Deck 13: Scaling

Question 1

One of the advantages of a Likert scale is that:

A)it is one of the quickest scales to administer.
B)the data are considered to be ordinal level, permitting the use of more powerful statistical procedures.
C)the data are considered to be interval level, permitting the use of the most powerful statistical procedures.
D)the single score based on the Likert scale is not a summary of responses to numerous items.

Answer 1

B

Question 2

Each item in a Likert scale can be considered to be:

A)a nominal measure.
B)an ordinal measure.
C)an interval measure.
D)a ratio measure.

Answer 2

B

Question 3

In selecting items for scales, one would want to make sure that:

A)no items are used until they have passed the test of construct validity.
B)extreme positions on the variable are left off so that regression toward the mean is avoided.
C)the scale achieves multi dimensionality.
D)the items cover the actual range of variation on the variable.

Answer 3

D

Question 4

A Likert scale can also be referred to as a:

A)Guttman scale.
B)multidimensional scale.
C)summated rating scale.
D)rapid assessment instrument.

Question 5

In assessing items for a Likert scale, a negative discriminatory power score on an item means that:

A)high scorers on the overall scale scored lower on this item than did the low scorers on the overall scale.
B)people in the upper and lower quartiles scored very differently on this item.
C)there is a negative relationship between how people score on this item and on the overall scale.
D)the item is a good candidate for inclusion in the final scale.

Question 6

The assumptions of which of the following scaling techniques can enable the researcher to assume that the resulting data are at the interval level of measurement

A)Likert scaling
B)Guttman scaling
C)Thurstone scaling
D)multivariate scaling

Question 7

A scale in which there is a series of statements, each followed by five ordered response alternatives, is called:

A)a Likert scale.
B)a Thurstone scale.
C)a semantic differential.
D)a Guttman scale.

Question 8

In Likert scales, a discriminatory power score for each item:

A)indicates the degree to which people's responses to an item range over the five response alternatives.
B)is a method of assessing the reliability of each item.
C)indicates the degree to which each item discriminates between high scorers and low scorers on the overall scale.
D)indicates how highly that item correlates with people's overall scale score.

Question 9

One of the advantages of using existing items or scales in research rather than developing new ones is:

A)existing scales usually achieve a higher level of measurement.
B)existing scales are more likely to be unidimensional.
C)existing scales are more likely to confirm research hypotheses.
D)the validity and reliability of existing scales have usually already been established.

Question 10

In developing a Likert scale, the best items for inclusion are:

A)those with the lowest DP scores.
B)those at the nominal level of measurement.
C)those whose DP score is zero or close to zero.
D)those with the highest DP scores.

Question 11

In assessing items for a Likert scale, a discriminatory power score would be calculated on:

A)the whole preliminary Likert scale.
B)every item of the preliminary Likert scale.
C)only on those items that are used in the final Likert scale.
D)those items that are being moved from a Thurstone to a Likert scale.

Question 12

A scale whose items measure only one variable is called:

A)a unidimensional scale.
B)a multidimensional scale.
C)a nominal scale.
D)a univariate scale.

Question 13

Which of the following are mentioned as possible sources for items for multiple-item measuring devices

A)the researcher's own imagination
B)people who are considered knowledgeable in an area
C)the people who are the focus of the research
D)all of the answers are mentioned as sources

Question 14

Which of the following is one of the disadvantages of Likert scales

A)It offers respondents too wide a range of response alternatives.
B)It gives respondents only a yes-no response alternative.
C)The data produced by Likert scales is only at the nominal level of measurement.
D)It can be hard to interpret a single Likert score because it summarizes so many responses.

Question 15

A "scale" would be best defined as:

A)a number of items that are combined to form a composite score on a variable.
B)an item that is both valid and reliable.
C)a measuring device that achieves at least an interval level of measurement.
D)a measuring device that involves asking people questions and recording their answers.

Question 16

What should be done if two items in a scale correlate nearly perfectly with one another

A)Keep them both because this shows they are unidimensional.
B)Drop one of them from the scale because they are redundant.
C)Combine the two questions into one so that no information is lost.
D)Drop them both because this shows they have insufficient range and variability.

Question 17

The upper quartile (Q₃)can be defined as:

A)the point in a distribution above which the highest 25 percent of the scores in the distribution are located.
B)the point in a distribution below which the lowest 25 percent of the scores in the distribution are located.
C)the point in a distribution above which the highest 50 percent of the scores in the distribution are located.
D)the point in a distribution above which the highest 10 percent of the scores in the distribution are located.

Question 18

One of the advantages of using scales over single item measures is that:

A)scales are less costly to develop and administer.
B)multiple item measures are generally more valid.
C)multiple item scales enable the researcher to shift the unit of analysis.
D)scales can achieve a lower and more simplified level of measurement.

Question 19

A scale in which a person's scale score is determined by summing the number of questions answered in a particular way is called:

A)a summated rating scale.
B)a Thurstone scale.
C)a unidimensional scale.
D)a semantic differential scale.

Question 20

Multiple item measures are more capable of producing data that are than are single item measures.

A)at the nominal level of measurement
B)partially ordered
C)at the interval level of measurement
D)at the organizational unit of analysis

Question 21

In developing a Thurstone scale, judges are used to:

A)provide the statistical analysis.
B)calculate discriminatory power scores.
C)find the adjective pairs that are used in the scale.
D)assess how favorable an item is toward the variable being measured.

