CH-5 Descriptive and Causal Research Designs – Essentials of Marketing Research : Book Guide

Chapter Five Descriptive and Causal Research Designs Answers to Hands-On Exercise 1. What was Lee Apparel Company’s overall goal for conducting such an extensive test market of its new line of jeans under the brand name “Riders”? In your opinion did the company achieve its goal? Why or why not? Lee wanted to use the benefits of a customer database to ensure a successful brand launch and future positioning strategy. It was highly successful. The approach generated new sales, response to direct mail pieces, new customer contacts for the database, and information for planning strategy. 2. Identify and explain the strengths and weaknesses associated with the test market process used by the Lee Apparel Company. The strengths are apparent based on the results Lee accomplished for its Riders brand, despite extensive competition in the jeans market. The weaknesses include the expense of conducting such an extensive test market, the time commitment (two years), and the potential for the competition to find out about the company’s plans. 3. In your opinion, should the company give consideration to the development and implementation of Internet-based test marketing strategies? Why or why not? Developing an Internet-based strategy may be worthwhile given the cost savings and added data provided through clickstream analysis. However, people still tend to buy clothes in physical stores and jeans do not require much in the way of information search. It may be beneficial but such a strategy is likely not as relevant as the one in place. ANSWERS TO REVIEW QUESTIONS 1. Identify and discuss the advantages and disadvantages of using quantitative survey research methods to collect primary data in marketing research. The primary pros associated with using quantitative survey research methods are: • The ability to distinguish small differences • The ability to capture data from larger samples at less cost • Data, once captured, is numerical and objective lending itself to statistical analysis • The ability to delve into factors and relationships between variables • Surveys can be distributed and administered efficiently Collecting primary raw data via survey research methods has a number of downsides: • Constructs and scales, when improperly constructed can lead to GIGO (garbage in, garbage out) • Data analysis procedures may be performed incorrectly and data results can be misinterpreted and misunderstood • Survey outcomes lack the “depth” associated with qualitative methodologies due to the inability of questions to probe • Low response rates and protracted return rates • Surveys can suffer from internal infirmities in their design, again leading to GIGO 2. What are the three factors that affect choice of appropriate survey method? How do these factors differ in person-administered surveys as opposed to self-administered surveys? Researchers must consider situational, task, and respondent factors when choosing a survey method. The situational factors include the budget, completion time frame, and quality requirements. Task factors include: • Task difficulty • Required stimuli • Amount of information asked from respondents • Sensitivity of the research topic. Respondent characteristics include: • Diversity • Incidence rate • Respondent participation—respondent participation involves the respondent’s ability to participate, willingness to participate, and knowledge. Person-administered surveys are more expensive and take longer to complete but are best for difficult tasks requiring stimuli. Self-administered surveys are best for sensitive topics, and when participants are able, willing, and knowledgeable. 3. Explain why survey designs that include a trained interviewer are more appropriate than computer-assisted survey designs in situations where the task difficulty and stimuli requirements are extensive. When it is difficult to complete a questionnaire (due to the nature of the directions, constructs, and scales associated with queries, or the length of a survey) computer-assisted survey designs run up against a number of roadblocks. First, a computer-assisted survey doesn’t require the presence of a trained human interviewer. Therefore, if a respondent is puzzled about a particular question or feels pressed for time there no support system to explain away concerns and answer legitimate questions. The other drawback associated with computer-assisted survey designs is one of “context”. While it’s clear more and more people feel comfortable interacting with a computer the world is still populated with many individuals who are anxious about technology and computers. In this respect, the presence of a trained interviewer cannot be underestimated. 4. Explain the major differences between in-home interviews and mall-intercept interviews. Make sure you include their advantages and disadvantages. A mall-intercept is a face-to-face personal interview which takes place in a shopping mall. In-home interviews speak of data that is captured in the “comfort” of one’s home. The word “comfort” is important here. Today, we live in a society where more people work outside the home, and are less inclined to “open the door” and invite in anyone but close friends and family. In this respect the comfort associated with an interviewee being on their “home turf” (more willing to share feelings and beliefs in an unhurried way) is a double-edged sword for the researcher. If the interviewee isn’t willing to let the interviewer in, everything grinds to a halt. A mall-intercept has an advantage that many more people can be canvassed in a shorter period of time, reducing costs to the research team and client. However, mall-intercept suffers from two problems: • The persons in the mall may not be closely aligned with the target market in question. • The purpose they are in the mall is to browse and shop, not to speak to an interviewer—so there may be a response rate problem. On the other hand, if the interview requires a good deal of equipment (recorders, projectors, computers, or test-products) a mall may be the best place from a logistical standpoint. 