Deck 7: Measurement

Question 1

To achieve standardization the content of job application tests should be the same for all applicants.

Answer 1

True

Question 2

Scoring keys for tests should be developed immediately after the test has been administered.

Answer 2

False

Question 3

A correlation between two variables does not necessarily that one causes the other.

Answer 3

True

Question 4

With very large samples, it is possible for a fairly weak relationship to still be statistically significant.

Question 5

Research shows that when an attribute is measured by both objective and subjective means, there is often relatively low agreement between scores from the two types of measures.

Question 6

If scores are classified as "low", "medium", and high", the scale of measurement is nominal.

Question 7

When numbers are assigned by category, this reflects an ordinal scale of measurement.

Question 8

The most appropriate measure of central tendency for nominal scale data is the median.

Question 9

Measures are methods or techniques for describing and assessing attributes of objects that are of concern to us.

Question 10

Numerical employment interview ratings represent objective measures.

Question 11

If an individual has a z-score of 2.0 on a performance test, this indicates this person's score is twice as high as the average test score.

Question 12

The standard deviation is a measure of the central tendency of a scale.

Question 13

Most staffing measures can be best described as being on a ratio scale.

Question 14

A correlation coefficient ranges from 0 to +1.

Question 15

Human body weight is an example of a variable measured on a ratio scale of measure.

Question 16

Standard scores are also useful for determining how a person performed, in a relative sense, on two or more tests. This is helpful for comparing relative standing across several tests.

Question 17

A standard level for statistical significance is p<.50.

Question 18

When developing measures, it is a good idea to give each rater some license to interpret the meaning of scores as fits the specific situation.

Question 19

Measurement standardization applies to measurement content, not the administration of measurements.

Question 20

A rank ordering of five job candidates in terms of overall qualification for the job is an example of an ordinal scale.

Question 21

Which of the following results represents statistically significance as it is commonly understood?

A) r = .30
B) p < .05
C) r > .10
D) all of the above

Question 22

While correlation is valuable as an indicator of the degree of association between variables, it is generally not used as a tool for prediction.

Question 23

Which of the following statements about measurement is not correct?

A) It assigns numbers to objects.
B) It is based on rules determined in advance.
C) It measures the object not the attribute.
D) It measures constructs.

Question 24

If the correlation coefficient between variables X and Y is .90, this means that the proportion of common variance shared by the two variables is ninety percent.

Question 25

A scale in which objects are rank ordered according to how much of the attribute they possess.

A) nominal
B) ordinal
C) interval
D) ratio

Question 26

A correlation score of minus one between two sets of scores indicates an exceptionally low association or relationship between the two sets.

Question 27

The most appropriate measure of central tendency for nominal data is the ______.

A) mean
B) median
C) standard deviation
D) mode

Question 28

Which of the following is not part of the process of measurement?

A) choosing an attribute of concern
B) assessing the attribute's acceptability to applicants
C) using the measure to gauge the attribute
D) constructing a measure of the attribute

Question 29

The correlation coefficient does not measure the change in one variable caused by another variable.

Question 30

Length in inches is an example of a(n) _________.

A) ratio scale
B) nominal scale
C) ordinal scale
D) interval scale

Question 31

A correlation coefficient of 1.0 between variables X and Y indicates that there is a perfect linear relationship between these two variables.

Question 32

A nominal scale of measurement has ________.

A) rank order
B) an absolute true zero point
C) classification by categories
D) equal differences between points on a scale of measurement

Question 33

A scale in which a given attribute is categorized, and numbers are assigned to the categories, but there is no order or level implied among the categories.

A) nominal
B) ordinal
C) interval
D) ratio

Question 34

The research literature indicates a high correlation between subjective and objective measures of performance.

Question 35

The process of ensuring that all test takers receive the same tests in the same context, as a way to eliminate extraneous influences on test performance is called ____________________.

A) standardization
B) optimization
C) rationalization
D) inflation

Question 36

The likelihood that a correlation exists in a population, based on knowledge of the actual value of r in a sample from that population is the ____________.

A) practical significance
B) likelihood ratio
C) functional statistic
D) statistical significance

Question 37

The variability of a measure is best captured by the _____________.

A) mean and standard deviation
B) range and standard deviation
C) range and median
D) mode and mean

Question 38

A scale in which objects are rank ordered according to how much of the attribute they possess, with equal space between objects.

A) nominal
B) ordinal
C) interval
D) ratio

Question 39

Rankings of the finishes of competitors in a foot race is an example of a(n) _______.

A) ratio scale
B) ordinal scale
C) interval scale
D) none of the above

Question 40

Performance appraisal ratings are subjective measures.

Question 41

The consistency of measurement of an attribute refers to its validity.

Question 42

Measuring the accuracy of a mechanical ability test in predicting the job performance of current employees involves a predictive validity design.

Question 43

The extent to which scores on a knowledge test truly reflect a job applicant's knowledge is a measure of the test's validity.

Question 44

The correlation coefficient expresses _________.

A) causality
B) variance
C) strength of the relationship between two variables
D) the range between two variables

Question 45

The larger the correlation coefficient, _________.

