- Sampling units are the:
- target population elements available for selection during the sampling process.
- identifiable groups of elements of interest to the researcher and pertinent to the information problem.
- complete set of elements identified for investigation.
- list of all population elements from which to draw a sample.
- number of units accidentally left out of the sampling frame.
Answer: A
Difficulty: Medium
Page: 137
- Spectrum Inc., an automobile manufacturer, is interested in determining the satisfaction of the users of one of its brands, Credence. Instead of sampling the entire population of car owners, Spectrum Inc. samples those who owned a Credence. Spectrum is drawing its sample from a:
- sampling unit.
- defined target population.
- designated sampling population.
- subset relevant population.
- full-set relevant population.
Answer: B
Difficulty: Medium
Page: 137
- Jamie Nelson works for a telephone company. She is interested in determining the satisfaction of people with the service provided by the company (for landline customers only). She decides to draw her sample from the telephone directory. For Jamie’s research, the telephone directory is serving as a(n):
- sampling unit.
- element chart.
- sampling frame.
- element.
- census.
Answer: C
Difficulty: Medium
Page: 138
- The difference between probability sampling and nonprobability sampling is that in nonprobability sampling:
- each sampling unit in the defined target population has a known probability of being selected for the sample.
- observed differences in sample findings can be partially attributed to the existence of sampling errors.
- the results can be generalized to the target population within a specified margin of error.
- the researcher is able to judge the reliability and validity of data collected by calculating the probability.
- sampling error is not known.
Answer: E
Difficulty: Medium
Page: 140
- _____ sampling is a probability sampling procedure in which every sampling unit has a known and equal chance of being selected.
- Judgment
- Quota
- Cluster
- Simple random
- Convenience
Answer: D
Difficulty: Medium
Page: 140
- Which one of the following is a nonprobability sample method?
- Simple random sampling
- Cluster sampling
- Snowball sampling
- Area sampling
- Stratified random sampling
Answer: C
Difficulty: Medium
Page: 140
- Which of the following is true of simple random sampling?
- It produces unbiased estimates of the target population’s characteristics.
- It has the benefit of obtaining a complete and accurate listing of the target population elements.
- It is less costly when compared to systematic random sampling because it can be done quickly.
- It requires that the defined target population be ordered in some way, usually in the form of a customer list, taxpayer roll, or membership roster.
- Unlike systematic random sampling, it does not require the sampling units to be given any special code prior to drawing the sample.
Answer: A
Difficulty: Medium
Page: 141
- The difference between simple random sampling and systematic random sampling is that systematic random sampling:
A. is attributable to mistakes in either drawing a sample or determining the sample size. - requires that the defined target population be ordered in some way.
- is a nonprobability sampling procedure.
- has a known and equal chance of being selected.
- is based on intuitive judgment or researcher knowledge.
Answer: B
Difficulty: Medium
Page: 141
- In a systematic sampling study, if the sampling frame has 1,500 names and the desired sample size is 50, the “skip interval” should be:
- 25.
- 0.03.
- 50.
- 30.
- 15.
Answer: D
Difficulty: Medium
Page: 141
- Which of the following is true of proportionately stratified sampling?
- Each stratum is dependent on its size relative to the defined target population.
- It is used when stratification of the target population produces sample sizes for subgroups that differ from their relative importance to the study.
- The size of each stratum is independent of its relative size in the population.
- The smaller strata are sampled more heavily because they make up a larger percentage of the target population.
- Sampling units are divided into mutually exclusive and collectively exhaustive subpopulations called clusters.
Answer: A
Difficulty: Medium
Page: 143
- Before polling students in the School of Business, the researcher divides all the current students into groups based on their class standing (freshman, sophomores, etc.). Then, she randomly draws a sample of 50 students from each of these groups to create a representative sample of the entire student body in the school. Which of the following sampling methods is the researcher practicing?
