Sample Exam

ISTQB Advanced Test Analyst

Answers

Exam Prepared By

Released March 2016

Question Answer Explanation / Rationale Learning Number
Objective of
(LO) Points
1 A A is correct. Planning for the resources should occur TA-1.3.1 1
during the planning stages and should be recorded in
the test plan. B, C and D are too late.

2 B B is correct. This information can be used to monitor TA-1.3.1 1

the project, determine trends and contribute to process
improvement. A and C are both activities that should
occur during the Planning phases of the project. D is a
testing task that the TA performs as part of project
testing.
3 C C is correct. Concrete test cases (with detailed steps TA-1.5.2 3
and defined data) should be used for the inexperienced
testers who will be running the regression tests.
Logical or high-level test cases should be used for UAT
because this will allow the business users to vary the
steps and data within the tests but will still provide
documented test cases.

4 B B is correct. The exit criteria determine when testing is TA-1.8.1 1

done. A is not correct because the exit criteria don’t
define priorities, just goals to be met. C is not correct
because the scope is defined by the exit criteria. D is
not correct because human analysis is still needed.
5 C C is correct. 9 of the 10 test cases have been TA-1.8.1 1
executed. Execution does not mean they have passed.
The only test excluded is the one that is still in
progress.
6 D D is correct. The 25 test cases that “passed with TA-1.8.1 1
exception” may be harboring significant defects even
though the test case functionality has “passed”. These
test execution results will need to be researched to
determine if the test cases should be considered
“passed” or “failed” for the exit criteria. A is not correct
because some number of the “passed with exception”
cases should probably be classified as “passed”. B is
not correct because some number of the “passed with
exception” cases should probably be classified as
“failed”. C is not correct, although it’s a really good
point and risk should also be a factor in evaluating exit
criteria. Risk assessment was not stated in the exit
criteria though, so it is not a factor with this answer.
7 C C is correct. This is an example of the software doing TA-1.7.1 2
something that is not specified in the requirements and
should be documented as an issue. A is not correct
because the tester can’t assume that the feature should
be there if it’s not specified. B is not correct because

2
Question Answer Explanation / Rationale Learning Number
Objective of
(LO) Points
the test case is correct per the requirements. D is not
correct because this should be documented and
investigated to see if it’s a defect, not a note in the test
case.
8 A A is correct. Agile models require the earliest TA-1.2.1 1
involvement from the TA. B requires activity from the
TA as soon as the requirements documents are ready
for review. C is an iterative model within a V-model, so
it works the same as B for the moment of involvement.
D is usually the latest involvement, often not happening
until the code is written.
9 C C is correct. Risk and priority assignment occur during TA-1.6.1 1
test design as the test cases are designed to test
particular test conditions. The test cases are ordered
for execution, often based on priority/risk, during the
Test Implementation step.
10 A A is correct. This information is often turned over to the TA-1.9.1 1
support and maintenance groups to help them with
understanding the known defects and workarounds for
those defects. B and C should be input to the Analysis
activities. D should be an output from the Analysis
activities and an input to the Design activities.
11 B B is correct. This risk needs to be raised to the Test TA-2.4.1 2
Manager because this is potentially a serious issue.
You should also work on mitigation approaches,
probably with the Test Manager. A is not correct
because if anyone pushes back on the CCB, it should
be the Test Manager. C is not correct because you’re
in the final cycle of testing so you won’t have any
benefit from re-prioritizing tests you’ve already run. D
is not correct because this would result in not meeting
the schedule and that may not be acceptable.
12 C C is correct. This is breadth-first because it is testing TA-2.4.1 2
across all areas while looking at the priorities within
each area. A is depth-first (pure priority order). B is
reverse priority order. D covers breadth, but doesn’t
consider priorities within the areas (U2 should come
before U1).
13 B B is a good communication practice that utilizes tools to TA-2.3.1 1
ensure everyone has access to the information they
need. A is not a good practice because it often leaves
the offshore team with software or environments that
aren’t working. C is not a good practice because
timezone differences may make conference calls
inefficient and it is not a good use of tools. D is not
correct because this won’t facilitate communication.
14 D D is correct. There are 5 partitions you need to test. TA-3.2.2 2
Invalid too low, $1 - $100, $101 - $500, >$500 and
invalid too high (if possible).
15 D D is correct. This provides one test for the invalid low TA-3.2.2 2

