6A: test data

Question 1

what is test data made up of

Answer 1

sets of inputs, put into algorithms/coded solutions to check processing + outputs

Question 2

main purpose of test data

Answer 2

makes sure all processes are checked + verified
- used efficiently when outputs are available/calculated

Question 3

different kinds of testing w/ test data

Answer 3

1. statement coverage testing: every statement executed to find unused code
2. decision coverage testing: every boundary condition tested
3. path coverage testing: checking every path to make sure no errors

Question 4

how are test data sets created

Answer 4

customized to the solution

Question 5

testing all pathways through algorithm/code

Answer 5

known as path coverage testing, requires unique set for each possible path

Question 6

testing algorithms using test data

Answer 6

compromises made:
- reduced number of test data sets
- subroutines tested independently

Question 7

two types of algorithm testing

Answer 7

1. bottom-up testing: tests lower-level subroutines first, using driver routines
2. top-down testing: tests main program/higher-level subroutines first, using stubs

Question 8

extra conditions = ??

Answer 8

increased number of test data sets
- CASE tools assist w/ this for coded solutions

Question 9

testing boundary conditions

Answer 9

selection + repetition use results of conditions to determine actions
- if conditions incorrect, major errors occur

Question 10

decision coverage testing

Answer 10

each condition checked in all conditions
- if too many conditions, some can be isolated + tested separately

Question 11

process of using test data

Answer 11

1. inconsistencies found when comparing test data results + expected results further investigated
2. errors corrected, tests continued
   - in large projects: test report from specialists sent to developers

Question 12

desk checking

Answer 12

process of evaluating each statement by hand and recording results in table
- fluctuating values of identifiers recorded
- if too many changes in a repetition, first and last ones usually prioritized

Question 13

special consideration for data types

Answer 13

arrays: column for array index to determine element being accessed
records: separate column for each field relevant to occurring processing

Question 14

stepping through a coded solution

Answer 14

• automated desk check
• function provided by IDEs
• usually possible to change value of identifiers + check code before execution is recommenced