- Simple random sampling
- Systematic random sampling
- Stratified random sampling
- Cluster sampling
- Snowball sampling
Answer: C
Difficulty: Medium
Page: 143
- The difference between stratified random sampling and cluster sampling is that in cluster sampling:
- the researcher must be able to secure a complete listing of the potential sampling units that make up the defined target population.
- a large number of respondents is interviewed in a relatively short time.
- representativeness of the sample cannot be measured because sampling error estimates cannot be calculated.
- the sampling units are divided into mutually exclusive and collectively exhaustive subpopulations.
- the target population is separated into different groups.
Answer: D
Difficulty: Medium
Page: 143
- Which of the following is true of cluster sampling?
- Each cluster is assumed to be representative of the homogeneity of the target population.
- The prospective sampling units are selected for the sample using systematic random sampling.
- Area sampling is a form of cluster sampling.
- In practice, cluster sampling is difficult to implement.
- Cluster sampling requires a substantially large budget.
Answer: C
Difficulty: Medium
Page: 144
- Which of the following is true of convenience sampling?
- Using convenience samples to develop constructs and scales can be risky.
- It is a time-consuming process when there are a large number of respondents.
- It is commonly used in the later stages of research.
- The representativeness of the sample can be measured because sampling error estimates are not calculated.
- It is commonly used to generalize results to the target population.
Answer: A
Difficulty: Medium
Page: 145
- A college professor wants to know if students in U.S. colleges will find the new textbook that he has written to be useful and interesting. He decides to sample all marketing majors at the school he teaches. He believes that this group of students will be representative of the U.S. college student population. His method of sampling is most probably:
- census sampling.
- convenience sampling.
- judgment sampling.
- quota sampling.
- cluster sampling.
Answer: C
Difficulty: Hard
Page: 145
- A researcher does not have the resources or time to do a probability sample. She wants to do a convenience sample, but she also wants to make sure that all ethnicities are fairly represented in her sample. What kind of sampling would you recommend to her?
- Judgment sampling
- Quota sampling
- Stratified sampling
- Cluster sampling
- Census sampling
Answer: B
Difficulty: Medium
Page: 146
- Which method of sampling relies on the assumption that rare groups of people tend to form their own unique social circles?
- Judgment sampling
- Quota sampling
- Cluster sampling
- Snowball sampling
- Convenience sampling
Answer: D
Difficulty: Medium
Page: 146
- When a market researcher asks if a project will be international, national, or regional in nature, the researcher is trying to determine:
- time frame requirements.
- degree of accuracy.
- scope of the research.
- availability of resources.
- research objectives.
Answer: C
Difficulty: Medium
Page: 147
- Which of the following is true of probability sample sizes?
- The more precise the required sample results, the larger the sample size.
- The larger the desired error, the larger the sample size.
- The shorter the variability in the data being estimated, the larger the sample size needed.
- The lower the level of confidence desired, the larger the sample size needed.
- The population standard deviation is a measure of the dispersion of the population and its square root is referred to as the population variance.
Answer: A
Difficulty: Medium
Page: 147
- If the estimated population standard deviation based on previous studies is 20 percent, and the acceptable error is plus or minus 5 percent, and the z-value for a 95 percent level of confidence is 1.96, the needed sample size is approximately:
- 48.
- 210.
- 62.
- 153.
- 196.
Answer: C
Difficulty: Hard
Page: 148
- If the estimated population proportion is 30 percent based on previous studies, and the acceptable error is plus or minus 5 percent, and the z-value for a 95 percent level of confidence is 1.96, the needed sample size is approximately:
- 323.
- 187.
- 36.
- 212.
- 554.
Answer: A
Difficulty: Hard
Page: 148
- If the estimate of expected population proportion having a desired characteristic based on intuition is 60 percent and the acceptable error is plus or minus 5 percent, and the z-value for a 95 percent level of confidence is 1.96, the needed sample size is approximately:
- 187
- 368
- 295
- 196
- 950
Answer: B
Difficulty: Hard
Page: 148
- For a given level of precision and confidence level, a researcher determines that he needs a sample size of 400. The study was to find out that the proportion of people who smoke is 30 percent. For the same level of precision and confidence, he would have needed the same sample size if the proportion of those who smoked were to change to:
- 40 percent.
