Subject Area Exam Testing

1. Why is regression testing an important part of any integration testing procedure?

Answer (Section 17.3.2):
he goa! of integration testing is to ma"e sure that independent modu!es that wor" correct!y
on their own do not interfere with one another when added to the same program
(unforeseen side effects are a!ways possi#!e). $egression testing chec"s for defects
propagated to other modu!es #y changes made to an e%isting program.
2. &escri#e o#'ect(oriented unit testing.
Answer (Section 17.).1):
*!ass testing for ++ software is e,ui-a!ent to unit testing for con-entiona! software. he
focus throughout is on designing and testing appropriate se,uences of operations to
e%ercise a!! c!ass states. .t ma"es !itt!e sense to test operations or a!gorithms indi-idua!!y for
c!asses.
3. /ist four types of systems tests.
Answer (Section 17.7):
$eco-ery testing
Security testing
Stress testing
0erformance testing
). What are the attri#utes of a good software test?
Answer (Section 11.1):
2as a high pro#a#i!ity of finding an error
3ot redundant
Shou!d #e capa#!e of unco-ering a who!e c!ass of errors
Shou!d not #e too simp!e or too comp!e%
4. &escri#e the differences #etween #!ac"(#o% testing and white(#o% testing.
Answer (Section 11.2):
5!ac"(#o% testing in-o!-es testing the functiona!ity of a software component without "nowing
the detai!s of its interna! !ogic. White(#o% testing in-o!-es testing the independent !ogic paths
with fu!! imp!ementation "now!edge.
6. What is e,ui-a!ence partitioning as it app!ies to software testing?
Answer (Section 11.6.2):
A #!ac"(#o% testing techni,ue in which the input domain is di-ided into c!asses of e,ui-a!ent
data items. est cases are deri-ed from com#inations of e!ements from each e,ui-a!ence
c!ass. 7%hausti-e testing of a!! input domain -a!ues is not necessary.
7. 2ow does unit testing differ for o#'ect(oriented testing as compared to con-entiona!
software unit testing?
Answer (Section 18.3.1):
9ou can no !onger test a sing!e operation in iso!ation (the con-entiona! -iew of unit testing)
#ut rather: as part of a c!ass. $ather than testing an indi-idua! modu!e: the sma!!est testa#!e
unit is the encapsu!ated c!ass. 5ecause a c!ass can contain a num#er of different operations
and a particu!ar operation may e%ist as part of a num#er of different c!asses: the meaning of
unit testing changes dramatica!!y.
1. What is scenario(#ased testing?
Answer (Section 18.).4):
he user tas"s descri#ed in the use(cases are used to construct the test cases. .t is used to
unco-er errors that occur when actors interact with the software (focus is on user #eha-ior:
not product #eha-ior). ;enera!!y it is #etter to spend more time re-iewing the use cases as
they are created than spending more time on testing.
8. &escri#e the how test cases are deri-ed from #eha-ior mode!s to faci!itate interc!ass
testing?
Answer (Section 18.6.2):
est cases must co-er a!! states in the state transition diagram
5readth first tra-ersa! of the state mode! can #e used (test one transition at a time and
on!y ma"e use of pre-ious!y tested transitions when testing a new transition)
est cases can a!so #e deri-ed to ensure that a!! #eha-iors for the c!ass ha-e #een
ade,uate!y e%ercised
1<. Summari=e the steps used in a We#App testing strategy..
Answer (Section 2<.1.3):
We#App content mode! is re-iewed to unco-er errors.
.nterface mode! is re-iewed to ensure a!! use(cases are accommodated.
&esign mode! for We#App is re-iewed to unco-er na-igation errors.
>ser interface is tested to unco-er presentation errors and?or na-igation
mechanics pro#!ems.
Se!ected functiona! components are unit tested.
3a-igation throughout the architecture is tested.
We#App is imp!emented in a -ariety of different en-ironmenta! configurations and
the compati#i!ity of We#App with each is assessed.
Security tests are conducted.
0erformance tests are conducted.
We#App is tested #y a contro!!ed and monitored group of end(users (!oo"ing for
content errors: na-igation errors: usa#i!ity concerns: compati#i!ity issues:
re!ia#i!ity: and performance).
11. What are the o#'ecti-es for content testing?
Answer (Section 2<.3.1):
>nco-er syntactic errors in a!! media (e.g. typos)
>nco-er semantic errors (e.g. errors in comp!eteness or accuracy)
@ind errors in organi=ation or structure of content presented to end(user
12. What is the difference #etween !oad testing and stress testing?
Answer (Section 2<.8):
/oad testing attempts to determine how the We#App and its ser-er(side en-ironments
respond to -arious !oad conditions. Stress testing is a continuation of !oad testing that see"s
to ha-e the num#er of user: transactions: and data !oad meet and e%ceed operationa! !imits.
13. What are the uni,ue characteristics of c!eanroom software engineering techni,ues?
Answer (Section 21.1):
Aa"e e%tensi-e use of statistica! ,ua!ity contro!
Berify design specification using mathematica!!y(#ased correctness proofs
$e!y hea-i!y on statistica! use testing to unco-er high impact errors
1). /ist the steps re,uired for certification in c!eanroom software engineering.
Answer (Section 21.).2):
>sage scenarios are created
A usage profi!e is specified
est cases are generated from the usage profi!e
ests are e%ecuted: fai!ure data is recorded and ana!y=ed
$e!ia#i!ity is computed and recorded
14. /ist the components of a forma! specification !anguage and descri#e their ro!es.
Answer (Section 21.7):
Synta% that defines the specific notation with which the specification is represented.
Semantics that define the o#'ects used to descri#e the system.
Set of re!ations that define the ru!es that indicate which o#'ects proper!y satisfy the
specification.