Deck 4: Identifying Reliable and Valid Predictors of Performance

Question 1

True score is equal to obtained score divided by error.

Answer 1

False

Question 2

Establishing construct validity is a difficult task and often beyond the capabilities of a typical HR manager.

Answer 2

True

Question 3

Domain sampling requires knowledge of the KSAs.

Answer 3

True

Question 4

To establish the reliability coefficient of a new employment test,first the proportion of total variance accounted by systematic variance has to be measured.

Question 5

Synthetic validity is best thought of as a rational approach to validation.

Question 6

A clock that is always five minutes fast is reliable.

Question 7

Reliability of a test establishes the lower bound of its validity.

Question 8

Cross validation is necessary only when the employment test in question is assessing attitudes and not potential performance.

Question 9

The spread around the true score of an unreliable test is:

A)lower
B)higher
C)the same as that of a reliable test
D)higher or lower depending on the type of test
E)none of the above is true of an unreliable test

Question 10

Concurrent validation,in many instances,is more difficult than predictive validation.

Question 11

A problem with test-retest reliability coefficient is the effect of memory on scores.

Question 12

Spearman-Brown correction is most useful for checking the validity of an employment test.

Question 13

If the length of an employment test increases,in general,its reliability increases.

Question 14

There are two types of measurement error.

Question 15

Meta-analysis is widely used in validity generalization studies.

Question 16

If a sales aptitude test containing 20 questions was split into two parts containing 10 items each and the scores of each group were correlated with the other,this will give us the convergent validity estimate of the test.

Question 17

The coefficient of equivalence is the correlation between two tests that are equally valid.

Question 18

Measurement error is the difference between the ideal score of a phenomenon and the operational measure of the same.

Question 19

Reliability can be determined by comparing the consistency between two sets of scores on a measure.

Question 20

If the items in an employment test are either very difficult or very easy,the score on the test is likely to be less reliable than when the test items are moderately difficult.

Question 21

John always gets nervous during exams,which results in him consistently receiving 20 fewer points on formal exams than on assignments done under identical conditions at home.The difference between his actual test score and true score (that is,the score he would have received normally)can be considered what sort of error?

A)Random
B)Systematic
C)Unpredictable
D)Instrumentation
E)None of the above

Question 22

Which of the following is/are empirical approaches to validation?

A)Concurrent validity
B)Synthetic validity
C)Content validity
D)A and B
E)A and C

Question 23

When present employees are used to validate an employment test,which validity is assessed?

A)Construct
B)Concurrent
C)Content
D)Face
E)None of the above

Question 24

Which of the following is not a step in the predictive validation of a new test?

A)Perform job analysis
B)Identify new predictor
C)Identify reliable criterion
D)Hire using scores received in the new test
E)All of the above steps are a part of predictive validation

Question 25

When Jane was writing her final exam in HR,sudden construction blasts in the neighbourhood shook her completely.She was not able to concentrate on the exam,although most of her classmates did not find the noise distracting.Jane received a 15% lower score than what she could have received under normal conditions.The difference between her actual test score and true score is best thought of as what sort of error?

A)Random
B)Systematic
C)Bounded
D)Instrumentation
E)Environmental

Question 26

The number of predictors to be used in a setting depends on:

A)the nature of the job
B)the validity of the predictors
C)the cost of collecting predictor information
D)B and C
E)all of the above

Question 27

When we say that the coefficient of determination for a test is 0.89,it means:

A)89% of the persons who took the test have scores significantly different from each other
B)89% of the persons who took the test have no significant differences in their test scores
C)89% of the variance in test scores among the persons who took the test is due to true variance
D)11% of the variance in test scores among persons who took the test is due to true variance
E)only 11% of the true score is being measured

Question 28

If reliability of a sales aptitude test is 0.81 and the reliability of the criterion (e.g. ,supervisor's rating of salesperson's performance)is 0.64,the validity of the test is most likely to be less than:

A)0.79
B)0.72
C)0.52
D)0.48
E)0.41

Question 29

Reliability of a test can be improved by:

A)shortening the length of the test
B)making the test extremely easy
C)asking multiple questions that measure the same construct
D)A and B
E)A and C

Question 30

Of the following,which focuses on evidence that shows that validity information collected on a predictor in a series of studies can be applied to a new setting?

A)Validity generalization
B)Moderator analysis
C)Psychological fidelity
D)Suppressor variable
E)None of the above

Question 31

If two predictors are highly correlated,in the normal course,how many should be used in a specific selection situation?

A)Both
B)Only one
C)Only the one with the lower validity
D)Only the one with lower reliability
E)None of the above

Question 32

What type of validity refers to the possibility that a predictor may validly measure a particular criterion for some people in some situations but have different validity for another group of people?

A)Divergent
B)Differential
C)Convergent
D)Face
E)Predictive

Question 33

Which of the following factors is/are suggested as reasons for differences among raters?

A)Raters may interpret the rating rules differently.
B)Raters may belong to different age groups.
C)Raters may belong to different races.
D)B and C are reasons.
E)All of the above are reasons.

Question 34

Which of the following is/are rational approaches to validation?

A)Content validity
B)Predictive validity
C)Synthetic validity
D)A and B
E)A and C

Question 35

A clock that is always five minutes fast is:

A)valid and reliable
B)reliable but not valid
C)valid but not reliable
D)unreliable but not invalid
E)none of the above

Question 36

The congruence between the psychological requirements of a job and the psychological requirements imposed by the predictor when it is measured is referred to as what?

A)Moderator variable
B)Suppressor variable
C)Psychological deviance
D)Psychological fidelity
E)Validity generalization

Question 37

The __ validity of an exam in human resource management is established by showing that the exam includes representative samples of all major topics covered in the course.

A)construct
B)concurrent
C)synthetic
D)content
E)none of the above

Question 38

Which type of variable correlates nearly zero with a criterion but is still useful when employed with a predictor with which it is highly correlated?

A)Moderator
B)Random
C)Suppressor
D)Criterion
E)Differential

Question 39

Based on available research,it is fair to assume that what percentage of organizations in Canada use properly validated selection tests?

A)53%
B)33%
C)23%
D)13%
E)3%