- 50 percent.
- 60 percent.
- 70 percent.
- 100 percent.
Answer: D
Difficulty: Hard
Page: 148
- Which of the following statements is true of sample size formulas for nonprobability samples?
- The sample size formula must be adjusted for confidence level before it can be used for nonprobability samples.
- The sample size formula must be adjusted for error tolerance before it can be used for nonprobability samples.
- The sample size formula must be adjusted for population variance before it can be used for nonprobability samples.
- The sample size formula must be adjusted for confidence level and error tolerance before it can be used for nonprobability samples.
- Sample size formulas cannot be used for nonprobability samples.
Answer: E
Difficulty: Medium
Page: 149
- In which stage of the sampling plan will random-digit dialing probably be considered?
- Selecting the data collection method
- Defining the target population
- Selecting the appropriate sampling method
- Creating an operating plan for selecting sampling units
- Identifying the sampling frame
Answer: E
Difficulty: Medium
Page: 151
True/False Questions
- Conducting a census is less time-consuming and less costly than sampling.
Answer: False
Difficulty: Easy
Page: 136
- A population is an identifiable group of elements of interest to the researcher and pertinent to the information problem.
Answer: True
Difficulty: Easy
Page: 137
- A voters list is a common source of a sampling frame.
Answer: True
Difficulty: Medium
Page: 138
- The central limit theorem (CLT) is the theoretical backbone of survey research.
Answer: True
Difficulty: Easy
Page: 138
- A benefit associated with detecting sampling error is that it can be determined before the sample is drawn and data collection is completed.
Answer: False
Difficulty: Medium
Page: 139
- Nonsampling errors reduce the overall quality of the data regardless of the data collection method.
Answer: True
Difficulty: Medium
Page: 140
- Sampling errors are related to the accuracy of data whereas nonsampling errors relate to the representativeness of the sample to the defined target population.
Answer: False
Difficulty: Medium
Page: 140
- Systematic random sampling works better than simple random sampling especially in those situations where there might be a hidden pattern in population.
Answer: False
Difficulty: Medium
Page: 141
- Simple random sampling succeeds in obtaining a complete and accurate listing of the target population elements.
Answer: False
Difficulty: Medium
Page: 141
- When executed properly, systematic random sampling creates a sample of objects or prospective respondents that is very similar in quality to a sample drawn using simple random sampling.
Answer: True
Difficulty: Medium
Page: 141
- The skip interval in systematic random sampling is computed by dividing the number of potential sampling units on the list by the desired sample size.
Answer: True
Difficulty: Medium
Page: 141
- The primary difficulty encountered with stratified sampling is determining the basis for stratifying.
Answer: True
Difficulty: Medium
Page: 143
- Cluster sampling is an example of a nonprobability sampling method.
Answer: False
Difficulty: Easy
Page: 143
- In cluster sampling, each cluster is assumed to be representative of the heterogeneity of the target population.
Answer: True
Difficulty: Easy
Page: 143
- Any geographical unit with identifiable boundaries can be used in area sampling.
Answer: True
Difficulty: Easy
Page: 144
- Referral sampling involves identifying a set of respondents who can help the researcher identify additional people to include in the study.
Answer: True
Difficulty: Medium
Page: 146
- Since quota sampling is a nonprobability sampling method, the representativeness of the sample can be measured.
Answer: False
Difficulty: Medium
Page: 146
- Snowball sampling prohibits bias to enter the study.
Answer: False
Difficulty: Medium
Page: 146
- The lesser the variability in the data being estimated, the larger the sample size needed.
Answer: False
Difficulty: Medium
Page: 147
- In any sampling plan, the first task of the researcher is to choose a method of data collection.
