Combining Test Techniques

Combining Test Techniques (ISTQB CTAL-TA): A Complete Guide with Exam Tips

Introduction
In the ISTQB Advanced Level Test Analyst (CTAL-TA) syllabus, the chapter on Selecting Test Techniques and Automating Test Design includes the topic of Combining Test Techniques. The idea behind it is simple: no single test technique finds every type of defect. A good Test Analyst therefore picks a mix of techniques that, together, give strong coverage of the test basis, the risks and the expected defect types. This guide explains what combining techniques means, why it matters, how it works in practice, and how to answer exam questions on it.

1. Why Combining Test Techniques Is Important
Each test technique is built to target certain kinds of defects. Each one also has blind spots.

Defect diversity: Each technique is good at finding certain defects and poor at others.
- Equivalence partitioning finds incorrect handling of a whole class of values. It can miss errors that occur exactly at the edges.
- Boundary value analysis catches off-by-one errors. It does not find errors in combinations of conditions.
- Decision table testing finds missing or wrong business rules. It does not check state-dependent behaviour.
- State transition testing finds invalid or missing transitions. It does not examine data boundaries.
- Experience-based techniques such as error guessing, exploratory testing and checklists find defects that the specifications do not anticipate.

Risk-based testing: Higher-risk areas justify more rigorous techniques. Combining techniques lets coverage intensity match the risk level.

The pesticide paradox: Repeating the same technique and tests eventually stops finding new defects. Mixing techniques keeps testing effective over time.

Efficiency: Techniques often build on each other. For example, the partitions from equivalence partitioning become the input for boundary value analysis. This reduces duplicated effort while raising thoroughness.

Covering gaps in the test basis: When specifications are incomplete or ambiguous, black-box techniques alone are limited. Experience-based techniques fill those gaps. Black-box techniques then provide systematic, measurable coverage where specifications exist.

2. What Combining Test Techniques Means
Combining test techniques is the deliberate use of two or more techniques on the same test object, feature or test condition. The aim is greater defect-detection effectiveness and coverage than any single technique could achieve. Factors that influence the selection and combination of techniques include:
- The type and complexity of the component or system
- Regulatory and contractual requirements, for example safety-critical standards that demand specific coverage
- Customer or stakeholder requirements
- Risk levels and risk types
- Test objectives
- Available documentation, meaning the quality and form of the test basis
- Tester knowledge and skills
- Available tools
- Time and budget
- Software development lifecycle model, for example Agile versus sequential
- Expected use of the software
- Previous experience with the techniques
- Types of defects expected in the component or system

3. How It Works: Common Combinations

a) Equivalence Partitioning + Boundary Value Analysis
This is the classic pairing. First, identify the valid and invalid partitions. Then derive test values at and around the boundaries of the ordered partitions.
Example: An age field accepts 18 to 65.
- Partitions: below 18, 18 to 65, above 65.
- Two-point BVA values: 17, 18, 65, 66. Add a representative value such as 40 for the valid partition.
This catches both class-level and edge-level defects.

b) Decision Table Testing + Boundary Value Analysis or Equivalence Partitioning
Decision tables handle combinations of conditions. Each condition, for example "order total is at least 100", is itself a partitioned input whose boundaries need testing.
Example: A discount applies when the customer is a member AND the order total is at least 100. The decision table covers the member/non-member and above/below combinations. BVA adds the values 99.99 and 100.00.

c) State Transition Testing + Decision Tables or EP
State models show which events are allowed in each state. Guard conditions on transitions often contain business rules or data ranges. Decision tables or EP and BVA test these guards in depth.

d) Use Case Testing + Black-Box Techniques
Use cases provide end-to-end flows: the basic flow, alternative flows and exception flows. Within each step, EP and BVA are applied to the input data. Decision tables are applied to business rules found in the flow.

e) Classification Tree or Pairwise Testing + EP
Classification trees break inputs into classifications and classes, which are essentially partitions. Pairwise or combinatorial techniques then reduce the number of combinations to a manageable set.

f) Black-Box Techniques + Experience-Based Techniques
Systematic techniques give measurable coverage. Exploratory testing, error guessing, defect-based and checklist-based testing then find what systematic techniques miss, such as usability quirks, odd sequences and historical defect patterns. A typical approach is to run specification-based tests first and then use timeboxed exploratory sessions guided by test charters.

g) Domain Analysis
Domain analysis is itself a combination. It merges equivalence partitioning and boundary value analysis for multiple interacting variables, using ON, OFF, IN and OUT points.