3
Question Answer Explanation / Rationale Learning Number
Objective of
(LO) Points
partition (too few characters), one for the valid partition,
and one for the invalid high partition (too many
characters). Note, the question specifies “characters”
so any alphanumeric or special characters could be
used in the answer. A has two tests in the valid
partition. B tests the boundaries, but has two tests in
the valid partition. C has two tests in the invalid too low
partition (too few characters). The negative value
doesn’t matter.
16 C C is correct. The only equivalence partition is for three TA-3.2.2 2
day combinations with two or more rainy days. There
are 7 of these possible (rainy, rainy, sunny/cloudy;
rainy, sunny/cloudy, rainy; sunny/cloudy, rainy, rainy;
rainy, rainy, rainy). All other combinations must be
tested. There are 27 total combinations. Of these, 7
are the same so only one has to be tested. As a result
you need 20 test cases for the other combinations and
1 test case for the 2+ rainy days). Within these 21
tests, only one will have no water, so the 5
combinations of spray types can be tested within the 21
tests for the weather combinations that result in water.
17 D D is correct. This covers all the boundaries with two- TA-3.2.3 2
value coverage. The boundaries to consider are 0, 1,
100, 500, and max spend (which might push the max
integer value and cause corruption).
18 C C is correct. You will need to test 74%, 75%, 76%, TA-3.2.3 2
84%, 85%, 86%, 94%, 95%, 96% to achieve 100%
three-value boundary coverage.
19 B B is correct. This provides one test for the value just TA-3.2.3 2
over the lower boundary (2 characters), one for the
lower boundary (3 characters), one for the upper
boundary (10 characters) and one for just over the
upper boundary (11 characters). Note, the question
specifies “characters” so any alphanumeric or special
characters could be used in the answer. A has two
tests in the valid partition. C doesn’t address the lower
boundaries at all and just includes the value over the
upper boundary. D includes the values outside the
boundaries, but not the boundaries themselves.
20 C C is correct. There are 16 columns required in the full TA-3.2.4 2
decision table. (see table A).

The collapsed decision table requires 6 columns. (see

table B). Please note, the use of the tildas indicates
that any value could be there – it doesn’t affect the test.
Also remember there are multiple ways to collapse a
table so if different columns were collapsed, that can be
right too as long as the logic is followed. Testers
always make assumptions on collapsing a table.

When reading the collapsed table:

4
Question Answer Explanation / Rationale Learning Number
Objective of
(LO) Points
1 is not needed, can’t take both credit and debit
2 is not needed, can’t take both credit and debit
3 is not needed, can’t take both credit and debit
4 is not needed, can’t take both credit and debit
5 is not needed, credit cards don’t require a PIN
6 is needed, covers valid credit card, insufficient funds
7 is needed, covers valid credit card, sufficient funds
8 is not needed, covered by 6
9 is needed, covers valid debit card, sufficient funds
10 is needed, covers valid debit card, insufficient funds
11 is needed, covers valid debit card, invalid PIN, funds
don’t matter
12 is not needed, covered by 11
13, 14, 15 are all covered by 16
16 covers both cards invalid

21 D D is correct. You have to check that the length is valid TA-3.2.4 2

(3-10) and you need to know if it’s >8. You also need
to check the characters in the password. A is not
correct because it doesn’t check the characters and
doesn’t check for a password that is 9 or 10 characters.
B is not correct because it can’t differentiate between a
password that is too long and one that is valid but more
than 8 characters. C is not correct because it is mixing
conditions and results.

22 B B is correct. The errors that can occur are either that TA-3.2.4 2
the password is the wrong length or that they didn’t
enter the right mix of characters. The proper symbol
should be displayed depending on the combination of
input characters. A symbol is only displayed if the
password is valid, so no special result of valid
password is needed. A is not correct because it
doesn’t define the reason for the invalid password and
doesn’t consider the symbols. C is not correct because
these should be Boolean values so there is no need for
valid length and invalid length. A “no” on one will
always be a “yes” on the other. It also doesn’t makes
sense to indicate which characters are missing since
that information as already supplied in the conditions –
it’s not the outcome of the decision. D is not correct
because the length is a condition, not a decision result.
It is also not a good practice to use a non-boolean for
the symbol displayed.
23 C C is correct. These are the correct labels for the TA-3.2.5 2
columns on a state transition table.
24 C C is correct. 8 test cases are required, one for each TA-3.2.5 2
row in the table. Since this is 0-switch coverage, only
one transition is being tested, one event to move from a
start state to an end state.