Answer: False
Difficulty: Easy
Page: 151
Essay Questions
- List some of the steps a researcher can conduct once he understands the basics of the central limit theorem.
Answer: With an understanding of the basics of the central limit theorem, the researcher can do the following:
- Draw representative samples from any target population.
- Obtain sample statistics from a random sample that serve as accurate estimates of the target population’s parameters.
- Draw one random sample, instead of many, reducing the costs of data collection.
- More accurately assess the reliability and validity of constructs and scale measurements.
- Statistically analyze data and transform it into meaningful information about the target population.
Difficulty: Medium
Page: 139
- Distinguish between sampling error and nonsampling error.
Answer: Sampling error is any bias that results from mistakes in either the selection process for prospective sampling units or in determining the sample size. Moreover, random sampling error tends to occur because of chance variations in the selection of sampling units. Even if the sampling units are properly selected, those units still might not be a perfect representation of the defined target population, but they generally are reliable estimates.
Nonsampling error occurs regardless of whether a sample or a census is used. These errors can occur at any stage of the research process. In general, the more extensive a study, the greater the potential for nonsampling errors. Unlike sampling error, there are no statistical procedures to assess the impact of nonsampling errors on the quality of the data collected. Nonsampling errors usually are related to the accuracy of the data, whereas sampling errors relate to the representativeness of the sample to the defined target population.
Difficulty: Medium
Page: 139-140
- List the advantages and disadvantages of simple random sampling.
Answer: Simple random sampling has several advantages. The technique is easily understood and the survey’s results can be generalized to the defined target population with a prespecified margin of error. Another advantage is that simple random samples produce unbiased estimates of the population’s characteristics. This method guarantees that every sampling unit has a known and equal chance of being selected, no matter the actual size of the sample, resulting in a valid representation of the defined target population.
Difficulty: Medium
Page: 141
- Discuss the advantages and disadvantages of quota sampling.
Answer: The greatest advantage of quota sampling is that the sample generated contains specific subgroups in proportions desired by the researcher. Use of quotas ensures that the appropriate subgroups are identified and included in the survey. Also, quota sampling reduces selection bias by field workers. An inherent limitation of quota sampling is that the success of the study is dependent on subjective decisions made by researchers. Since it is a nonprobability sampling method, the representativeness of the sample cannot be measured. Therefore, generalizing the results beyond the sampled respondents is questionable.
Difficulty: Medium
Page: 146
- List the steps to be followed in developing a sampling plan.
Answer:
Step 1: Define the target population.
Step 2: Select the data collection method.
Step 3: Identify the sampling frames needed.
Step 4: Select the appropriate sampling method.
Step 5: Determine necessary sample sizes and overall contact rates.
Step 6: Create an operating plan for selecting sampling units.
Step 7: Execute the operational plan.
Difficulty: Medium
Page: 151
Chapter 07
Measurement and Scaling
Multiple Choice Questions
- Which of the following is true of measurement?
- As part of the measurement process, researchers assign labels to phenomena they measure but do not assign numbers to them.
- As part of the measurement process, researchers assign numbers to phenomena they measure but do not label them.
- The goal of the construct development process is to precisely measure each construct.
- It is the process of developing methods to systematically characterize information about persons, events, ideas, or objects of interest.
- The process of measurement begins with scale measurement followed by construct development.
Answer: D
Difficulty: Medium
Page: 158
- A _____ is a hypothetical variable made up of a set of component responses or behaviors that are thought to be related.
- scale point
- construct
- relationshipCoefficient alpha
- statistical parameter
- non-bipolar descriptor
Answer: B
Difficulty: Easy
Page: 159
- Which of the following statements about constructs is true?
A. They are directly measured.
B. They are made up of a combination of several unrelated indicator variables.
C. Construct development primarily involves measuring each construct. - They are an observable concept.
- Construct development begins with an accurate definition of the purpose of the study.
