Version 2020

AT*SQA Essentials of Software Testing

— Sample Exam Answer Key
65 possible points. 43 required to pass (65%).

Question Answer Rationale Learning Number

Objective of
(LO) Points

1 B B is correct It can be considered fit for purpose if LO-1.1.b 1

all the users will always use it in the tested
environment.

2 A A is correct. Test analysts are the most likely folks LO-1.2.a 1

to work with the users during UAT.

3 A A is correct. A developer is also a tester when LO-1.4.a 1

they are writing and executing unit tests. B, C and
D are not correct because they are still acting as a
developer rather than a tester.

4 B B is correct. This is one of the purposes of a test LO-1.3.a 1

plan. A is the description of a test strategy. C is
the description of a test condition. D is the
description of a test case.

5 C C is correct. User stories are usually used in an LO-1.1.a 1

Agile SDLC. A is not an SDLC. B is not correct as
these are more likely to use formal requirements
documents. D is not correct because this is not a
recognized SDLC.

6 A A is correct. With a project that must pass LO-2.1.a 1

regulatory commission requirements, the
documentation will be critical to its acceptance. All
the others are factors, but not as important as
having the appropriate documentation to present
for approval.

7 D D is correct. A mature team with good skills can LO-2.8.a 1

make any testing approach work and are not
dependent on particular documentation or lifecycle
guidelines to be successful.

8 B B is correct. OAT is conducted by system LO-2.2.b 1

operators who will be responsible for the software
running on the production systems. A is not

Testing Essentials — Sample Exam Answers, Version 2019 2

© Association for Testing and Software Quality Assurance
Question Answer Rationale Learning Number
Objective of
(LO) Points

correct as these testers will do the system testing.

C is not correct as these folks will be involved in
UAT. D is not correct as these are beta testers.

9 B B is correct. Getting early feedback from users or LO-2.3.a 1

potential users is possible with an iterative
environment because they see the product as it is
being developed. A is not correct as this is a
characteristic of the sequential models. C is true,
but it is a disadvantage because of the time and
skill requirements rather than an advantage. D is
characteristic of a sequential model where there is
time with stable requirements before the testers
receive the software.

10 C C is correct. In this case there is not a lot of LO-2.5.a 1

information for the input and there is no
requirement for detailed test documentation, so C
is the best answer. A is not a good approach
because it will tend to move the responsibility for
requirements documentation to the testers. B is
not correct unless a complete UAT will be
conducted and the testers are required to supply
that documentation. Usually process documents
are more suitable as the guideline for UAT. D is
not correct because the completed code can
always serve as a basis for test cases.

11 C C is correct. In this case, you need to save time LO-2.7.a 1

and money and the only way to do that is to test
earlier and help to ensure better quality sooner.
Sharing the environment with development is only
possible if testing is occurring at the same rate as
the development as each release will need to be
tested as soon as it is available. This is a
challenging, but very real situation. A is not correct
because sequential will be slower and more
expensive in the long run. B is not correct because
eliminating UAT could result in a product that
meets the requirements but is not fit for purpose for
the users. C is not correct because this would
result in the majority of defects being found in UAT
which will be the most expensive and most time
consuming approach of all.

12 A A is correct. Decision tables are best suited for LO-3.3.b 1

looking at combinations of conditions and verifying

Testing Essentials — Sample Exam Answers, Version 2019 3

© Association for Testing and Software Quality Assurance
Question Answer Rationale Learning Number
Objective of
(LO) Points

that the combinations are handled properly by the

software. D is not correct because these
combinations should not interact, so there is no
concept of handling combinations correctly as the
combination shouldn’t matter.

13 C C is correct. This is an example of when LO-3.7.a 1

exploratory testing would be most efficient and
would give the quickest feedback.

14 D D is correct. There should be an exposed API that LO-3.6.a 1

will allow you to test the database interactions
without having to test through the UI.

15 A A is correct. Usability can be assessed with LO-3.5.a 1

exploratory sessions, particularly regarding
navigation and clarity of instructions/prompts. B
must be tested in a controlled environment over a
prescribed period of time. C is best determined
with static analysis tools. D is not a non-functional
test area.

16 C C is correct. The boundaries for $100 and $101 LO-3.2.a 1

overlap and only need to be tested once. The
boundaries to be tested are: -1, 0, 1, 99, 100, 101,
102, 499, 500, 501

17 B B is correct. LO-3.3.a 1
See table below

18 C C is correct. A and B are the investments in the LO-4.2.a 1

ROI and C is the expected return. D is not a factor
in the equation.

