McGill PSYC 406 Psychological Tests Exam 2 - Questions with
Correct Answers (Verified Update)
50 questions
1. Why is a single test score an insufficient basis for a clinical diagnosis?
A. Tests are never reliable
B. Diagnoses must be based only on interviews
C. Scores contain error and must be integrated with history, interview and observation
D. Scores cannot be compared to norms
Correct Answer: C. Scores contain error and must be integrated with history, interview and
observation
Rationale: Good assessment uses multiple methods and sources.
2. A psychologist administers a test in a noisy room and allows extra time not specified in the
manual. The resulting scores are problematic mainly because:
A. The test has low face validity
B. The test was not criterion-referenced
C. The SEM becomes zero
D. Standardized administration was violated, so norms no longer apply
Correct Answer: D. Standardized administration was violated, so norms no longer apply
Rationale: Norms assume the standard conditions used with the norm group.
3. Which ethical issue arises when test results are sent to an employer without the client's
consent?
A. Lack of face validity
B. Test bias
C. Breach of confidentiality
D. Restriction of range
Correct Answer: C. Breach of confidentiality
Rationale: Disclosure requires consent except in legally defined circumstances.
4. If every score on a test is increased by 10 points, the standard deviation:
A. Increases by 10
B. Doubles
C. Decreases by 10
D. Stays the same
Correct Answer: D. Stays the same
Rationale: Adding a constant shifts the mean but not the spread.
5. If every score on a test is multiplied by 2, the standard deviation:
A. Stays the same
B. Is halved
C. Increases by 2
D. Doubles
Correct Answer: D. Doubles
Rationale: Multiplying by a constant multiplies the SD by that constant.
, 6. Why are standard scores (z, T) preferred to raw scores when comparing performance on two
different tests?
A. They are always higher
B. They remove all error
C. They place scores on a common scale of standing relative to each test's norm group
D. They make tests more valid
Correct Answer: C. They place scores on a common scale of standing relative to each test's
norm group
Rationale: Comparing raw scores across tests with different means and SDs is meaningless.
7. Jamal scores 70 on a test (M = 60, SD = 10) and 80 on another test (M = 75, SD = 10). On which
did he do relatively better?
A. The first test (z = 1.0 vs. 0.5)
B. The second test
C. Equally well
D. Cannot be compared
Correct Answer: A. The first test (z = 1.0 vs. 0.5)
Rationale: z1 = 1.0; z2 = 0.5.
8. The correlation between height and shoe size is high, but buying bigger shoes will not make
you taller. This illustrates that:
A. Correlation does not imply causation
B. The correlation is invalid
C. Height has low reliability
D. Shoe size is nominal
Correct Answer: A. Correlation does not imply causation
Rationale: Third variables or reverse causation can explain correlations.
9. A test standardized on a sample that underrepresented rural students is used to classify rural
students. The most likely problem is:
A. The test will become more reliable
B. Scores may be interpreted against norms that do not represent them
C. Item difficulty will be zero
D. Face validity will increase
Correct Answer: B. Scores may be interpreted against norms that do not represent them
Rationale: Unrepresentative norms can systematically misclassify a group.
10. Why do some tests report age-based norms rather than one overall norm table?
A. Age norms are more reliable by definition
B. Performance changes developmentally, so people should be compared with same-age peers
C. Age norms remove error
D. It is required for criterion-referenced tests
Correct Answer: B. Performance changes developmentally, so people should be compared
with same-age peers
Rationale: Raw scores mean different things at different ages.
11. In classical test theory, reliability can be defined as the ratio of:
A. Error variance to true-score variance
B. Observed variance to error variance
C. The mean to the SD
D. True-score variance to observed-score variance
Correct Answer: D. True-score variance to observed-score variance
Rationale: r(xx) = var(T) / var(X).
Correct Answers (Verified Update)
50 questions
1. Why is a single test score an insufficient basis for a clinical diagnosis?
A. Tests are never reliable
B. Diagnoses must be based only on interviews
C. Scores contain error and must be integrated with history, interview and observation
D. Scores cannot be compared to norms
Correct Answer: C. Scores contain error and must be integrated with history, interview and
observation
Rationale: Good assessment uses multiple methods and sources.
2. A psychologist administers a test in a noisy room and allows extra time not specified in the
manual. The resulting scores are problematic mainly because:
A. The test has low face validity
B. The test was not criterion-referenced
C. The SEM becomes zero
D. Standardized administration was violated, so norms no longer apply
Correct Answer: D. Standardized administration was violated, so norms no longer apply
Rationale: Norms assume the standard conditions used with the norm group.
3. Which ethical issue arises when test results are sent to an employer without the client's
consent?
A. Lack of face validity
B. Test bias
C. Breach of confidentiality
D. Restriction of range
Correct Answer: C. Breach of confidentiality
Rationale: Disclosure requires consent except in legally defined circumstances.
4. If every score on a test is increased by 10 points, the standard deviation:
A. Increases by 10
B. Doubles
C. Decreases by 10
D. Stays the same
Correct Answer: D. Stays the same
Rationale: Adding a constant shifts the mean but not the spread.
5. If every score on a test is multiplied by 2, the standard deviation:
A. Stays the same
B. Is halved
C. Increases by 2
D. Doubles
Correct Answer: D. Doubles
Rationale: Multiplying by a constant multiplies the SD by that constant.
, 6. Why are standard scores (z, T) preferred to raw scores when comparing performance on two
different tests?
A. They are always higher
B. They remove all error
C. They place scores on a common scale of standing relative to each test's norm group
D. They make tests more valid
Correct Answer: C. They place scores on a common scale of standing relative to each test's
norm group
Rationale: Comparing raw scores across tests with different means and SDs is meaningless.
7. Jamal scores 70 on a test (M = 60, SD = 10) and 80 on another test (M = 75, SD = 10). On which
did he do relatively better?
A. The first test (z = 1.0 vs. 0.5)
B. The second test
C. Equally well
D. Cannot be compared
Correct Answer: A. The first test (z = 1.0 vs. 0.5)
Rationale: z1 = 1.0; z2 = 0.5.
8. The correlation between height and shoe size is high, but buying bigger shoes will not make
you taller. This illustrates that:
A. Correlation does not imply causation
B. The correlation is invalid
C. Height has low reliability
D. Shoe size is nominal
Correct Answer: A. Correlation does not imply causation
Rationale: Third variables or reverse causation can explain correlations.
9. A test standardized on a sample that underrepresented rural students is used to classify rural
students. The most likely problem is:
A. The test will become more reliable
B. Scores may be interpreted against norms that do not represent them
C. Item difficulty will be zero
D. Face validity will increase
Correct Answer: B. Scores may be interpreted against norms that do not represent them
Rationale: Unrepresentative norms can systematically misclassify a group.
10. Why do some tests report age-based norms rather than one overall norm table?
A. Age norms are more reliable by definition
B. Performance changes developmentally, so people should be compared with same-age peers
C. Age norms remove error
D. It is required for criterion-referenced tests
Correct Answer: B. Performance changes developmentally, so people should be compared
with same-age peers
Rationale: Raw scores mean different things at different ages.
11. In classical test theory, reliability can be defined as the ratio of:
A. Error variance to true-score variance
B. Observed variance to error variance
C. The mean to the SD
D. True-score variance to observed-score variance
Correct Answer: D. True-score variance to observed-score variance
Rationale: r(xx) = var(T) / var(X).