Answer: E
Difficulty: Medium
Page: 160
- The scale measurement process primarily:
- determines how to precisely measure each construct.
- identifies what is to be measured.
- defines what is to be measured.
- defines the purpose of a study.
- identifies the characteristics that define the concept studied in a research.
Answer: A
Difficulty: Medium
Page: 160
- A researcher designs his questionnaire in such a way that a respondent can communicate his/her intensity of like/dislike for a product by circling an appropriate number from the response options that go from 1 to 7 (where 1 is “Strongly disagree” and 7 is for “Strongly agree”). These numbers (1 to 7) are commonly referred to as:
A. construct points.
B. scale portions.
C. scale points.
D. construct ordinals.
E. scale ratios.
Answer: C
Difficulty: Medium
Page: 161
- A professor wants to determine the male-female ratio in his college. He designs a questionnaire in which a respondent only needs to choose between two choices, male and female. The scale used by the professor is an example of the _____ scale.
A. ordinal
B. nominal
C. ratio
D. interval
E. Likert
Answer: B
Difficulty: Medium
Page: 162
- Which of the following is true of an ordinal scale?
- It cannot measure quantitative data.
- It can be used to determine the absolute difference between rankings.
- It allows responses to be rank-ordered in a hierarchical pattern.
- It is less powerful than a nominal scale.
- It enables researchers to make absolute comparisons between responses.
Answer: C
Difficulty: Medium
Page: 162
- Which of the following statements is true about interval scales?
A. They can measure absolute differences between scale points.
B. The standard deviation cannot be calculated for interval scales.
C. They are the highest level scale.
D. They do not allow for comparison between measured objects.
E. They are less powerful than ordinal scales.
Answer: A
Difficulty: Medium
Page: 163
- Which of the following statements is true about ratio scales?
- They cannot measure quantitative data.
- They cannot be used to determine the absolute difference between rankings.
- They are the most basic and least powerful scale design.
- They are less powerful than nominal scales.
- They enable researchers to make absolute comparisons between responses.
Answer: E
Difficulty: Medium
Page: 164
- In a(n) _____, the scale questions are divided into two parts equally and the resulting parts’ scores are correlated against one another.
- split-half test
- test-retest reliability technique
- equivalent form technique
- nominal scale
- ordinal scale
Answer: A
Difficulty: Easy
Page: 166
- Internal consistency specifically refers to the:
- extent to which a scale can reproduce the same measurement results in repeated trials.
- degree to which the individual questions of a construct are correlated.
- degree to which the respondents give similar responses.
- degree to which a specific investigated issue leads the respondent to give a socially expected response.
- level of preparation required to create an appropriate environment for the respondents participating in a research.
Answer: B
Difficulty: Easy
Page: 166
- A measurement scale should:
A. use complex sentence construction. - not have negative descriptors.
C. be internally consistent.
D. not have discriminatory power.
E. not include a neutral descriptor.
Answer: C
Difficulty: Medium
Page: 166
- Using a scale, a researcher determines that the brand attitudes for Volvo are very positive (6.5 on a 7-point scale). Using the same scale for the same population, another researcher discovers that the brand attitudes for Volvo are very negative (2.2 on the same scale). This indicates that there was some problem with the scale being used. This problem is most likely to be related to:
A. scale reliability.
B. scale validity.
C. face validity.
D. discriminant validity.
E. convergent validity.
Answer: E
Difficulty: Medium
Page: 167
- A scale that has a neutral descriptor but has a larger number of response options on one side, either positive or negative, is referred to as a(n) _____ scale.
- unbalanced
- free-choice, even-point
- forced-choice, even-point
- forced-choice, odd-point
- free-choice, odd-point
Answer: A
Difficulty: Easy
Page: 168
- The _____ represents the distance between the largest and smallest response.
- mean
B. standard deviation
C. range
D. median
E. frequency distribution
Answer: C
Difficulty: Easy
Page: 170
- If there is significant variation among subjects regarding their attitudes toward a given brand, this variance will be reflected most clearly in which of the following measures?