5. How might measurement and design errors affect respondent errors? Respondent errors occur when responses do not reflect the “true” answer. If questions are written in a way that lead, bias, distort, or provide inappropriate choices, respondent error will occur. 6. Develop three recommendations to help researchers increase the response rates in direct mail and telephone-administered surveys. Response rates can be enhanced using several techniques (particularly those recommended by Dillman in his Total Design Method (TDM)). The following are commonly used to increase response rates in direct mail and telephone-administered surveys: • Offer incentives • Emphasize the sponsor to enhance credibility • Personalize the invitation 7. What is “nonresponse”? Identify four types of nonresponse found in surveys. “Nonresponse” refers to the lack of response from units in the sample. It may be caused by refusals or by problems reaching the unit. Nonresponse error is a systematic bias that occurs when the final sample differs from the planned sample, usually due to some of the preselected respondents’ failure to participate. Nonresponse error limits the generalizability of the findings. Some prospective respondents do not trust the research sponsor and have little commitment to participate. Some respondents also feel that the questions violate their privacy. In some cases the questions deal with a subject that respondents feel is too sensitive. In many cases, the researcher fails to contact planned respondents because they are not home or the contact information is out-of-date or wrong. Sometimes the survey instrument is too long or uninteresting and respondents just give up in frustration part way through the survey. 8. What are the advantages and disadvantages associated with “online” surveys? Online surveys offer speed of data collection and relatively low costs. They also offer more information on the response process because of the meta-data collected in web logs. The disadvantages include relatively low response rates and the limitations due to coverage (not everyone has Internet access). In addition, while probability samples are possible online, many online surveys utilize a self-selected, convenience sample. 9. How might a faulty problem definition error affect the implementation of a mail survey? If the problem is not correctly defined, it is likely that the wrong questions will be asked. It is also likely that the wrong respondents will be questioned. The results of the research will not really address the true nature of the marketing problem. Management action based upon such flawed research may be inappropriate and even counterproductive. The most important step in any research project is to gain a clear understanding of the problem and establish appropriate research objectives. 10. Explain the difference between internal validity and external validity. Internal validity refers to the extent to which the research design accurately identifies causal relationships, while external validity means the results of the experiment can be generalized to the target population. Generally internal validity is enhanced by lab experiments because there is more control over the variables and experimental manipulation. However, this enhanced internal validity may come at the expense of external validity, which is generally higher in realistic settings. 11. What are the advantages and disadvantages of field experiments? A field experiment is performed in a natural setting, which is more realistic but provides a lesser degree of control. That means it offers enhanced external validity but lower internal validity. It also may be expensive and provides the competitor an opportunity to know about the strategies being examined. ANSWERS TO DISCUSSION QUESTIONS 1. Develop a list of the factors used to select from person-administered, telephone-administered, self-administered, and computer-assisted survey designs. The factors relevant in selecting the appropriate data collection method are listed below: • Situational factors ○ Budget ○ Completion time frame ○ Quality requirements—completeness of data, data generalizability, and data precision • Task factors ○ Task difficulty ○ Stimuli needed to elicit the response ○ Amount of information needed from the respondent ○ Research topic sensitivity • Respondent factors ○ Diversity ○ Incidence rate ○ Respondent participation—the respondent’s ability to participate, the respondent’s willingness to participate, and the respondent’s knowledge 2. What impact, if any, will advances in technology have on survey research practices? Support your thoughts. Telecommunications and computer technology have already affected survey research practices. Already, most survey-based marketing research is conducted online. Further, mobile phone surveys are being tested and pursued and will likely experience growth in the near future. Online focus groups are also becoming increasingly popular. 3. EXPERIENCE THE INTERNET. Go to the Gallup Poll site (www.gallup.com) and locate information about the Gallup World Poll. After reviewing the material, make a list of the challenges in conducting polls that represent 6 billion citizens across the world. Students’ responses will vary for this question. Their answers may contain points like the language differences will make the work of researchers difficult because they have to make sure that all participants interpret the question in the way they (researchers) want them to. Coordinating the researches will also pose a challenge because of the number of participants involved. 