An expression that appears in a conditional statement such as if, while, do, for, etc. At run time, the variables in conditional expression are substituted with actual values, and the expression evaluates to either true or false. An example of a conditional expression is (a>1&&b!=NULL).
Sequence of nodes in a flow graph. A sequence of program statements such as the statements are connected with branches.
An edge in a flow graph. A true branch is taken when the conditional expression in the conditional statement evaluates to true.
Conditional Diversity Vector
A vector containing conditional diversities <c1,…,cn>, where each ci is associated with a particular conditional expression in the source code. Each ci is the value of the conditional diversity corresponding to a conditional expression ei in the code .
Conditional Diversity Matrix
A matrix containing conditional diversity vectors <v1,…,vk>, where each vi corresponds to a particular execution of a test case.
Multi Branch Statement
A statement that has more than two possible outcomes. An example is a C switch statement, or a VB select statement.
Conditional Diversity Mean Value
An arithmetic average of the vectors in the diversity matrix. Each value mi in the mean vector <m1,…,mk> is the arithmetic mean of the values ci in the vectors <v1,…,vk>. This is the usual statistics textbook definition of mean.
Standard Deviation Vector
A vector containing standard deviations <s1,..,sk>, where each si is the standard deviation of the values ci in the vectors <v1,…,vk>. That is, each si is the arithmetic mean of the squares of ci in the vectors <v1,…,vk> minus the square of mi.
The expected value of the square of a a random variable minus the square of the arithmetic mean. This is the usual statistics textbook definition.
A test measurement technique aimed at measuring the percentage of covered (executed) source code. Examples are statement, branch, and path coverage.
Covering control statement in the code.
Data Diversity Vector
A vector containing data diversity values <d1,..,dk>, where each di is the percentage of distinct ci values in <v1,…,vk>.
Test Case Independence
Test cases are independent if the failure/success of one does not influence the failure/success of the others. The joint probability of failure of the test cases is the same as the product of their individual failures rates.
Execution probabilities associated with each input for a program. Indicates the expected frequency of program feature usage in the field.