentation, and why each matters during development, use and maintenance.
- The distinction between data, information and knowledge, data types, and the validation and verification techniques that protect data quality.
A CCEA A-Level Digital Technology answer on data and information: the distinction between data, information and knowledge, common data types, and the validation and verification techniques that protect data quality on entry.
Sources & how we know this
- CCEA GCE Digital Technology specification — CCEA (2016)