Research Quiz 6

Professor Jacobson tried to use a Guttman scale on his data set that he had seen published in a major journal. Thames was upset because the items did not form a Guttman scale on his data. Confused, he asked for your help. You should tell him:

items may form a Guttman scale in one sample but not in another.

Exhibit 6-2 Patterns for a Guttman scale on life satisfaction (no mixed types). VARIABLES HEALTH HOBBIES CITY Number of Cases A + + + 100 B + + - 75 C + - - 50 D - - - 25 Refer to Exhibit 6-2. Given the patterns for a Guttman scale on life satisfaction, the coefficient of reproducibility is:

1.00 You have to understand how this coefficient of reproducibility is calculated and the purpose of this concept. Since there is no mixed type, the answer should be "1." Guttman developed this criterion based on the percentage of correct predictions, called the coefficient of reproducibility. It is used to assess the reliability of these scales by arranging items from least extreme to most extreme. It is calculated by counting the number of times than responses fall "outside" the predicted pattern, divided by the total number of responses; which is subtracted from 1.0. A coefficient of .9 or higher is commonly accepted for an acceptable scale. In this case, the reproducibility is high.

Exhibit 6-2 Patterns for a Guttman scale on life satisfaction. VARIABLES HEALTH HOBBIES CITY Number of Cases A + + + 100 B + + - 75 C + - - 50 D - - - 25 Refer to Exhibit 6-2. What is the index score for Line A?:

3 Please go to page 192. Table 6-5. The coefficient of reproducibility is the percentage of correct predictions. So please take a look at Table 6-3. The Mixed types are considered as errors. Since there is no mixed types in the question, the coefficient of reproducibility is 1. A higher coefficient is desirable. We simply add up the number of "yes" responses for the three questions above. For example, if a respondent answered "yes" to two of the four questions, his or her index score would be 2, meaning that life satisfaction is high. If a respondent answered "no" to all three questions, his or her life satisfaction score would be 0, indicating that he or she has low life satisfaction.

The following items measure the extent to which Americans are willing to associate with extraterrestrials. 1. It's OK if my child marries an extraterrestrial. 2. It's OK to have an extraterrestrial for a friend. 3. It's OK to have an extraterrestrial living on the block. 4. It's OK to have an extraterrestrial living in my state. These items illustrate a:

Bogardus social distance scale. The Bogardus social distance scale is a psychological testing scale created by Emory S. Bogardus to empirically measure people's willingness to participate in social contacts of varying degrees of closeness with members of diverse social groups, such as racial and ethnic groups. The scale asks people the extent to which they would be accepting of each group (a score of 1.00 for a group is taken to indicate no social distance): As close relatives by marriage (score 1.00) As my close personal friends (2.00) As neighbors on the same street (3.00) As co-workers in the same occupation (4.00) As citizens in my country (5.00) As only visitors in my country (6.00) Would exclude from my country (7.00) Researchers have used the scale to study racism and measure the attitude of whites toward blacks.

_____________ is the process of testing the validity of a measure by examining its relationship to other indicators of the same variable.

External validation If a respondent scores high on a particular scale, the respondent should score high on similar items not included in the index. For instance, if one is scored as political conservative on some measures, they should be conservative on other measures as well. The process is called external validation. An index score should be related to similar measures of the index that were not included in the index.

Which type of validity is used to select items for an index?

Face Face validity means, on the surface, the item must be logical and the quality of the measure seems reasonable.

Internal validation refers to the process of testing the validity of a measure, such as an index or scale, by examining its relationship to other, presumed indicators of the same variable.

False The correct answer should be external validation. Scales and indexes have to be validated. Internal validation checks the relation between the individual measures included in the scale, and the composite scale itself. External validation checks the relation between the composite scale and other indicators of the variable.

Index construction refers to the process of assigning scores to patterns of responses.

False There are FOUR main steps in the process of index construction, which are selecting possible items, examining their empirical relationships, scoring the index, and validating it.

Researchers use scales much more frequently than they use indexes .

False Babbie provided two possible reasons: 1. scales are often difficult or impossible construct from the data at hand. 2. methods of index construction seem so obvious and straightforward that they aren't discussed much in the literature.

The principle of item selection that helps develop items with sufficient range on the variable you are measuring is known as coefficient.