Practical process for a Test Analyst:
1. Analyse the test basis and identify test conditions.
2. Assess risks and decide the required rigour for each area.
3. Choose a primary technique suited to the structure of the requirement:
- Ranges: EP and BVA
- Rules: decision tables or cause-effect graphing
- Workflows and modes: state transition testing
- Scenarios: use case testing
- Many parameters: pairwise testing or classification trees
4. Add complementary techniques to cover the weaknesses of the primary one.
5. Supplement with experience-based techniques, especially for high-risk areas or a weak test basis.
6. Remove redundant test cases. One test case can often satisfy several techniques' coverage items, for example a boundary value inside a decision-table rule.
7. Track coverage per technique for reporting.

4. Key Points to Remember
- No single technique is sufficient. Combinations increase defect-detection effectiveness.
- Techniques are often layered. The output of one, such as partitions, feeds another, such as boundaries.
- Experience-based techniques complement specification-based ones. They do not replace them, especially where coverage must be demonstrated.
- Risk drives how many techniques are used and how rigorously.
- Combining techniques does not mean simply adding up test cases. Overlapping test cases should be merged to stay efficient.
- Regulated domains, for example under standards such as DO-178C or ISO 26262, may mandate specific techniques. These are then combined with others.

Exam Tips: Answering Questions on Combining Test Techniques

Tip 1: Match the technique to the structure of the requirement. Read the scenario carefully and look for clue words:
- Ranges, limits, minimum or maximum point to EP plus BVA.
- If/then, combinations, rules point to a decision table.
- Modes, status, events, sequences point to state transition testing.
- Actors, flows, scenarios point to use case testing.
- Many parameters with many values point to pairwise testing or classification trees.
- Poor or missing documentation, time pressure point to exploratory testing or error guessing.
The best answer usually combines the techniques whose clue words appear.

Tip 2: Prefer answers that complement rather than duplicate. If the options pair two techniques that target the same defect types, look for an option where the second technique covers a weakness of the first. For example, "decision table plus BVA on the numeric conditions" is stronger than "EP plus more EP".

Tip 3: Remember that EP and BVA are the most natural pairing. If a question asks which techniques are most commonly combined, or which technique builds on the result of another, the answer is usually BVA building on EP partitions.

Tip 4: Use risk as the tie-breaker. When a scenario describes a high-risk or safety-critical feature, choose the answer applying more rigorous or multiple systematic techniques, plus experience-based testing. For low-risk areas, a lighter combination such as EP plus exploratory testing may be the best choice.

Tip 5: Watch for the word "best" or "most appropriate". Several options may be technically valid. Pick the one that addresses all the characteristics in the scenario, for example both the business rules and the data ranges, while remaining efficient.

Tip 6: Do not dismiss experience-based techniques. The CTAL-TA syllabus emphasises that experience-based techniques complement systematic ones. Be wary of options claiming that exploratory testing replaces specification-based testing where coverage must be measured or reported. Also be wary of options claiming it is never useful.

Tip 7: Counting questions with combined techniques. In K3 questions you may need to derive the minimum test cases that satisfy two techniques at once. Strategy:
- First list the coverage items for each technique, such as partitions, boundary values and decision rules.
- Then find which test cases can cover several items at once.
- Remember that a single test case can cover a valid partition and a boundary together.
- Invalid values normally need separate test cases, one invalid input per test, to avoid defect masking.

Tip 8: Know the selection factors. Questions may ask which factor most influences technique selection in a scenario. Common correct answers include the risk level, the available test basis, regulatory requirements, the expected defect types and tester skills.

Tip 9: Eliminate absolute statements. Options containing "always", "only" or "never" are often wrong. Examples include "BVA should always be used alone" and "use only one technique per feature to avoid redundancy".

Tip 10: Think like a Test Analyst, not a developer. The CTAL-TA focuses on black-box, experience-based and quality-characteristic testing. White-box combinations such as statement plus decision coverage belong to the Technical Test Analyst syllabus. If an option relies on code coverage in a TA question, it is likely a distractor unless the scenario specifically requests it.

Worked Exam-Style Example
Scenario: An insurance quote system calculates premiums based on driver age (18 to 80), years of driving experience (0 to 60), and whether the driver has had an accident in the past 3 years. Different combinations of these factors trigger different premium rules. The specification is detailed. Which combination of techniques is MOST appropriate?
A) Exploratory testing only
B) State transition testing and use case testing
C) Decision table testing combined with equivalence partitioning and boundary value analysis
D) Pairwise testing only

Answer: C.
- The premium rules depend on combinations of conditions, so a decision table fits.
- Age and experience are numeric ranges, so EP and BVA fit.
- Option A wastes a detailed specification.
- Option B does not fit, because no states or flows are described.
- Option D reduces combinations but ignores boundaries and specific rule outcomes.

Summary
Combining test techniques is about building a complementary toolkit. Choose a primary technique based on the structure of the test basis. Add techniques that cover its weaknesses. Scale rigour according to risk. Round off with experience-based testing. In the exam, identify the clue words in the scenario, choose combinations that complement each other, use risk as the deciding factor, and avoid absolute or single-technique answers unless the scenario clearly justifies them.