4. EXPERIENCE THE INTERNET. Go to Kinesis Research (www.kinesissurvey.com) and view the short wireless survey demonstration video. What are the advantages and disadvantages of wireless surveys? Wireless surveys enable surveying at the point of purchase wherever respondents may be. They can deliver many of the benefits of online surveys. However, there is also a potential bias given that respondents may be distracted by uncontrolled stimuli during the survey. It may also be difficult to encourage responses. 5. Comment on the ethics of the following situations: a. A researcher plans to use invisible ink to code his direct mail questionnaires to identify those respondents who return the questionnaire. b. A telephone interviewer calls at 10:00 p.m. on a Sunday and asks to conduct an interview. c. A manufacturer purchases 100,000 e-mail addresses from a national e-mail distribution house and plans to e-mail out a short sales promotion under the heading “We Want to Know Your Opinions.” a. Generally, survey respondents are assured that their input will be held confidential and their identities protected. Secretly coding the survey instruments violates those assurances. The fact that it was done solely for administrative purposes may somewhat mitigate the sin. Why not just be open about the coding and explain the reasoning behind it? Unless the research topic is highly sensitive, most respondents would not be negatively influenced by an administrative action. b. Obviously poor timing. Some people are hard to catch at home, but most of Sunday should be out of bounds. Unless, prior arrangements have been made, this is an unreasonable time. This is not so much a matter of ethics as it is of practical courtesy and consideration. c. Selling under the guise of research (SUGGING) is unethical and a Federal offense. There are no mitigating circumstances that can change this hard fact. It is practitioners like this that give marketing and marketing research a bad name. Such practices create bad impressions and make the task of legitimate market researchers just that much harder. 6. The store manager of a local independent grocery store thought customers might stay in the store longer if slow, easy-to-listen to music were played over the store’s intercom system. After some thought, the manager considered whether he should hire a marketing researcher to design an experiment to test the influence of music tempo on shoppers’ behaviors. Answer the following questions: a. How would you operationalize the independent variable? b. What dependent variables do you think might be important in this experiment? c. Develop a hypothesis for each of your dependent variables. a. The independent variable is music tempo. It can be operationalized as slow, moderate, or fast. However, the experiment may also want to examine specific types of music (classical, jazz, easy listening, and so on). b. The dependent variables could include: • The average amount spent per visit per customer • The average length of store visit • The average number of items purchased per visit per customer c. The following are alternate hypotheses: • The average amount spent per visit will be higher when the background music is slow than when background music is fast. • The average length of a store visit will be shorter when the background music is fast than when it is slow. • The average number of items purchased will be higher when background music is slow than when it is fast. Chapter Six Sampling: Theory And Methods Answers to Hands-On Exercise 1. How many questions should the survey contain to adequately address all possible new menu items, including the notion of assessing the desirability of new cuisines? In short, how can it be determined that all necessary items will be included on the survey without the risk of ignoring menu items that may be desirable to potential customers? The number of questions should be limited to enable participants to complete the survey quickly. Otherwise, participants may be annoyed at the time investment they were asked to make. It is not necessary to examine a set of menu items on the survey. Rather it is necessary to find out whether it is important to customers that new items be offered periodically, and if so, how often, how many items, and what kind of items (including cuisine). 2. How should the potential respondents be selected for the survey? Should customers be interviewed while they are dining? Should customers be asked to participate in the survey upon exiting the restaurant? Or should a mail or telephone approach be used to collect information from customers/noncustomers? An exit interview might be appropriate. As diners receive their checks they could be given a short survey and asked to complete it before they leave. 3. How many new menu items can be examined on the survey? Remember, all potential menu possibilities should be assessed but you must have a manageable number of questions so the survey can be performed in a timely and reasonable manner. Specifically, from a list of all possible menu items that can be included on the survey, what is the optimal number of menu items that should be used? Is there a sampling procedure one can use to determine the maximum number of menu items to place on a survey? First, Santa Fe Grill is a niche player. There is a reason for that. If they try to be all things to all people, they will need to have a great deal more resources to put into their operation. They obviously don’t have that option. Therefore, if they are thinking about expanding the menu, they should think of items that are congruent with, or complimentary to, their present offering. That