Prev | Current Page 286 | Next

Peter Farrell-Vinay

"Manage Software Testing"

69
NEW
COUNT
26
12
14
0.46
TOTAL
26
20
6
0.77
Test path coverage metric
Testing Processes and Infrastructure 123
7.9.2 Test Effectiveness
Test Effectiveness Ratios (TERs) 1, 2, and 3 have been defined, where:
The three denominators in the above formulae can be identified during static analysis and used as
constants in the equations. The numerators are calculated by an inspection of an execution history file,
generated at run time by the instrumented code. Coverage is maximized by running a number of subject
program executions with suites of stimulating data. TER 1 and 2 normally reach unity without great
effort (although infeasible branches may be discovered), but TER 3 often lags TERs 1 and 2 by some
margin, because TER 3 requires a demanding testing strategy to achieve unity. If however, unity can be
achieved for TER 3, then the number of undetected bugs remaining in the subject program are substantially
reduced.
TER
number of statements exercised at least once
1 =
totalnumber of executable statements
TER
number of branches exercisedat least once
to
2 =
talnumber of branches
TER
number of LCSAJs exercised at least once
tota
3 =
ll number of LCSAJs
125
8
Test Documents
8.1 General
The documents outlined in this book are generated by the four essential processes of software testing:
test planning, specification, execution, and reporting. In this book however, the documents are generalized
to cover the three principal kinds of software tests: System, Integration, and Unit tests.


Pages:
274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298