False It is known as variance. In the process of item selection, you want to ensure that items have sufficient variation. In other words, if everyone gives the same response to a particular item, then the item would have no use in constructing the scale.

The Thurstone scale was named after Rensis Likert.

False The Likert scale was named after Rensis Likert. The Thurstone scale was developed by Louis Leon Thurstone in 1928, as a means of measuring attitudes towards religion. In psychology and sociology, it was the first formal technique to measure an attitude. It is made up of statements about a particular issue, and each statement has a numerical value indicating how favorable or unfavorable it is judged to be. People check each of the statements to which they agree, and a mean score is computed, indicating their attitude.

The Bogardus social distance scale measures how much alienation people experience.

False The Bogardus Social Distance Scale measures perceptions of the social distance separating groups from one another. Emory Bogardus developed the scale in the 1920s to measure the willingness of members of one ethnic group to associate with members of another ethnic group. The scale can be used with other groups, including religious, political, and deviant groups.

The coefficient of reproducibility is used with which scale?

Guttman The Guttman scale is a type of composite measure used to summarize several discrete observations and to represent some more general variable. A hypothetical, perfect Guttman scale consists of a unidimensional set of items that are ranked in order of difficulty from least extreme to most extreme position. An important objective in Guttman scaling is to maximize the reproducibility of response patterns from a single score. A good Guttman scale should have a coefficient of reproducibility (the percentage of original responses that could be reproduced by knowing the scale scores used to summarize them) above .85. The Guttman scale is used mostly when researchers want to design short questionnaires with good discriminating ability. The Guttman model works best for constructs that are hierarchical and highly structured such as social distance, organizational hierarchies, and evolutionary stages.

Professor Swanson developed an index of marital happiness. Several items in Swanson's questionnaire were not part of the index (i.e., "How many serious quarrels or arguments have you had with your spouse in the past year?") Swanson compared the index scores of marital happiness with these questions. Swanson was doing:

Index validation Index validation should show that that the index provides measure of some variable and those ranked-ordered for one variable should be rank-ordered for similar variables.

The unit of observation in "A study of Differences in Business Ethical Values in Mainland China, the U.S. and Jamaica" is:

Individual The unit of observation may not be the same as the unit of analysis. In this case, the unit of observation is the individual and the unit of analysis is the group.

Both scales and indexes are which level of measurement?

Ordinal There is no true zero. There are however differences in ranking.

Professor Duncan administered a questionnaire containing the following items: Please tell me how you feel about your supervisor's leadership style on the following three items: valuable ----:----:----:----:---- worthless easy ----:----:----:----:---- demanding critical ----:----:----:----:---- uncritical Duncan was using a:

Semantic differential scale. Semantic differential refers to a questionnaire format in which the respondent is asked to rate something in terms of two, opposite adjectives (e.g. rate textbooks as "boring" or "exciting"), using qualifiers such as "very", "somewhat", "neither" to bridge the distance between the two opposites.

What are the two decisions to be made in deciding how to score an index?

The desirable range of index scores and the assignment of scores for each response. There are two important steps in index scoring. The first is to determine the desirable range of the index scores. The second is to determine whether to give each item in the index equal or different weights. Standard: items should be weighted equally unless there are compelling reasons for differential weighting.

Unless there are compelling reasons to the contrary, items should be equally weighted in index scoring

True This is part of index scoring. You must first decide the desirable range of the index scores. Then you have to assign scores to each response. According to Babbie, equal weighting should be the norm.

For an index or scale to be considered unidimensional, its component items should be indicators of only one dimension.

True

If an index fails to predict external validation items, the conclusion to be drawn is more ambiguous than if item analysis shows inconsistent relationships between the items and the index itself.

True

If two items are not related to one another empirically, it is unlikely that they measure the same variable.

True

The coefficient of reproducibility refers to the ability of a Guttman scale to reproduce the pattern of individual responses.

True

Although it is referred to as Likert scaling, to call it a scale is really a misnomer.

True According to Babbie, Likert created a technique for combining the items into scale. However, Likert's scaling technique is rarely used today, his answer FORMAT is one of the most widely used in survey research. In the vast majority of cases, Likert scaling refers to the format instead of scaling.

Item analysis is an assessment of whether each of the items included in a composite measure makes an independent contribution or merely duplicates the contribution of other items in the measure.