5
Question Answer Explanation / Rationale Learning Number
Objective of
(LO) Points
25 B B is correct. It shows both state transitions, one for a TA-3.2.5 2
valid combination of username/password, and one for
an invalid combination. A is not correct because it
doesn’t show the transition for an invalid combination.
C is not correct because the question states that it just
wants the transitions from the Login state. D is not
correct for the same reason and it duplicates the first
transition. A state transition table should not have
duplicate rows.
26 B B is correct. 12 test cases are needed. The longest list TA-3.2.6 2
(4) multiplied by the second longest (3) gives the
answer needed. Table C shows the table that would be
generated from a full pairwise expansion.

27 B B is correct per the following table: TA-3.2.6 2

1 1 1
2 1 2
3 1 1
1 2 2
2 2 1
3 2 2

Pairwise coverage means that each pairing of options

is covered. In this case, the first column represents the
3 options for color. In column 1 and 2, all combinations
of those two values are created. In columns 2 and 3,
all combinations of brand and slot are covered. In
columns 1 and 3, all combinations of color and slot are
covered.
28 A A is correct. Singleton coverage means that each item TA-3.2.6 2
must be represented at least once. By using the
number for the largest set of options, you are
guaranteed that everything can be used once.

29 C C is correct. Classification trees provide a graphical TA-3.2.7 2

representation of the combinations and allow the user
to see how the test values are derived. A and D are
not graphical techniques. B is not real.
30 B B is correct. EP is often used to limit the options to TA-3.2.7 2
create a reasonable number of nodes for the tree from
all possible options, which could be very large.
31 A A is correct. Pairwise coverage will be achieved if each TA-3.2.7 2
of the 9 test cases are run. Singleton coverage will
also be achieved because each item will be tested at
least once.
32 C C is correct. Alternate paths should be tested TA-3.2.8 2
individually if possible because this limits the risk of
masking a defect. A is not correct because 100%
coverage has not been achieved because the alternate
paths were tested together. B is not correct because

6
Question Answer Explanation / Rationale Learning Number
Objective of
(LO) Points
step 7 can’t occur after step 6. D is not correct
because test case 2 does cover all the alternate paths,
but does them in one test case.
33 C C is correct. A minimum of 5 cases are needed, one TA-3.2.8 2
for the main path and one for each alternate path (1a,
3a, 4a, 6a). In reality, a good tester would have many
more test cases than 5. You could test several of the
alternate paths together, but then there is the risk of
masking a failure. It’s safer, when going for minimum
coverage, to test each alternate path separately. For
example, testing 1, 2, 3a, 3b, 3c, 4, 5, 6, 7.
34 A A is correct. 5 tests are needed for minimum coverage TA-3.2.8 2
and only 2 of those have been executed.
35 C C is correct. The problem seems to be occurring with TA-3.2.11 3
3 transitions, which will be tested with 2-switch testing.
D is not correct because invalid transitions don’t seem
to be the problem. The problem seems to be with not
allowing valid transitions.
36 B B is correct. BVA will be needed to ensure all the TA-3.2.11 3
monetary transactions are handled properly,
particularly with regards to rounding. A classification
trees are applicable for non-interacting conditions. This
question is dealing with determining accurate amounts
based on rounding and transfer information. C is not
correct because there is no indication that states are
used in the wire transfers. D is not correct because
there are no business rules cited in the question that
would indicate that decision tables would be needed.
37 C C is correct. Decision tables would be the best TA-3.2.11 3
approach because there are numerous rules
embedded in the software. The other techniques may
be useful for other parts of the software, but the
example shown in the question would be best tested
with decision tables.
38 C C is correct. Experience-based techniques can be TA-3.4.1 1
used to fill in the gaps of incomplete specifications.
Exploratory testing is frequently used in this capacity.
A is not correct because experience-based techniques
should be used by experienced testers. B is not correct
because these techniques work well to fill in the gaps of
other techniques that are applied when there are good
requirements. D is not correct. Experience-based
techniques tend to give poor coverage and traceability
information.
39 A A is the most correct charter. It explains what areas to TA-3.4.2 2
test and then specifies the valid and invalid card usage.
This test session will result in some coverage
information and could support traceability back to the
requirements for credit and debit card handling. B and
D are too vague to be useful. C doesn’t really make
sense because removing items after you buy them may