19 B B is correct. If done correctly, test automation can LO-4.4.b 1

improve the overall efficiency of testing and
thereby reduce costs. A is not necessarily a cost
reducer because those people will have to be
backfilled as the manual testing scope expands. C
is not correct because while low cost tools may
have a lower entry cost, they may have a higher
usage cost over the lifetime of the tool. Just using
low cost tools will not ensure that automation can
reduce overall testing costs. D is not correct
because manual testers are not usually replaced
by automation – their roles are just expanded and
their coverage increased.

Testing Essentials — Sample Exam Answers, Version 2019 4

© Association for Testing and Software Quality Assurance
Question Answer Rationale Learning Number
Objective of
(LO) Points

20 A A is correct. The build acceptance/verification LO-4.6.b 1

tests provide the highest return and visibility for the
automation project. They also help to encourage
better testing from the developers as they know
their software will be rejected if it doesn’t pass
these first tests.

21 C C is correct. Having shared code allows re-use of LO-4.3.b 1

modules across different automation
implementations, thus reducing development and
maintenance efforts. A is not correct because the
shared code is part of the function library. B is not
correct because this doesn’t affect the engineer’s
ability to use new technologies. D is not correct
because one of the advantages of a shared
framework is a standard naming convention that is
used by all automation engineers.

22 B B is correct. Keyword-driven is well suited for this LO-4.3.a 1

as action words can be assigned for the state
transitions and the data can be identified and
stored separately from the scripts. A is not correct
because keyword-driven will be more maintainable
in the frequently changing Agile environment. C is
not correct because this is not a type of test
automation. D is not correct because in-line scripts
have all the data and action within a test script
which will lead to higher maintenance efforts and
costs.

23 D D is correct. Defects that are found earlier are less LO-4.4.a 1

expensive for the team and earlier fixes improve
the efficiency of future development and testing,
thus further reducing the costs. A is not correct
because there may be an initial cost for
implementing the automation, but since the
releases are frequent, the overall cost should be
reduced quickly. Production defects are generally
very costly to fix. B is not correct because it would
not be practical to implement automation for
functional testing rather than regression testing.
Even if that was done, the manual testers would
probably not be able to keep up with frequent
releases to production. C is not correct because
there generally isn’t an efficiency gain from fixing a
large number of defects. Large numbers of fixes

Testing Essentials — Sample Exam Answers, Version 2019 5

© Association for Testing and Software Quality Assurance
Question Answer Rationale Learning Number
Objective of
(LO) Points

tend to result in unstable software which results in

more defects.

24 B B is correct. Automating chaos just results in LO-4.5.a 1

faster chaos.

25 D D is correct. This is a role that is often taken by LO-4.6.a 1

the manual testers. This allows the manual tester
to understand the coverage and effort still required
by the manual testing. A and B are the
responsibilities of the automation engineers. C is
the responsibility of the test manager.

26 A A is correct. If there are no requirements, then the LO-5.3.b 1

targets to be reached cannot be identified and test
results can’t be determined to have passed or
failed. B is not correct. This may be true even if
the requirements are specified. C is not correct
because tuning the tools is not a factor, although
tuning the system may be. D is not correct
because the exit criteria can’t even be defined
without the requirements being defined.

27 A A is correct. This is a major goal in running a LO-5.5.b 1

preliminary test. B is not correct as this should be
a part of data checking rather than environment
checking. Running a single test will not ensure
there is enough data for the entire test. C and D
are not correct as the preliminary test will not prove
this.

28 C C is correct. The performance tests must be run LO-5.6.a 1

again after changes, particularly tuning changes,
are made to the system. A is incorrect because
most of the monitoring tools also log what is
happening. B sometimes happens, but even
transient issues must be diagnosed. D may also
happen, but this isn’t the primary reason for
needing to repeat tests.

29 C C is correct. The concept of “acceptable” LO-5.2.b 1

performance is subjective and varies between
users. A is not correct because specific
performance requirements should be in the overall
requirements for a system. B is not correct
because performance will definitely matter if it’s
slow. D is not correct because performance
corrections in production will be expensive and

Testing Essentials — Sample Exam Answers, Version 2019 6

© Association for Testing and Software Quality Assurance
Question Answer Rationale Learning Number
Objective of
(LO) Points

potentially disruptive.

30 D D is correct. This should occur during acceptance LO-5.2.c 1

testing as part of contractual acceptance and OAT.
A should occur during requirements. B should
occur during design and unit testing. C should
occur during integration testing.

31 C C is correct. The metrics must be meaningful to LO-5.4.a 1

the stakeholders or they will not be useful.

32 B B is correct. The frequency of the activities is not LO-5.4.b 1

known and that is necessary to execute the
performance scripts with the correct volume for
each transaction type.