- Mean
B. Median
C. Mode
D. Standard deviation
E. Frequency distribution
Answer: D
Difficulty: Medium
Page: 170
- Which of the following types of scales asks the respondent to state his/her agreement or disagreement with a series of statements about a specific brand in terms of a 5-point scale ranging from “strongly agree” to “strongly disagree”?
A. Likert scale
B. Semantic differential scale
C. Paired-comparison scale
D. Constant sum scale
E. Numerical scale
Answer: A
Difficulty: Medium
Page: 171
- Which of the following is true about a semantic differential scale?
- It is a type of nominal scale.
- It cannot capture a person’s attitudes or feelings about a given object.
- Only the endpoints of the scale are labeled.
- It is not possible to calculate the mean for each attribute.
- It is a type of ratio scale.
Answer: C
Difficulty: Medium
Page: 173
- Which of the following is the strongest predictor of future behavior?
A. A person’s stated attitudes
B. A person’s past behavior
C. A person’s stated purchase intentions
D. A person’s exposure to advertisements for a product
E. A person’s stated dislike for competing products
Answer: B
Difficulty: Medium
Page: 175
- A researcher working at Arcade Inc. is interested in finding out people’s perception of Arcade’s prices and quality relative to those of its top market rival. Given this objective, which of the following would prove to be most efficient?
- Graphic rating scale
B. Single-item scale
C. Comparative rating scale
D. Nominal scale
E. Likert scale
Answer: C
Difficulty: Medium
Page: 175
- Which of the following statements about graphic rating scales is true?
A. They are a type of comparative rating scales.
B. They are recommended for use with adults only.
C. They cannot be used to collect emotional data.
D. They do not contain narrative descriptors. - They use a scaling descriptor format.
Answer: E
Difficulty: Medium
Page: 175
- Rank-order rating scales:
A. are easy to use in traditional or computer-assisted telephone interviews, and are especially effective if the number of items being asked is more than six or seven.
B. incorporate a scale point format that allows respondents to compare objects by indicating their order of preference. - can collect more than ordinal data structures, helping the researcher to learn about the intensity or reasoning of the respondents.
D. present the respondent with some type of descriptive graphic continuum as the set of possible raw responses to a given question. - allow respondents to compare their responses against that of others.
Answer: B
Difficulty: Easy
Page: 176
- Which of the following types of scales typically requires the respondent to allocate a given number of points, usually 100, among each separate attribute or feature relative to all the other listed ones?
- Nominal scales
- Rank-order scales
- Constant-sum scales
- Graphic rating scales
- Behavioral intention scales
Answer: C
Difficulty: Easy
Page: 176
- The difference between a single-item scale and a multiple-item scale is that a:
- single-item scale involves collecting data about only one attribute of the object being investigated.
- multiple-item scale is a type of nominal scale.
- single-item scale simultaneously collects data on several attributes of a construct.
- single-item scale provides more than one possible response to its respondents.
- multiple-item scale cannot collect ordinal data.
Answer: A
Difficulty: Medium
Page: 178
- Which of the following statements is true regarding the various scale measurement issues?
A. The researcher should avoid using specific terms and instead use general, broad words.
B. The researcher should use “leading” phrases.
C. The researcher should give clear and simple instructions.
D. The researcher should ensure that scale descriptors have minimal discriminatory power.
E. The researcher should use double-barreled items.
Answer: C
Difficulty: Medium
Page: 179
True/False Questions
- The goal of the construct development process is to precisely identify and define what is to be measured.
Answer: True
Difficulty: Medium
Page: 159
- At the heart of construct development is the need to determine exactly what is to be measured.
Answer: True
Difficulty: Medium
Page: 161
- A feature is an abstract construct if it can be directly measured using physical characteristics.