said, the way to identify a few new items for the menu is a multi-stage process, much like the new product development model. Make a list of all the possible choices for new items. Use secondary sources; a literature review, look at competitor’s offerings, and expert interviews. Also, use primary sources of information; stage a contest to solicit new menu offering ideas, brainstorm your staff, and ask customers for suggestions. Once you have a massive list, start to eliminate the obvious “losers.” Items that are incongruent with your restaurant’s theme, items that don’t fit the rest of your offering’s price range, items that the cooks can’t handle, etc. Once you have drastically cut the list, get your target population to give feedback on their desires. The result should be a prioritized short-list of items that will work in your setting and that your target market finds attractive. How you decide on an optimal menu offering is a business decision, not based on consumers’ tastes and preferences. Make an estimate of contribution to margin that each menu item will generate. Start at the top and work down through the list. Stop when you get to zero. Stop before that if you get to the place that you don’t have resources to support the offering. Include some items that are complimentary to the real profit drivers. There is no “optimum” number of items for a menu. It all depends on the individual situation. Menu items can even change seasonally, to keep up with ingredient availability and/or customer’s seasonal expectations. 4. Determine the appropriate sample design. Develop a sample design proposal for the Santa Fe Grill that addresses the following: Should a probability or nonprobability sample be used? Given your answer, what type of sampling design should be employed (simple random, stratified, convenience, etc.)? Given the sample design suggested, how will potential respondents be selected for the study? Finally, determine the necessary sample size and suggest a plan for selecting the sample units. It is possible to use a probability sample in this case and the results can be generalized to the target population. A systematic random sample is probably adequate for this exploratory research situation. The sampling plan will have something to do with how much traffic flows through your restaurant, the desired sample size, and the resources available to do the sample. Let’s say that you sample the person who pays the bill for every 15th order. The tickets have control numbers, so identifying every 15th ticket (after a random start) should be fairly easy. Sample size is a function of the degree of precision required, the amount of variation within the target population, and the resources available to do the study. We don’t know anything about any of these factors, so it is pretty hard to select a proper sample size. Certainly anything less than 30 would be too small. Anything over 500 would be too costly and cumbersome. Somewhere in that range would be the “right” size. ANSWERS TO REVIEW QUESTIONS 1. Why do many research studies place heavy emphasis on correctly defining a target population rather than a total population? When performing marketing research today, it’s important to insure the people and objects under scrutiny have a direct relationship to the objectives of the research project. A total population is a grouping of people, objects, products, and organizations that are interest to the researcher and client. However, drawing a representative sample from a population such as “all people studying for a Bachelor of Science Degree” or “everyone who orders out for fast food on the weekend” is a challenge, since it’s unlikely the data captured would be accurate or useful at all to Brown University or KFC. In sum, heavy emphasis is placed upon correctly defining a target population in that the focus is placed on a precisely defined subset of the total population of interest to the research team and the client. 2. Explain the relationship between sample sizes and sampling error. How does sampling error occur in survey research? Sampling error is any type of bias that can be attributed to how the sample was drawn or how sample size was determined. Estimated standard error measures the sampling error and gives the research team and the client an idea of how far the sample outcome “strays” from the estimate from the actual defined target population. Estimated standard error is “joined at the hip” with sample size for the following reason: The more variability, and the more precision and confidence desired and the bigger the sample must be. 3. The vice president of operations at Busch Gardens knows that 70 percent of the patrons like roller-coaster rides. He wishes to have an acceptable margin of error of no more than +/- 2 percent and wants to be 95 percent confident about the attitudes toward the “Gwazi” roller coaster. What sample size would be required for a personal interview study among on-site patrons? This is an interesting review question; however, it should be assigned as a take-home question as a way of preparing your class participants for some of the more sophisticated discussion questions concerning sample size below. For this reason it’s advised that you simply point them toward the section which discusses determining sample size and let them “have a go at it”. At this juncture it’s important to encourage them to “do the best they can” and that they will learn more if they come to class prepared. There’s an old saying you should share with your students as you begin to drill down into the more “quantitative” side of our text; namely: “Fifty percent of something is better than one-hundred percent of nothing at all.” Determining sample size, like the many and varied statistical techniques can cause the culture of your