33 A A is correct. This is one of the important uses of a LO-5.6.b 1

tool and is generally impossible to do without tool
support, particularly for a long period of time.
Because this is an important requirement for
performance testing, it is often used to justify the
cost of the tool. B does not need a tool because
response time for an individual transaction can be
done manually. C is not a goal of performance
testing. D is not true. Performance tools don’t
know what to test and can’t automatically generate
the test scripts.

34 A A is correct. Cybersecurity should be considered LO-6.4.a 1

from the point of design all the way through the
lifecycle of the system.

35 D D is correct. Intrusion detection determines that an LO-6.5.d 1

intrusion has occurred and should activate the
necessary responses and communication to
prevent further damage to the system.

36 A A is correct. Verification verifies that the system LO-6.2.a 1

was built to the requirements (built right) and
validation will verify if the right thing was built
meaning that it can deal with both expected and
unexpected threats.

37 A A is correct. Cybersecurity testing does not just LO-6.3.a 1

expose problems, it also looks at what the system
does to recover when problems have occurred. B
is not correct because that is not the only focus of
the testing. The testing looks at detection,

Testing Essentials — Sample Exam Answers, Version 2019 7

© Association for Testing and Software Quality Assurance
Question Answer Rationale Learning Number
Objective of
(LO) Points

response and recovery. C is not correct because

security testing can occur at the component level.
D is not correct because other types of testing
such as performance testing may also require the
same skills.

38 C C is correct. A port scan is done using a tool that LO-6.4.b 1

examines the system without executing any code
(static testing). The others are all dynamic tests
that are conducted while the system is operating.

39 C C is correct. These are a part of the Detect stage. LO-6.4.c 1

A is Identify. B is Protect. D is Recover.

40 A A is correct. White-box testing is used to defend a LO-6.5.b 1

system against someone who has inside
knowledge of an organization or system and does
not depend on reconnaissance to find out about
system vulnerabilities. B is an example of black-
box testing threat. C and D are not correct.

41 C C is correct. Black-box testing is focused on LO-6.5.c 1

threats that might be available to a person who is
external to a company. This concentrates on what
they can discover without having any insider
knowledge of the system or the organization.

42 A A is correct. User guides should provide LO-7.2.a 1

information regarding the day-to-day activities of a
user which will help the tester to understand what
areas need to be covered in the usability testing. B
is not correct because this information would be
more helpful for functional testing. C is not correct
because error recovery is not the primary focus of
usability testing (although it is sometimes a
consideration). D is not correct because user
guides don’t usually have performance information
in them, unless the performance is remarkably
slow, and this would not particularly help with
usability testing anyway.

43 D D is correct. UX considers the user’s experience LO-7.4.b 1

with the software before, during and after use.

44 D D is correct. Accessibility is the degree to which a LO-7.7.a 1

component or system can be used by people with
the widest range of characteristics and capabilities
to achieve a specific goal in a specified context of

Testing Essentials — Sample Exam Answers, Version 2019 8

© Association for Testing and Software Quality Assurance
Question Answer Rationale Learning Number
Objective of
(LO) Points

use. A, B and C are subsets of the larger set.

45 A A is correct. This includes the quality characteristic LO-7.6.a 1

usability and describes the sub-characteristics of it.

46 C C is correct. This is a description of Operability. LO-7.6.b 1

47 B B is correct. In this case breadth is the goal LO-7.2.b 1

because it will provide the widest possible sample.

48 D D is correct. The goal of giving general tasks is to LO-7.5.b 1

see if the user can figure out how to accomplish
their tasks with little or no instruction.

49 B B is correct. This is a standard used for any LO-7.7.b 1

software developed for the federal government in
the US.

50 B B is correct. The user group for connected devices LO-8.4.a 1

such as smart phones is huge and their
expectations are wildly varying. A is not correct
because users are becoming more savvy
regarding what they want. C is not correct
because connected devices and their applications
do have good flexibility particularly for usability. D
is not correct because users expect a consistent
experience as they move between their connected
devices.

51 A A is correct. The earlier the better with these types LO-8.4.b 1

of requirements.

52 A A is correct. In this case you are testing the LO-8.3.a 1

interaction of the software with the actual
refrigerator and need to be sure the refrigerator
responds correctly. An emulator, if there is one,
will be the most reliable way to test this software.
B is not correct because a simulator will not show if
the hardware is responding correctly. C is not
correct because you’d need a lot of refrigerators
and that’s the same with D (or maybe the cloud
couldn’t support that much weight ☺).

53 D D is correct. There is a wide range of users in the LO-8.4.c 1

connected devices world and many potential
usages and that makes planning and executing
usability testing quite difficult because the usages
may not be known and finding a representative set

Testing Essentials — Sample Exam Answers, Version 2019 9

© Association for Testing and Software Quality Assurance
Question Answer Rationale Learning Number
Objective of
(LO) Points

of users can be tricky. A is not correct because

users may not be able to articulate what they want,
but they do have expectations for usability. B is
not correct because there are many users and
some will undoubtedly be available for the testing.
C is true but that isn’t a factor in making usability
testing difficult.