Answer: False
Difficulty: Easy
Page: 161
- A nominal scale enables respondents to express relative magnitude between the answers to a question and responses can be rank-ordered in a hierarchical pattern.
Answer: False
Difficulty: Medium
Page: 162
- Ordinal scales can be used to determine the absolute difference between rankings.
Answer: False
Difficulty: Medium
Page: 162-163
- For interval scales, it is not possible to calculate mean and standard deviation.
Answer: False
Difficulty: Medium
Page: 163
- Random error produces inconsistency in scale measurements that lead to lower scale reliability.
Answer: True
Difficulty: Medium
Page: 165
- Validity is a measure of accuracy in measurement.
Answer: True
Difficulty: Easy
Page: 166
- The more scale points, the lesser the discriminatory power of the scale.
Answer: False
Difficulty: Medium
Page: 168
- The median is the arithmetic average of all the data responses.
Answer: False
Difficulty: Easy
Page: 170
- A Likert scale cannot measure attitudes and behavior.
Answer: False
Difficulty: Easy
Page: 171
- Semantic differential scales typically use an odd number of scale points.
Answer: True
Difficulty: Easy
Page: 173
- In a well-designed semantic differential scale, the individual scales should be truly bipolar.
Answer: True
Difficulty: Medium
Page: 173
- Behavioral intention scales are difficult to construct.
Answer: False
Difficulty: Easy
Page: 174
- A comparative rating scale is used when the objective is to have a respondent express his/her attitudes, behavior, or intentions about a specific object.
Answer: False
Difficulty: Easy
Page: 175
- A graphic rating scale is a type of comparative rating scale.
Answer: False
Difficulty: Easy
Page: 175
- Only ordinal data can be obtained using rank-order scales.
Answer: True
Difficulty: Medium
Page: 176
- Constant-sum scales should preferably use more than seven attributes.
Answer: False
Difficulty: Medium
Page: 177
- Age is an example of a single-item scale.
Answer: True
Difficulty: Medium
Page: 178
- The number of dimensions of the construct influences the decision to use a single-item versus a multiple-item scale.
Answer: True
Difficulty: Medium
Page: 178
Essay Questions
- Discuss the difference between a concrete variable and an abstract construct.
Answer: A concrete variable can be directly observed whereas an abstract construct must be indirectly measured. An example of a concrete variable is a person’s age or income. An example of an abstract construct is a person’s personality or intelligence.
Difficulty: Medium
Page: 159-161
- A researcher measures the temperature (in degree Fahrenheit) inside a grocery store at which the customers feel most comfortable. Suggest a scale that the researcher can use. Why is it not possible to use the ratio scale in this case?
Answer: The researcher can use the nominal scale, ordinal scale, as well as the interval scale. However, the ratio scale cannot be used in this case. Given that the zero F temperature is not a true zero, the scale can never be a ratio scale. For example, it will be erroneous to say that 64°F is twice as warm as 32°F. To see the fallacy of this argument, simply convert these two temperature points into degree Celsius and you will find that the argument falls apart.
Difficulty: Hard
Page: 162-164
- What are the requirements of designing measurement scales?
Answer: Designing measurement scales requires (1) understanding the research problem, (2) establishing detailed data requirements, (3) identifying and developing constructs, and (4) selecting the appropriate measurement scale.
Difficulty: Medium
Page: 167
- Discuss the measurement problems associated with the semantic differential scale pole descriptors.
Answer: Descriptors used to express the two ends of the scale should be true bipolar ends of the scale. The scale is not valid if bipolar ends lack the expression of extreme intensity associated with end poles. The scale design might not allow for significant magnitudes to exist between two pole descriptions.
Difficulty: Medium
Page: 173
- What are the two factors that drive the use of multiple-item scales in research?
Answer: If the construct under study is multidimensional, then these unique dimensions have to be captured using a multiple-item scale. Also, multiple-item scales generally have more reliability and validity compared to those of single-item scales.
Difficulty: Medium
Page: 178-179