True Item analysis is an important first step in validating the index.

The purpose of multivariate analysis in index construction is to discover the simultaneous interaction of the items to determine whether they are all appropriate for inclusion in the same index.

True Multivariate analysis is the analysis of the simultaneous relationships among several variables. If two items are highly correlated, then they may not be included in the same index.

Semantic differential asks respondents to choose between two opposite positions on various dimensions.

True The Semantic Differential (SD) measures people's reactions to stimuli in terms of ratings on bipolar scales defined with contrasting adjectives at each end.

Generally, a composite measure should represent only one dimension of a concept.

True The process is called scale and index construction. Unidimensionality is an important characteristics of the process. For instance, according to Babbie, items reflecting religious fundamentalism should not be included in a measure of political conservatism, even though the two variables might be empirically related to each other.

Both index and scale are ordinal and unidimensional.

True There is no true 0 in index or scale. They ought to be uni-dimensional, that is a composite measure should reflect only one dimension of a concept.

Scale scores convey more information than index scores.

True This is because scale accounts for the relative intensity of various items.

Typologies are typically nominal composite measures, dealing with more than one dimension.

True Typology refers to the classification (typical nominal) of observations in terms of their attributes on two or more variables. The classification of newspapers as liberal-urban, liberal-rural, conservative-urban, or conservative-rural would be an example. Typology is an important concept when there are multiple dimensions to a concept. Let us use the political orientation case in a slightly different way. Voters tend to be described as conservative or liberal in their political views. In this case, political orientation is treated as a single dimension variable. However, some liberals may have very conservative views on certain issues, such as abortion. some conservatives might have very liberal views on some issues, such as the legalization of undocumented immigrants in the U.S. Therefore, researchers have to rely on typologies to describe the multidimensional nature of the variable--political orientation.

The whole point of scoring is to create a single composite index out of several items.

True When you have chosen the best items for your index, your next step is to assign scores for particular responses, thereby creating a single composite measure out of the several items.

A(n) ____ summarizes the intersection of two or more variables.

Typology Typology is a composite measure that involves the classification of observations in terms of their attributes on two or more variables. Such classification is usually done on a nominal scale. An example of a typology would be classification such as by age and health: young-healthy, young-sick, old-healthy, old-sick. It is also called a "taxonomy."

Sydney wanted to analyze the intersection between two variables-school type (public or private) and school location (urban or rural) - in her study of student values. Which approach should she use?

Typology Typology is the answer, because it is the classification (typical nominal) of observations in terms of their attributes on two or more variables. The classification of newspapers as liberal-urban, liberal-rural, conservative-urban, or conservative-rural would be an example.

Robert notices that several respondents did not answer the question of whether they took out loans to fund their education. However, later in his survey, Robert finds that respondents told him about the source(s) and the amount(s) of their college loans. In handling the missing data on the question of whether or not they had a college loan, Robert's best approach is to:

Use the latter information about amount and source of loans as grounds for assigning values to the cases which had missing values.

Which of the following steps in index construction is the first one to be out of sequence?

Validate the index

When selecting items for a composite index, which of the following should be kept in mind? e-correct

a-Each item should have face validity. b-A composite measure should be unidimensional. c-There should be concern about the amount of variance provided by the items. d- Each item should have face validity and a composite measure should be unidimensional. e-All of these choices are correct.

Among the reasons for the frequent use of composite measures is(are) that: e-correct

a-the researcher is seldom able to develop in advance single indicators of complex concepts. b-a single data item might not have enough categories to provide a range of variation c-composite measures give a more comprehensive and accurate indication of a given variable d-they are efficient data reduction devices e-all of these choices

Using Likert-type scale items, Professor Wong developed an index to assess job morale. Wong administered the index to 2,000 workers and found that 1 to 4 responses were missing on each of the items. What is the best option for Wong:

assign the middle value to cases with missing data.

Nora has about 6 missing cases on her index of voting interest out of 1500 respondents. The best strategy for handling these missing cases would be to:

exclude them from the construction of the index and the analysis. The number of missing cases are relatively small in Nora's case. She does not have the information about why the cases are missing. Voting interest tends to change over time. Therefore, it is impractical to do follow-up studies of the missing cases.