Software Testing - Test Design Techniques
- 2. Test Design Technique A procedure used to define test conditions, design test cases, and specify test data. (1) Why is it important? - To cut down number of tests we need to run. ~ Efficiency
- 3. Test Design Technique (Categories) • Specification-based (Black-box) - derive test cases from the specification or a model of a system - testing of what a system should do • Structure-based (White-box) - derive test cases from the code written - testing based on how the specified behaviour was implemented • Experience-based - derive test cases from the tester’s experience of similar systems and general experience of testing
- 4. Test Condition A testable aspect of a component or system identified as a basis for testing. (2) Why is it important? - To create / design good test case Example: • test condition - field can accept up to 20 alphanumeric characters • test case # 1 - blank input • test case #2 - symbol input • test case #3 - 21 alphanumeric characters
- 5. Importance of Test Coverage - quantitative measure of the quality of testing that has been done - way of estimating how much more testing needs to be done Example: “I have tested 75% of the decisions” “I have tested 80% of the requirements”
- 6. (5) Specification-based Techniques • Equivalence Partitioning • Boundary Value Analysis • Decision table testing • State transition testing • Use case testing
- 7. Equivalence Partition • In this technique, you divide the set of test condition into a partition that can be considered the same • Input Partitions - inputs to a program can be ‘chunked’ into groups of similar inputs 2 Types of Input Partition: • Valid Equivalence Partition • Invalid Equivalence Partition
- 8. Equivalence Partition Examples: 1. Valid input: integers in the range 100 to 999 • Valid partition: 100 to 999 inclusive • Invalid partition: less than 100, more than 999, real (decimal) numbers and non-numeric characters 2. Valid input: names with up to 20 alphabetic characters • Valid partition: strings up to 20 alphabetic characters • Non-valid partition: strings of more than 20 alphabetic characters, strings containing non-alphabetic characters
- 9. Equivalence Partition Exercise: Suppose you have a bank account that offers variable interest rates: 0.5% for the first $1,000 credit; 1% for the next $1,000; 1.5% for the rest. If you wanted to check that the bank was handling your account correctly, what valid and invalid input partitions might you use? • Output Partitions - outputs to a program can also be partitioned Example: Exercise above could produce outputs of 0.5%, 1%, and 1.5%
- 10. Equivalence Partition • Other Partitions - partitioning other values Example: A mail-order company charges $2.95 postage for deliveries if the package weighs less than 2 kg, $3.95 if the package weighs 2 kg or more but less than 5 kg, and $5 for packages 5 kg or more. Generate a set of valid test cases using equivalence partitioning. The valid input partitions are: under 2 kg; 2 kg or over but less than 5 kg; and 5 kg or over. Input values could be 1 kg, 3.5 kg, 5 kg. These would produce expected results to $2.95, $3.95, and $5 respectively. In this case there are no invalid inputs (unless the scales fail).
- 11. Boundary Value Analysis • It is the process of testing boundaries between partitions’ of the input values. • Works well with equivalence partitioning because partitions must have boundaries. 2 Types of Boundary Values: • Valid boundary values • Invalid boundary values
- 12. Boundary Value Analysis Boundary Value Analysis Approach: • 2 value boundary approach - most practical - to identify just two (2) values at each boundary • 3 value boundary approach - uses three (3) values at each boundary
- 13. Boundary Value Analysis Example #1: • The boiling point of water - the boundary is at 100 degrees Celcius. • 3 value - boundary values will be 99 degrees, 100 degrees, and 101 degrees; unless you have a very accurate digital thermometer, they could be 99.9 degrees, 100.0 degrees, 100.1 degrees • 2 value - boundary values would be 100 and 101
- 14. Boundary Value Analysis Example #2: • Exam pass - if an exam has a pass boundary at 40%, merit at 60% and distinction at 80% • 3 value - 39, 40, 41 for pass; 59, 60, 61 for merit; 79, 80, 81 for distinction • 2 value - 39 and 40, 59 and 60, and 79 and 80 respectively.
- 15. References • (1) - https://glossary.istqb.org/en/search/test%20design%20tec hnique • (2) - https://glossary.istqb.org/en/search/test%20condition
Editor's Notes
- #3: This is usually applied when there is a time constraint and we need to test smart.
- #4: Specification-based does not need to know what is going on ‘inside the box’ Experience-based applied when specs or structure of implementation is not provided or insufficient Q: Which of these categories do you frequently use for your projects? It’s important to ensure references (test basis) such as functional and technical specs to design test effectively - Using all or each category will ensure good test coverage
- #5: Test condition - derived from the requirement Test case - traceable back to the test condition
- #6: Why is it related to Test Design? It provides a quantitative assessment of the extent and quality of testing. It can be applied to any systematic technique - “How deep is our testing?”
- #7: We use these technique to reduce the number of test cases we need to write
- #8: Input Partition - we only need to test a “representative” of the partition Valid Partition - partitioning the collection of valid inputs Invalid Partition - partitioning the collection of invalid inputs Q: If it’s not included in the specs, should we test it? - There are far more ways to provide incorrect input than there are to provide correct input
- #10: We could use test cases that generate each of these outputs as an alternative to generating input partitions.
- #11: Limit the test cases to the ones that are most likely to crop up in real situation.
- #12: Based on experience, common mistakes that programmers make is that errors tend to cluster around boundaries. Faults coincide with particular programming structures such as the # of times a program loop is executed or the exact point at which a loop should stop executing. Valid boundary values - boundaries on the inside of a valid partition Invalid boundary values - boundaries on the outside of a valid partition; or boundaries on the inside of a invalid partition
- #13: 2 value - boundary value itself and one value opposite side of the boundary 3 value - boundary value itself and the extremes