class to take a “turn for the worse” if the process of acclimation isn’t managed properly. Unlike the discussion questions below—which come complete with outcomes you can use—approach this review question by “thinking outside the box.” The rewards will be rich when you return to cover the landscape of quantitative analysis later in the term. ANSWERS TO DISCUSSION QUESTIONS 1. Summarize why a current telephone directory is not a good source from which to develop a sampling frame for most research studies. The current telephone directory is a less than adequate source from which to develop a sampling frame because of what’s referred to as an under-registration condition. Under registration occurs when eligible sampling units are accidentally (and in this case purposefully) left out of the sampling frame. For example, in its most basic sense a “telephone” book ought to list names, numbers, and addresses for each and every person who owns a telephone. However, your telephone book does not include people whose numbers were activated after the book has been published, cell phone numbers, pager numbers, and/or unlisted numbers. Therefore, a telephone book should never be used as the primary or only source for developing a sampling frame because of this sampling gap. 2. EXPERIENCE THE INTERNET. Go to www.surveysampling.com and select from the menu “the frame.” Once there, select “archive” and go to a particular year (e.g., 2012) and review the articles available on the topic of sampling. Select two articles and write a brief summary on how sampling affects the ability to conduct accurate market research. This response will vary by students. Students should be able to summarize the importance of sampling in their own words. Chapter Seven Measurement And Scaling Answers to Hands-On Exercise 1. In your judgment, what level of scale design would be the most appropriate in creating the necessary scale measurements for collecting primary data on each construct? If we are going to use a number of measures of product/service satisfaction, price, and quality of customer service to obtain a composite measure of customer satisfaction, an interval scale for each would be appropriate. Actually, if we can obtain a performance metric for each of these factors, then obtain a factor weighting score (by using a constant sum technique for all factors), we could obtain a really good composite score. Use the same approach for the other two. 2. For each construct, design an example of the actual scale measurement that could be used by Burke, Inc. to collect the data. A single-item measure for repeat purchase intention (one of Burke’s components of loyalty) could be: How likely are you to dine at the Santa Fe Grill at least once in the coming month? The response format could be a five-point scale beginning at “not at all likely” to “extremely likely.” 3. What are several weaknesses associated with how Burke, Inc. measured its Secure Customer Index® (SCI®)? Make sure you clearly identify each weakness and explain why you feel it is a weakness. The Secure Customer Index® is seeking to determine customer loyalty. Customer loyalty is not necessarily a measure of customer satisfaction. It is a multi-dimensional construct that is partially a matter of satisfaction, but also consists of the customer’s commitment to buy in the future and their willingness to make purchase recommendations to others. They do ask somewhat the right questions, but it’s not mentioned how they measure the data, how they capture it, or how they calculate their Secure Customer Index®. The ideal way to measure this characteristic of interest is to use an ordinal-interval scale for each of the three dimensions. Decide on a cut-off score for each (not necessarily the same value) and calculate the percent of customers that meet all three standards. Unless there is uniformity in the way data is measured and uniformity in the way data is evaluated, it is not meaningful to compare one store’s SCI® with another’s. 4. If you were the lead researcher, what types of scale measurement designs would you have used to collect the needed data for calculating SCI®? Why? Write some scale measurements you would use. Because the constructs are attitudes and behavioral intentions, the scales in the chapter can be used as scale measurement designs. 5. Do you agree or disagree with the Burke, Inc., interpretation of the value they provide their clients using the Customer Loyalty Index? Support your response. It is difficult to tell whether to agree or disagree with the Burke, Inc., interpretation because it is not exactly told how the index is calculated. Let us say that only the responses of 1 or 2 count as agreement with the first statement. Let us further assume that a response of 1, 2, or 3 on the other two questions would count as agreement. Now it is possible to calculate the percentage of respondents that can be counted as “secure customers” because they responded “in the zone” on all three questions. ANSWERS TO REVIEW QUESTIONS 1. What is measurement? Measurement is the process of determining the intensity (or amount) of information about persons, events, ideas, and/or objects of interest and their relationship to business problems or opportunities. 2. Among the four basic levels of scale measurements, which one provides the researcher with the most information? When compared to nominal, ordinal, and interval levels, a ratio scale provides the most data and information. This is because it simultaneously activates all four scaling properties (assignment, order, distance, and origin). Consequently all forms of data analysis can be used with ratio scaled data. 3. Explain the main differences between interval and ratio scale measurements. Interval scale measurements have the scaling properties of assignment, order, and distance, but not origin. Ratio scale measurements along with the properties of assignment, order, and distance also have the property of origin. 