WRITTEN CASE DISCUSSION GUIDES

Instructor's Resource Manual

Written and Video Cases

Written Case Discussion Guides

Cases can be an asset in teaching marketing research methods as they involve the student in the decision making and analysis processes. High levels of student involvement are positively correlated with student learning and concept retention.

Some instructors prepare for case discussion by developing detailed answers for case questions. In this section, we have provided you with answers to the questions that conclude each case.

Other instructors prefer to prepare lists of questions rather than answers, with the order of the questions designed to build the student's understanding of the complex issues revealed by the case. As the students answer the questions the class delves more deeply into text concepts.

Still other instructors prepare anticipated student answers and plan the tangents to which those answers will lead. Lecture material is planned for the tangents.

One guideline has served case users well: be prepared to take advantage of student learning opportunities to which an unexpected student answer may quite naturally lead.

Case: AgriComp

Abstract: AgriComp, a supplier of computer systems for farmers, has surveyed it dealers on whether to change its procedure for settling warranty claim disputes. Currently local dealers handle warranty services for customers via local repair followed by a reimbursement claim to Agri Comp. Denied claims follow an internal company appeal process. Dealers have been complaining about the fairness of the appeal process and in a recent survey were asked to respond to an alternative process, an impartial mediator. The student is asked to review survey results and determine whether the costly external mediator process would be worth implementing to keep the dealers happy.

This case offers a chance for students to deal with the data before it's crosstabulated. Nothing very fancy is required, but the students will need to recognize that a crosstabulation is in order (or at the very least that some separate tabulations are needed). The dealer preferences are different for those who have used the existing appeals process than for those who haven't. The more they've used the process, the less they perceive a need for change. If the data are tabulated in the aggregate, this trend is not apparent.1. Jody wonders just how important the process is to the dealers? Was there widespread discontent or had he just heard from a few malcontents at the dealers' meeting?

You can start with a question like "How do the dealers feel?" and follow it up with "Do all of the dealers feel that way?" This will normally bring out a cross tabulation or something like the series of MINITAB dotplots given below. Then you should turn the discussion to how best to summarize or display the conclusions. Either some sort of crosstabulation (using appropriate percentages instead of counts) or plots like those below will work. The main point is to be sure students don't simply declare a "significant" lack of independence and let it go at that. They should have to say something about what kind of dependence they find, not just assert the absence of independence.The more the dealers have used the existing appeals process, the less they agree with the statement that it should be replaced, so it appears that Jody was hearing from some malcontents. The cross-tabulation of responses by number of uses is given on the next page. The corresponding chi-squared is 82.16 on 12 degrees of freedom, so something is clearly going on. Students may offer a variety of summaries of just what is going on, and you should prompt them for such summaries if all they offer is chi-squared. The general trend is illustrated by such diagrams as the dotplots given on the page following the crosstabulation.

Crosstabulation (count) of REP by USE

USE

REP 0 1 2 3

+------+------+------+------+

1 | 12 | 12 | 4 | 6 |34

+------+------+------+------+

2 | 6 | 27 | 18 | 12 |63

+------+------+------+------+

3 | 8 | 18 | 16 | 18 |60

+------+------+------+------+

4 | 4 | 9 | 8 | 52 |73

+------+------+------+------+

5 | 1 | 12 | 12 | 37 |62

+------+------+------+------+

31 78 58 125 292

USE (percent)

REP 0 1 2 3

+------+------+------+------+

1 | 38.7 | 15.3 | 6.9 | 4.8 |34 11.6%

+------+------+------+------+

2 | 19.4 | 34.6 | 31.0 | 9.6 |63 21.6%

+------+------+------+------+

3 | 25.8 | 23.1 | 27.6 | 14.4 |60 20.5%

+------+------+------+------+

4 | 12.9 | 11.5 | 13.8 | 41.6 |73 25.0%

+------+------+------+------+

5 | 3.2 | 15.5 | 20.7 | 29.6 |62 21.2%

+------+------+------+------+

31 78 58 125 292

100.0 100.0 100.0 100.0

Dotplot of REP by USE

USE 0

:

: :

: : :

: : : :

: : : : .

1 2 3 4 5

USE 1

.

:

:

:

:

: :

: :

: :

: : : :

: : : . :

: : : : :

: : : : :

: : : : :

: : : : :

1 2 3 4 5

USE 2

:

: :

: :

: : :

: : :

: : : :

: : : :

: : : : :

: : : : :

1 2 3 4 5

USE 3 (Each dot represents 3 points)

:

:

: .

: :

: :

: :

: : :

: : : :

: : : : :

1 2 3 4 5

Case: BBQ Products Cross Over the Lines of Varied Tastes

Abstract: This case asks students to assess measurement and scaling issues in the context of the introduction of a frozen, microwaveable BBQ product line into the southeast by Rich Products, Buffalo, NY. The new line is being introduced with commercials depicting Ruby, a fictitious waitress at Pork-O-Rama who prefers the taste of the new frozen line.

This case deals with measurement and scaling issues. The student must select a method to measure attitudes toward the product category and the specific brand. In addition, the student must decide on a measurement method to measure preference among a number of brands in the product category. Student may attack issues such as validity and reliability of the measures.

1. What measurement and scaling issues should be considered when developing a study to measure consumers attitudes toward barbecue in general and specifically Rich Products Barbeque?

Chapter 14 presents difference methods for measuring attitudes. The student should select a method and defend why they chose that method. Their defense should rest on the issues of validity and reliability. Some additional issues that may be addressed are:

A. How will the marketing managers use the attitude measures? If they simply want to determine the overall attitude toward Richs barbecues then different approaches are available. However, if they are interested in developing advertising copy, employing segmentation, or altering the product, then the measurement device must collect information on the salient attributes. In essence, the student must decide whether they want to collect formative or reflective indicators of the attitude.