Question 22

The term "rapid assessment instrument" refers to:

A)a statistical technique for assessing the validity of multiple-item scales.
B)short scales used in the evaluation of human service practice.
C)a scale discrimination technique.
D)a technique for avoiding response bias.

Question 23

Which of the following would be the best illustration of "response bias"

A)Women refuse to participate in an interview because the interviewer is male.
B)Interviewers respond more personally to respondents who are white than those who are nonwhite.
C)A person chooses answers in an interview that he thinks the interviewer would like him to give.
D)The cover letter is designed such that it increases the likelihood that people will participate in the interview.

Question 24

In developing a Thurstone scale, items would be good candidates for inclusion in the scale if:

A)there is low agreement about them among the judges.
B)there is high agreement about them among the judges.
C)they have low DP scores.
D)they have high reproducibility.

Question 25

"Funneling" is a technique that helps to reduce:

A)response bias.
B)discriminant functioning.
C)scale discrimination.
D)the unrepresentativeness of samples.

Question 26

An advantage of the semantic differential (SD)over Likert or Thurstone scales is:

A)SDs provide interval level data.
B)SDs provide nominal level data.
C)adequate reliability and validity can be achieved with a smaller number of items.
D)the items in an SD are easier to interpret in terms of operational definitions and measurement.

Question 27

Which of the following is a disadvantage of semantic differential scales

A)They generate ordinal, but not interval, level data.
B)They require a large number of items to achieve adequate reliability.
C)It takes people a long time to complete semantic differential scales.
D)They are difficult and time-consuming to construct.

Question 28

In looking at agency screening for alcohol-related problems in Research in Practice 13.2, research showed that the short assessment scales used by agencies during the brief intake screening could accomplish which of the following

A)detect a potential alcohol problem.
B)detect the nature of an alcohol problem.
C)detect the severity of an alcohol problem.
D)it could do all of these.

Question 29

The concept of "reproducibility" relates to which type of scaling technique

A)Likert scales
B)semantic differential scales
C)Thurstone scales
D)Guttman scales

Question 30

In a semantic differential, a determination of which of the three dimensions a given pair of adjectives measures can best be made by:

A)intuition.
B)a statistical procedure called factor analysis.
C)calculating a discriminatory power score for each adjective pair.
D)calculating a Q score for each adjective pair.

Question 31

The scale type that is based on presenting respondents with a series of polar opposite adjectives is:

A)the semantic differential.
B)the Likert scale.
C)the Thurstone scale.
D)the Guttman scale.

Question 32

According to the developers of the semantic differential, which of the following is NOT one of the dimensions of a concept that is commonly measured by a semantic differential scale

A)reliability
B)potency
C)evaluation
D)activity

Question 33

"The tendency for people's answers to questions to be influenced by things other than their true feelings, beliefs, and behaviors" is called:

A)response rate.
B)response bias.
C)contingency response.
D)pilot testing.

Question 34

Which of the following is a (are)danger(s)of the uninformed use of scales

A)A poorly constructed scale will provide false or inaccurate information.
B)People will believe they have gained some understanding of the world when they have not.
C)Social policy will be based on erroneous data.
D)All of the answers represent such dangers.

Question 35

An advantage of a Thurstone scale over a Likert scale is:

A)Thurstone scales achieve ordinal level of measurement.
B)people can respond to the items in Thurstone scales more quickly.
C)Thurstone scales avoid the ecological fallacy.
D)Thurstone scales have higher reliability.

Question 36

In a Guttman scale, the "easiest" items are the ones that:

A)the most people are likely to agree with.
B)the least people are likely to agree with.
C)the average person is likely to agree with.
D)have the highest reproducibility coefficient.

Question 37

In the administration of multiple item scales, "response set" refers to:

A)the tendency for some people to agree or disagree with statements regardless of their content.
B)the tendency for people to become anxious if they have to repeat a "yes" or a "no" response all of the time.
C)the total score of a person's responses that is achieved by summing up his or her individual responses.
D)the misperception of a question or statement owing to a perceptual set.

Question 38

Which of the following scales involves procedures in its construction that help ensure its unidimensionality

A)Likert scales
B)Thurstone scales
C)semantic differential scales
D)Guttman scales

Question 39

Which of the following is NOT identified in the text as a source of response bias

A)response set
B)response pattern anxiety
C)social desirability effect
D)all of the answers represent sources of response bias

Question 40

In the development of a scale for measuring intimate justice in couple relationships, described in Research in Practice 13.1, which of these steps occurs before the others:

A)apply Cronbach's alpha to the scale.
B)conduct reliability tests on the scale.
C)develop possible items for the scale.
D)compute a discriminatory power score.

Question 41

Identify which scale types purport to yield interval level measurement and how they are constructed to achieve this characteristic.

Question 42

Beyond the general concerns of question wording, identify and explain the factors to be considered in the selection of scale items.

Question 43

What is response bias Identify its sources and what can be done to reduce it

Question 44

Describe in detail the steps involved in developing and using a Thurstone scale.

Question 45

Describe the basic steps that are common to the construction of most scales.

Question 46

Identify the scaling technique that utilizes the concept of reproducibility and explain the meaning of the term as it relates to that scaling technique.

Question 47

Compare and contrast how Guttman and Semantic Differential scales are developed.

Question 48

Describe in detail the steps involved in developing and using a Likert scale.

Question 49

What are the advantages of scales over single item measures

Question 50

What are some of the issues to which human service professionals must be especially attuned in regard to scaling