4. What are the major differences between ordinal and interval scales? In your response include an example of each type of scale While interval scale measurements have the scaling properties of assignment, order, and distance, ordinal scales have only assignment and order. Ultimately this affects the data analysis tools that are appropriate. An example of an ordinal scale is: Rank the following television programs in your order of preference such that 1 indicates the show you like the best. An example of an interval scale is: Indicate which of the income ranges best approximates your household’s total income. 5. Explain the major differences between “rating” and “ranking” scales. Which is a better scale measurement technique for collecting attitudinal data on sales force performance of people who sell commercial laser printers? Why? A rating scale asks for the respondent to express his or her attitudes, feelings, or behaviors about an object while a ranking scale asks the respondent to indicate an order of preference about several objects. Rating scales might be a better technique to use here as their main objective is to have a respondent express his or her attitudes, behavior, or intentions about a specific object or event (in this case attitudinal data on sales force performance of people who sell commercial laser printers). 6. What are the benefits and limitations of comparative scale measurements? Design a ranking order scale that will enable you to determine brand preference between Bud Light, Miller Lite, Coors Light, and Old Milwaukee Light beers. Comparative scales require the respondent to make a direct comparison between two products or services, whereas noncomparative scales rate products or services independently. Data from comparative scales is interpreted in relative terms. Following is a design of a ranking order scale that can be implemented to determine brand preference between Bud Light, Miller Lite, Coors Light, and Old Milwaukee Light beers: For each of the following pairs of beers brand name, please circle the brand that you prefer in that pairing. Make sure you select only one brand name per pairing: ANSWERS TO DISCUSSION QUESTIONS 1. Develop a semantic differential scale that can identify the perceptual profile differences between Outback Steakhouse and Longhorn Steakhouse restaurants. Listed below is a possible semantic differential scale design that can be implemented: Thinking about your perceptions of Outback Steakhouse, please select the space that best expresses your opinion about that item: Thinking about your perceptions of Longhorn Steakhouse restaurants, please select the space that best expresses your opinion about that item: 2. Design a behavior intention scale that can answer the following research question: To what extent are college students likely to purchase a new automobile within six months after graduating? Discuss the potential shortcomings of your scale. Listed below is a possible behavior intention scale design that can be implemented: Please check the one response that best expresses how likely you are to purchase a new automobile within six months after graduating from college? [ ] Definitely will purchase [ ] Probably will purchase [ ] Probably will not purchase [ ] Definitely will not purchase The shortcomings include the following: • It is a single-item measure • It asks people to think about a fairly long time period out. 3. For each of the scales shown below (A, B, and C), answer the following questions: a. What type of data is being collected? b. What level of scale measurement is being used? c. What is the most appropriate measure of central tendency? d. What is the most appropriate measure of dispersion? e. What weakness, if any, exists with the scale? A. Example: Raw data being collected here is the mode of payment for travel. This is a nominal question so only an assignment property is activated. The only measure which can be used here is a measure of central tendency, the mode. There are two problems which may cause glitches for the research team after the data is collected: • It might be better to change the question to “What is the most frequent (or preferred) way to pay for your travel expenses?” in order to avoid having to deal with respondents who check multiple (or every) category. • The open-ended line following the “Other” option. This space may not be long enough for a respondent to articulate something of substance for the research team to include in the analysis. What is worse, it asks a respondent to “think”. For any type of nominal scale design that focuses on having the respondent just select one category [or multiple] descriptor, one should try to initially include all possible choices. If the research is unsure that all category choices are included, it is reasonable to add a “some other: please specify _____” choice and have the respondent write in their choice answer. B. Example: The raw data being collected is information on the frequency of doing a specific type of behavior [travel]. The level is interval; the scaling properties of assignment, order, and distance are activated. The most appropriate measure of central tendency would be the mean, with dispersion—standard deviation. One can argue there are two problems with this question: • It is double-barreled, coupling traveling for business and pleasure together when they should appear in separate questions. • The set up portion of the question in this scale needs to include a reasonable time frame [e.g. How often do you normally travel for business or pleasure in a typical year?]