54 C C is correct. Detailed test cases are not needed, LO-8.5.a 1

but enough information needs to be captured so
the tests are repeatable. A is not correct because
test cases are not the appropriate place to
document requirements, regardless of the
approach. B is not correct because these
techniques do not actually generate the test cases
but rather identify test conditions. D is not correct
because exploratory should not be the only
approach used.

55 C C is correct. Combinatorial techniques such as LO-8.4.f 1

pairwise or classification trees are usually used to
reduce the test set from all possible combinations
to a testable set of combinations.

56 A A is correct. If security isn’t designed into the LO-8.4.e 1

software and tested throughout the lifecycle it will
be expensive, if not impossible, to implement later.
B is not correct. Testing is done in case there are
vulnerabilities that are missed, not because the
developers are incompetent. C is not correct
because security testing is needed throughout
development regardless of the intended use of the
software. No one wants their refrigerator to be
hacked. D is not correct because the testing
needs to occur regardless of what the users
understand about security testing.

57 A A is correct. The biggest challenge is that the LO-8.4.h 1

testing tends to take long periods of time running in
a stable environment, but these long periods of
time are difficult to find in the high speed release
environment of the connected device world. B is
not correct because applications should still be
tested, even if they are unlikely to fail. C is not
correct because the testing can be done, but
generally is more compressed than we would like.
D is true but does not explain why reliability testing

Testing Essentials — Sample Exam Answers, Version 2019 10

© Association for Testing and Software Quality Assurance
Question Answer Rationale Learning Number
Objective of
(LO) Points

is challenging.

58 C C is correct. This is an expected benefit of a LO-9.1.b 1

DevOps project. A and D are not correct. The
same people are needed, but their roles and extent
may change. B is not correct because test
automation is not necessarily easier when built
incrementally – there may be more maintenance
required than would be needed if the automation
was built at the end.

59 B B is correct. Testing during planning means to LO-9.3.a 1

gather accurate requirements and define testable
acceptance criteria for them. A, C, and D are all
good activities and are needed in DevOps, but are
not the goals of testing during planning.

60 C C is correct. Continuous delivery is dependent on LO-9.2.a 1

continuous integration (and continuous testing). A
is not correct because the integration can occur
without deployment. B is not correct because
continuous testing can occur without continuous
integration or deployment. It is, of course, more
effective with the CI/CD components. D is not
correct because deployment can occur without
monitoring.

61 D D is correct. This is the correct order of the stages. LO-9.2.d 1

62 A A is correct. The purpose of infrastructure as code LO-9.4.b 1

is to have a set of scripts that can provision and
manage the test and production environments.
This allows the environments to be completely
rebuilt using scripts that can be modified to include
necessary configuration changes. B is not correct
because this is a part of continuous deployment
but it is helped if IaC is in place. C is the goal of
monitoring tools. D is not correct because this is
not the primary purpose. Although these tools
could be used to modify live production
environments, it’s unlikely that the entire live
production environment would be rebuilt without
impacting users.

63 B B is correct. While performance testing should be LO-9.3.d 1

run throughout the SDLC, the full performance
testing executions occur in the staging

Testing Essentials — Sample Exam Answers, Version 2019 11

© Association for Testing and Software Quality Assurance
 Question Answer Rationale Learning Number
Objective of
(LO) Points

environment which is representative of the

production environment running most of the
software for the release. A and C are not correct
because performance testing can and should be
run throughout the SDLC. D is not correct
because active testing is still in progress at this
time.

64 C C is correct. Test automation tools are vital to the LO-9.4.a 1

implementation of continuous testing. Even a large
number of manual testers would not be able to
keep up with the workload and execute the
required tests fast enough to provide the return the
developers need. A is not correct because this is
an automated process. B is not correct because
well documented regression tests are probably out
of date by the time the documentation is completed
– better to automate than to document. D is not
correct because a DevOps environment requires
testing to start while the software is being
developed and can’t wait for it to stabilize.

65 B B is correct. This set of tools is normally called a LO-9.4.c 1

toolchain.

Table for question 17 above.

Condition

Too wet Y N N N N N

Watering allowed - N Y Y Y Y

Rain predicted tomorrow - - N Y N Y

> 1 day rain predicted in next - - N N Y Y

three days

Results

No water Y Y

Normal water Y

½ normal amount of water Y Y

¼ normal amount of water Y

Testing Essentials — Sample Exam Answers, Version 2019 12

© Association for Testing and Software Quality Assurance