7
Question Answer Explanation / Rationale Learning Number
Objective of
(LO) Points
not be allowed and it specifies steps rather than the
purpose of the testing.
40 D D is correct. Checklist testing would be the most TA-3.4.3 3
suitable because you already have the outline of what
you want to test and that can easily be made into a
checklist for these experienced users who will be doing
the testing. A is not correct because there is nothing in
the question about known error patterns or scenarios.
B is not correct because the testing will use the existing
test cases for guidelines allowing the user to determine
the data and steps, but within the confines of the
existing test cases. C is not correct because there are
no use cases but if there were they would have the
steps defined.
41 A A is the best choice here because you are concerned TA-3.4.3 3
with the boundaries of the allowed purchase limits and
it’s likely that equivalence partitions may be helpful as
well. Domain analysis gives you both BVA and EP. B
would not be helpful because there are no states being
transitioned here. C is not the best choice. Although
decisions are being made, concentration around the
particular values is needed and that will be covered
better by A. D is not correct because, while error
guessing might be useful, it is not the best technique.
42 B B is correct. Given that there are a large number of TA-3.4.3 3
known defect areas, it makes sense to create a
taxonomy and conduct defect-based testing. A would
make sense if the concentration were on known failure
areas, but it’s concentrating on heavy usage areas,
which may or may not be high risk. C is not correct for
the primary testing, although it might be useful for
regression testing. D is not correct because that is not
what pairwise testing does.
43 A A is correct. 5 test cases should pass and 10 should TA-3.2.10 2
fail as shown in table D below.
44 A A is correct. A cause-effect graph is a graphical TA-3.2.1 1
presentation of decision information that can be
contained in a decision table.
45 D D is correct. The acceptance criteria are used to TA-3.2.9 1
measure coverage of a story. Of course, the accuracy
of this method depends on how well the acceptance
criteria have been defined.
46 D D is correct. With interoperability testing it is likely that TA-4.2.1 1
many combinations of environments will need to be
tested and these should not influence the behavior of
the software in an adverse way. Combinatorial testing
is well suited for this. A might be used, but would likely
only be used as a final check, not a primary technique.
B is not applicable since it concentrates more on EP
and BVA combinations. C may be useful, but only if
there is a taxonomy available from prior interoperability

8
Question Answer Explanation / Rationale Learning Number
Objective of
(LO) Points
testing.
47 A A is correct. Given that this is your first day, you first TA-4.2.4 3
need to figure out what the software should do. After
you know that, B would make sense to clarify what the
customer is expecting. C and D will take too long.
48 D D is correct. 1 is a test for syntax (what can be entered TA-4.2.4 3
in an input field). 5 is a test for semantics (reasonable
and meaningful error message). 2 is testing navigation.
3 is testing learnability. 4 is testing accessibility.
49 D D is correct. A survey would be the best way to get this TA-4.2.4 3
information since it is deployed and there are a number
of users out there who will respond. A is not correct
because this should have been done when the
interface was being designed. B is not correct because
the lab is quite restricted in the number of participants
and D gives a wider set of inputs from real users. C is
not correct because there is no indication that the
customers who are not interested might have
accessibility issues. Also, the abandonment rate is not
high, indicating that once people figure out the site,
they stay there.
50 A A is correct. When you leave the requirements review TA-5.1.1 1
meeting you should have enough information to
support the test effort. B is not correct and should
happen during the review or with another scheduled
review. C and D are information you would expect to
get from a design review, not a requirements review.
51 B B is correct. No alternate flows are shown in this set of TA-5.2.1 3
steps. And you may need to wipe up the counter given
the order of steps 3 and 4! A is not correct because
this is the main flow. C is not correct because this
would only be applicable to a user story. D is not
correct because accessibility requirements are not
normally shown at the use case level.
52 D D is correct. It’s likely that the error messages weren’t TA-5.2.1 3
defined and either weren’t implemented correctly or are
misleading. A is not correct because the question stem
says they were defined. B and C should not be
significant for a mobile application.
53 A A is correct. A user story should only contain one item TA-5.2.2 3
of functionality. Save and print are two items and
should be recorded in two separate stories. B and C
are applicable to use cases not use stories. D is not a
problem because there appear to be no dependencies.
54 C C is correct. The story is too big for a two-week sprint. TA-5.2.2 3
The acceptance criteria are not all defined (for
example, is there a limit to the number of points? Can
you split the points between purchases?) and are not
all testable (usable? Efficient?). The functionality is not
clearly defined and the story contains multiple items of

