Is Branch Coverage a Good Measure of Testing Effectiveness? - Empirical Software Engineering and Verification - page 203

1.1

ACTIVE_LIST
ARRAY
ARRAYED_LIST
ARRAYED_SET
ARRAYED_STACK
BINARY_SEARCH_TREE
BINARY_SEARCH_TREE_SET
BINARY_TREE
FIXED_LIST
HASH_TABLE
HEAP_PRIORITY_QUEUE
LINKED_CIRCULAR
LINKED_LIST
PART_SORTED_TWO_WAY_LIST
Median of medians

1

0.9

0.8

0.7

0.6

0.5

0.4

30 60 90 120 150 180 210 240 270 300 330 360

Time (minutes)

Fig. 6. Fault detection similarity for each class over time; their median

0.3

ACTIVE_LIST
ARRAY
ARRAYED_LIST
ARRAYED_SET
ARRAYED_STACK
BINARY_SEARCH_TREE
BINARY_SEARCH_TREE_SET
BINARY_TREE
FIXED_LIST
HASH_TABLE
HEAP_PRIORITY_QUEUE
LINKED_CIRCULAR
LINKED_LIST
PART_SORTED_TWO_WAY_LIST
Median of stdevs

0.25

0.2

0.15

0.1

0.05

0

0 30 60 90 120 150 180 210 240 270 300 330 360

Time (minutes)

Fig. 7. Standard deviation of the fault detection similarity for each class over time; their median

To study the correlation between branch coverage level and fault detection ability,
Figure 8 superimposes the median of the branch coverage level and the median of the
normalized number of faults for the tested classes. In the first few minutes of testing,
when the branch coverage level increases quickly, faults are also found quickly. After
a while, the increase of branch coverage slows down. The speed of fault detection also