. It could be suggested that once the original question is separated into two questions, one might want to use a ratio sc ale design to capture the data responses in order to maximize analytical options later. C. Example: The raw data being captured here is a respondent’s income level before taxes. This is an interval question so assignment, order, and distance are being activated. The measures of central tendency and dispersion will be the same as those listed for Example A, unless of course some of your class participants raise worthwhile arguments to classify the question as an ordinal one. There is a pair of problems with this question: • Some hesitate to answer income questions accurately. • The question might be better rendered as a ratio question since the analytical possibilities are more robust. 4. For each of the listed concepts or objects, design a scale measurement that would enable you to collect data on that concept/object a. An excellent long-distance runner b. A person’s favorite Mexican restaurant c. Size of the listening audience for a popular country and western radio station d. Consumer’ attitudes toward the Colorado Rockies professional baseball team e. The satisfaction a person has toward his or her automobile f. Purchase intentions for a new tennis racket a. Example: State-of-being, mind, behavior, and intention all figure in here. If the excellence refers to “timing” and/or “distance” a ratio question format might be most appropriate from a response and analytical standpoint. b. Example: State-of-mind, behavior, and intention are good candidates here. If the question seeks to define a “choice” within a local competitive environment, a nominal question might suffice. c. Example: State-of-being and perhaps behavior could be argued to figure in here. An interval question would be the best choice from a format standpoint, especially given the ratings issue tied to audience size. d. Example: State-of-mind and perhaps intention qualify as candidates here. e. Example: This is clearly a state-of-mind question. Again, the best scaling properties involve assignment, order, and distance so an interval question/scale design would be a logical choice. f. Example: State-of-intention appears clearly at work in this option. An interval question/scale would be the best candidate. 5. Identify and discuss the key issues a researcher should consider when choosing a scale for capturing consumers’ expressions of satisfaction. Probably the first factor that should be considered is the research objective. Most of the rest of the project, including the measurement flows from this beginning. Next, we might consider the available resources to do the project. Time, money, and talent can serve as constraints in research design. Another important factor to consider is your population of interest—How complex a measure can be applied? For instance, children cannot adequately respond to highly detailed, sophisticated measurements. The scale selected should be appropriate for the data collected. As the scales become more complex, their degree of discrimination becomes greater. Finally, we should consider the reliability of the scale as a measurement of the characteristic of interest. 6. AT&T is interested in capturing the judgments people make of its new wireless cell phone services. Determine and justify what service attributes should be used to capture the performance of its wireless cell phone service. Design two scale measurements that would allow AT&T to accurately collect the data. Three constructs relating to AT&T’s new cell phone services might be: • Coverage [the size of the service area] • Reliability [the signal consistency] • Quality of reception [voice quality clarity] The following are examples of possible performance scales: a. For each of the aspects of AT&T’s new cell phone services listed below, please indicate the response that best expresses your evaluative judgment of how well AT&T is performing on that aspect (Scale = 5 points; 1 = Terrible, 5 = Outstanding). Coverage Reliability Reception quality b. How satisfied are you with AT&T’s cell phone services? (1 = Very unsatisfied, 5 = Very satisfied) 7. The local Ford dealership is interested in collecting data to answer the following information research question: How likely are young adults to purchase a new automobile within a year after graduating from college? Design a nominal, ordinal, interval, and ratio scale measurement that will enable the dealership to collect the required data. In your opinion, which one of your designs would be most useful to the dealership and why? Students’ responses to this question will vary. Possible answers are listed below: The main information question is “How likely are young adults to purchase a new automobile within a year after graduating from college?” Nominal Scale Design: How likely are you to purchase a new automobile within one year after graduating from college? [ ] Do intend to purchase [ ] Do not intend to purchase Ordinal Scale Design: Please check the one response that best expresses how likely you are to purchase a new automobile within one year after graduating from college? [ ] Definitely will purchase [ ] Unlikely will purchase [ ] Probably will purchase [ ] Definitely will not purchase Interval Scale: Which one of the following percentages ranges best approximates the likelihood of you purchasing a new automobile within one year after graduating from college? [ ] less than 10% chance [ ] 11% - 49% chance [ ] 50 – 79% chance [ ] 90 – 100% chance Ratio Scale: In the space provided below, please write the percentage value that best approximates the likelihood [chance] of you purchasing a new automobile within one year after graduating from college. Solution Manual for Essentials of Marketing Research Joseph F. Hair, Mary Celsi, Robert P. Bush, David J. Ortinau 9780078028816, 9780078112119