9
Question Answer Explanation / Rationale Learning Number
Objective of
(LO) Points
functionality (registration as well as using the points). 5
is done correctly, there is a priority on the story. 4 is
unknown, but since it is not defined, one would
probably conclude there aren’t any dependencies (of
course, that could be dangerous…).
55 C C is correct. The cost of a defect is minimized when TA-6.2.1 1
the defect is found in the same phase in which it was
introduced. This is perfect phase containment.
56 C C is correct. The purpose of root cause analysis is to TA-6.5.1 1
find common causes of problems and eliminate those
problems, thus improving the overall process. A is not
correct because although you may find the defects
earlier because you know what to look for, this isn’t the
purpose of RCA. B is definitely not correct because we
are not trying to target the developer but rather look for
patterns. D is not correct because root cause analysis
tells us what causes problems, not where they are.
57 A A is correct. These are all false positives, meaning that TA-6.4.1 3
a problem is reported when there isn’t one. This is a
tester error as the tester should have done further
investigation before reporting the problem. B is not
correct because there is not a problem with the code
logic. C and D might be problems, but the tester
should be checking for those before documenting the
defect.
58 B B is correct. Since the requirements were clearly TA-6.4.1 3
stated, it’s not a requirements issue. It’s likely that
design decisions were made regarding the type of
implementation that are causing the software to behave
incorrectly on the smaller screens. If this were a
problem with just a particular browser, it might be a
coding issue, but since the problem is more wide
spread, it’s likely a design issue.
59 C C is correct. This is a data handling problem. It might TA-6.4.1 3
have been due to a missing requirement, but incorrect
data handling is a more clear classification and
immediately indicates that further testing is needed with
a wider variety of data.
60 B B is correct. This is the job of the TA because they TA-7.2.2 1
know the domain and so can supply accurate data and
keywords for use with the automation. A is true for
data-driven. C is the job of the TTA. D is not correct
because the TA has an active role in defining the input
parameters.

10
#20, Table A:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Valid Y Y Y Y Y Y Y Y N N N N N N N N
credit
Valid Y Y Y Y N N N N Y Y Y Y N N N N
debit
Valid Y Y N N Y Y N N Y Y N N Y Y N N
PIN
Funds Y N Y N Y N Y N Y N Y N Y N Y N
OK

Approve Y N Y N Y N Y N Y N N N N N N N

#20, Table B:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Valid Y Y Y Y Y Y Y Y N N N N N N N N
credit
Valid Y Y Y Y N N N N Y Y Y Y N N N N
debit
Valid Y Y N N Y ~ ~ N Y Y N N Y Y N ~
PIN
Funds Y N Y N Y N Y N Y N ~ N Y N Y ~
OK

Approve Y N Y N Y N Y N Y N N N N N N N

#26, Table C:

Manufacturer Buy/Sell/Trade Facility Credit/Debit

N B 1 C
N S 2 D
N T 3 C
TM B 3 D
TM S 2 C
TM T 1 D
C B 2 C
C S 3 D
C T 1 C
T B 3 D
T S 1 C
T T 2 D

11
#43, Table D:

Var Cond 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Food F >=0 On 0
Off -1
Out -2
F <=20 On 20
Off 21
Out 22
Typical In 10 10 10 10 10 10 10 10 10
Toy T >= 0 On 0
Off -1
Out -5
T <=5 On 5
Off 6
Out 10
Typical In 0 0 0 2 3 4 5 6 7
F/T F-T>=5 On 5
Off 4
Out 3
Typical In 18 18 18 10 11 15 5 4 0
Result P F F P F F P F F P F F P F F

12