Deck 7: Fairness and Bias

Question 1

Test fairness relates mainly to which of these psychometric characteristics of a test?

A) validity
B) reliability
C) norms
D) scoring

Answer 1

A

Question 2

Some students (call them Group A) have studied the material on Chapter 7 for 20 hours. Other students (call them Group B) have not even scanned the chapter. Both groups take a quiz on Chapter 7. What do we expect to find when we compare average scores of Groups A and B on the Chapter 7 quiz?

A) We expect Group A to have a higher mean.
B) We expect Group B to have a higher mean.
C) We expect the group means to be about the same.
D) We can't develop any expectation about the outcome.

Answer 2

A

Question 3

A test intended to measure reading ability contains many items about public transportation systems. Rural children are not very familiar with such systems, hence score low on the test. In the terminology of validity, knowledge of public transportation systems in this instance is contributing ________.

A) Unreliable variance
B) Construct irrelevant variance
C) Construct under-representation
D) Internal inconsistency

Answer 3

C

Question 4

We compare the performance of groups A and B on a test. Group B performs noticeably less well than Group A. What does this tell us about the fairness of the test?

A) The test is biased against Group A.
B) The test is biased against Group B.
C) It tells us nothing about bias in the test.

Question 5

In the "panel review" method for studying test bias, pane members are typically selected to represent different _____.

A) Levels of testing expertise
B) Subgroups of people
C) Academic content fields
D) Types of psychopathology

Question 6

What does the text suggest about the frequency with which the "panel review" procedure is used in test development?

A) It is used with nearly all recently developed tests.
B) It is used with only a minority of recently developed tests.
C) In earlier days, it was widely used, but its use has declined recently.

Question 7

When attempting to identify test items that show bias against some group, it is possible to compare results of statistical analysis of the items with results of panel review of the items. What does research show about the degree of agreement between these two procedures? Agreement is _____.

A) very high
B) moderate
C) very low

Question 8

In the context of studying fairness of tests, the term panel review refers to review of test items by _____.

A) content experts in whatever the test attempts to cover
B) bias-detection item analysis procedures
C) differential prediction models
D) representatives of different cultural groups

Question 9

Differential item functioning (DIF) is used mainly in the study of ___.

A) item bias
B) item reliability
C) test validity
D) test norms

Question 10

Which of these terms refers to test-takers of equal ability differing in their responses based on their group membership (e.g., race or gender) ___.

A) differential item functioning
B) item characteristic curve
C) item information function
D) item response theory

Question 11

In the most typical application of differential item functioning (DIF) analysis, the equivalence of the groups being compared is defined in terms of ___.

A) total score on the test
B) performance on an external criterion
C) GPA
D) clinical judgment

Question 12

The Mantel-Haenszel procedure is a technique for studying ___.

A) item reliability
B) differential item functioning
C) the guessing parameter
D) criterion-referenced tests

Question 13

We are using differential item functioning to study potential bias in test items. What technical terms do we use to designate the groups being compared?

A) Large and small groups
B) High and low groups
C) Focal and reference groups
D) Norm and criterion groups

Question 14

We are using differential item functioning to study potential bias in test items. We contrast performance in a minority group and a majority group. What technical term do we use to designate the minority group?

A) Target
B) Focal
C) Reference
D) Bias

Question 15

When using item characteristic curves (ICCs) to study bias in test items, which of these suggests lack of bias?

A) Overlapping curves
B) Curves with no overlap
C) Curves that match at top and bottom but not in the middle
D) Curves that match in the middle but not at top and bottom

Question 16

The concepts of slope bias and intercept bias are especially applicable to ________ validity.

A) content
B) criterion-related
C) construct
D) consequential

Question 17

The two types of bias in the context of criterion-related validity, especially predictive validity are __________.

A) intercept and slope bias
B) intercept and response bias
C) response and slope bias
D) integral and incidental bias

Question 18

We are studying potential test bias using differential prediction. We find that the correlation between a test and the criterion we are trying to predict is exactly the same (say, r = .45) in groups A and B. Based on just that information, what can we conclude about slope bias and intercept bias?

A) There is no slope or intercept bias.
B) There is both slope and intercept bias.
C) There is no intercept bias but there may be slope bias.
D) There is no slope bias but there may be intercept bias.

Question 19

Studying whether the structure of a test, as indicated by factor analysis, is the same in two different groups is an example of applying concepts of ____________.

A) Norm-referencing
B) Test-retest reliability
C) Measurement invariance
D) Standard error of estimate

Question 20

The manual for a test of depression reports a correlation of .35 between this depression measure and a measure of anxiety based on the standardization sample which is 95% White. We now get the correlation between this depression measure and a measure of anxiety based on an Asian sample. This is an example of applying concepts of ____________.

A) Norm-referencing
B) Test-retest reliability
C) Measurement invariance
D) Standard error of estimate

Question 21

A test of math problem solving is administered in a large print version to a student with a visual disability. Which is the technical term for this change in strict adherence to standardized administration of the test?

A) technical adjustment
B) justified alteration
C) accommodation
D) modification

Question 22

The state's math competency assessment consists of 30 multiple-choice items in the usual paper-and-pencil format. To assess the math competency of a student with severe, multiple disabilities, the teacher interviews the student. This type of change in the assessment is called a(n) ________.

A) accommodation
B) modification
C) transformation
D) conversion

Question 23

Two types of changes in standardized administration of a test for persons with disabilities are called accommodation and modification. For which are the norms for the test presumed to be applicable?

A) accommodation only
B) modification only
C) both accommodation and modification
D) neither accommodation nor modification

Question 24

Strict adherence to standardized administration of a test is sometimes made in the test's administration to individuals with disabilities. What is the principal rationale for such changes?

A) To improve the test's reliability
B) To facilitate scoring the test
C) To better measure the test's target construct
D) To enhance applicability of the test's norms

Question 25

Which popular term does the text use to describe the goal of using an accommodation for a test?

A) level the playing field
B) take your mark
C) not-so-fast
D) flipping a coin

Question 26

What conclusion applies regarding applicability of a test's norms for accommodations and modifications? The norms should be applicable __________.

A) For both accommodations and modifications
B) For neither accommodations nor modifications
C) For accommodations but not modifications
D) For modifications but not accommodations

Question 27

What does the research suggest about whether accommodations work as intended?

A) They usually work as intended.
B) They usually do not work as intended.
C) Results are too varied to allow a conclusion.

Question 28

What conclusion does the text suggest as the profession's position about the question of whether scores derived from a test administered with accommodations should be "flagged," that is, marked to indicate the nonstandard administration?

A) Scores should always or nearly always be flagged.
B) Scores should never or rarely be flagged.
C) Professionals have no consensus position on this matter.

Question 29

What does the research suggest about the degree of bias against minority group members in cognitive ability tests?

A) There is little or no bias.
B) There is a great deal of bias.
C) There is bias at younger ages but not at older ages.
D) Too little research has been conducted to draw any conclusion.

Question 30

What does the research suggest about the degree of bias against minority group members in tests such as the MMPI and Rorschach?

A) There is little or no bias.
B) There is a great deal of bias.
C) There is bias at younger ages but not at older ages.
D) Too little research has been conducted to draw any conclusion.