A) the greater the practical significance
B) the smaller its practical significance
C) the larger its range
D) the smaller its statistical significance

Question 46

The proper test to determine that a given sample correlation is statistically significant as an estimate of a correlation in a population is the __________.

A) z score
B) standard deviation
C) squared correlation coefficient
D) the t test

Question 47

Even when measurement error is present, true scores can be measured with perfect precision.

Question 48

In staffing the scores of individuals are treated as if they were the attribute itself, rather than merely indicators of the attribute.

Question 49

The standard error of measurement is a useful indicator of how accurate actual scores of applicants are in assessing the true scores of a given measure.

Question 50

Perfect reliability is virtually impossible to achieve because of the presence of measurement error.

Question 51

Error caused by failing to measure a key aspect of the attribute of interest (i.e., the attribute we wish to measure), is known as contamination error.

Question 52

Asking different questions of job applicants during comparisons of interview ratings between these same job applicants is an example of contamination error.

Question 53

Comparing scores of objective measures within the same time period is a measure of internal consistency.

Question 54

A complete lack of correlation between two variables is expressed by a correlation coefficient of _________.

A) .50
B) 1.00
C) -1.00
D) zero

Question 55

A z score of 3.0 for a job applicant's score on an aptitude test indicates the applicant ______.

A) achieved a score 3 points higher than the average score
B) received a score 3 standard deviations above the mean score
C) had a score 3 percent higher than the average score
D) had a score plus or minus 3 points relative to the mean score

Question 56

Calculation of the test-retest reliability of scores between time periods is done for objective measures, not subjective measures.

Question 57

If an attribute of job performance is "planning and setting work priorities," and the raters fail to rate people on that dimension during their performance appraisal, then the performance measure is contaminated.

Question 58

If all the members of a panel interview reach the same conclusion regarding a person who is being interviewed, it could be said that the interview ratings are reliable.

Question 59

A measure with a coefficient alpha of .55 should generally be regarding as having adequate reliability.

Question 60

Reliability sets the lower limit on validity.

Question 61

The assessment of concurrent validity is more convenient and more efficient than the assessment of predictive validity.

Question 62

Criterion measures are not used in the calculation of content validity.

Question 63

When predictor and criterion scores have been obtained, the predictor can be considered valid if __________.

A) the correlation coefficient has the desired practical and statistical significance
B) the correlation coefficient has statistical significance alone
C) the correlation coefficient is insignificant
D) the correlation coefficient has practical significance alone

Question 64

Which of the following is a true statement?

A) true score divided by error equals actual score
B) actual score equals true score plus error
C) actual score plus error equals true score
D) true score equals variance plus error

Question 65

Contamination error _________.

A) is easy to minimize and control
B) is the same as deficiency error
C) represents unwanted sources of influence on a measure
D) shows a lack of an acceptable coefficient alpha

Question 66

Reliability of a measure places ________.

A) no limit on the validity of a measure
B) a lower limit on the validity of a measure
C) an average limit on the validity of a measure
D) an upper limit on the validity of a measure

Question 67

The standard error of measurement allows ________.

A) calculation of confidence intervals for true scores
B) estimates of content validity
C) calculation of confidence intervals for actual scores
D) calculation of confidence intervals for error

Question 68

Which of the following is an example of contamination error?

A) irrelevant material on the test
B) guessing by the test taker
C) different time limits for the same test
D) all of the above

Question 69

Content validation is most appropriate when ____________.

A) reliability is too high
B) sample size is large
C) criterion measures are readily available
D) sample size is too small for criterion validity calculation

Question 70

Reliability of measurement is defined as __________.

A) frequency of measurement
B) magnitude of measurement
C) accuracy of measurement
D) consistency of measurement

Question 71

The process of criterion-related validation begins with the identification of criterion measures.

Question 72

Comparing the same supervisor's rating of an employee's performance at T1 to T2 is a method for assessing ______ reliability.

A) intrarater
B) fixed-point
C) test-retest
D) interrater

Question 73

Comparing objective scores from T1 to T2 is a method for assessing _____ reliability.

A) maximal asymptotic
B) internal consistency
C) test-retest
D) interrater

Question 74

Validity of a measure is best defined as ________.

A) the upper limit of reliability
B) the degree to which the measure does indeed measure what it is intended to measure
C) the lower limit of reliability
D) the consistency of measurement

Question 75

Deficiency error would not indicate a failure to _________.

A) measure some portion of the attribute of interest
B) adequately define the attribute of interest
C) construct a proper measure of the totality of the attribute
D) calculate the standard deviation

Question 76

Even if methodological and statistical differences across criterion-related validation studies are not controlled for statistically, it is still probable that validity can be generalized from one specific situation to another.

Question 77

Which of the following is a type of validity?

A) test-retest
B) coefficient alpha
C) criterion
D) parallel forms

Question 78

Coefficient alpha assesses ____________.

A) reliability within a single time period
B) reliability between time periods
C) reliability between samples
D) test-retest reliability
E) A and B are correct

Question 79

One guideline for effective staffing practice is that all predictors should be routinely subjected to content validation.

Question 80

Organizations that collect assessment data need to attend to professional standards that govern their use.