B. A fundamental issue for the collection of attitudes for Richs product is consumer awareness of the brand. The brand is new; consequently, will people have an attitude toward it? Will the company need to develop concept storyboards, allow consumers to try the product, etc.?

C. Another issue is the target market. It would appear that some consumers will have well developed attitudes toward barbecue, but others will not. The question is: Can the same instrument or measurement device be used for both types of consumers?

2. Assume Richs wanted to test peoples preferences for their barbecue versus the other leading brands (of which there are five). What would you recommend to measure these preferences?

A. The first choice the student will have to make is whether to use a comparative or noncomparative methods. Comparative scaling results in data that must be interpreted in relative terms and has ordinal data properties. An attractive feature of comparative scaling is that relatively small differences among objects being compared can be detected. It must be remembered that the respondent is instructed to directly compare objects; consequently, differences are forced to surface. Comparative scales are, in general, easily understood by respondents but can become time consuming as the number of objects to rate increases leading to respondent fatigue.

B. With noncomparative scaling the respondent is not instructed to compare the object being rated against either another object or some specified standard. Hence, small differences among objects may not surface.

C. Once students have discussed these issues they should choose a particular method and defend their choice. The advanced student may want to discuss other potential analysis for the data such as multidimensional scaling approaches.

Case: Calling Up Attendance

Abstract: A study by Prince Marketing for TCS Management Group, which markets TeleCenter System software, measures customer satisfaction with the current edition of the software and aims to predict attendance at a two-day educational event, Users Forum.

This case has numerous tie-ins to various chapters. Used with Chapters 4 and 5, you can build the management-research question hierarchy. Used with Chapter 8, your students can build the research process model up through data collection. Used with Chapter 16 and 17, you can discuss sample frames, and screening for qualified respondents, as well as various methods for drawing a sample from a sample frame (customer list). Used with Chapter 13, you can discuss the types of data being collected as well as reliability and validity issues. Used with Chapter 14, you could discuss why a rating scale is appropriate to evaluate customer service and likely attendance. Used with Chapter 11, you can discuss the appropriate communication methods. Used with Chapter 15, you could discuss preliminary analysis planning. Used with Chapter 18, you could discuss what to do with the 16% of the sample that felt ill-equipped to answer the issue questions and what preliminary analysis would be appropriate to answer the management questions.

1. This question gets the student in the habit of formulating the management-research question hierarchy (Chapter 4) as the beginning step of a research project. The management dilemma facing TCS is how to prepare for the scheduled Users Forum when it doesn't know what specific presentations to build into the program, nor how many people to expect at the Opryland Hotel event.

Management questions: What topics should be addressed in the presentations or materials distributed at the Users Forum? How should solutions be delivered, given the attendance expected (written materials? whole session devoted to the issue?)?

Research questions: What issues cause the most concern among the current users of the TeleCenter System software? How many current users see attendance at Users Forum as a viable means to address troublesome issues, such as ease of use, technical support access and responsiveness?

Investigative questions: What is the current users' evaluation of customer service, especially on ease of use, software-generated reports, technical support service, and effectiveness of software for its intended purpose? Who, if anyone, from a current user's company will attend the Users Forum? Do new users of the software have different concerns than more established users? How many representatives from a single company may come to the Users Forum? What type of respondent (user or manager) is most likely to attend the Forum? Do the different types of respondents have different concerns with TeleCenter System software?

Measurement questions: We don't have the actual instrument, but we know that Prince Marketing collected information relative to investigative questions using a 7-point rating scale, where "7" was the most positive on the scale. We also know they were able to classify the respondent as a software user or call center manager.

2. The obvious choice, as indicated by Exhibit 11-1, is the telephone interview. Because TCS needs the information is a relatively short amount of time, one must exclude personal interviewing as a collection method. Given the technical skills of the respondents, self-administered via computer was a possibility, if TCS had e-mail addresses or Prince felt the response timeframe would be met. Prince Marketing chose a phone interview because of the ability to control the process and meet their three-week promise, but also because of the limited number of data variables they were collecting. They could choose any of the three methods (human, CATI, or computer-delivered), but human-administered definitely gives Prince and TCS the most control. CATI, however, gives the telephone interviewer the ability to enter the responses in process, permitting Prince to tally the results more quickly, possibly making them available to TCS in real time, so they could prepare for presentations without waiting for the formal estimate of attendance.

3. The request for a preliminary analysis plan relates well to Exhibits 15-2 (and 15-1) and would be fairly simple for this study. TCS would want to frequencies on the attitudinal-scale variables. And we would want to cross-tabulate such variables by whether the respondent definitely would be, might be, or definitely would not be in attendance at the Users Forum. Frequencies (or cross-tabulated frequencies) would tell us the most pressing issues, assuming Prince has done a good job of anticipating issues in the design of the data collection instrument or was able to transform the data during data preparation into homogenous groups of responses on issues. We also want to be able to predict attendance, so we would want frequencies on the attendance variable, as well as their interest in more information about attending the Users Forum. Hopefully we can cross-tabulate this information against classification variables such as industry of the firm and years of software use that would be readily available from the customer database the TCS provided to Prince Marketing.

4. In data analysis, the new users who felt ill-equipped to evaluate the software ease of use question (16% of the sample), would likely not be included (counted as missing cases) in the analysis of this issue. However, their likely attendance could still be important to the prediction of attendees overall. If some of these new users did identify software installation, technical support, or use concerns, you could cross-tabulate a status variable (new user vs. established user) against the various issues variables. Such data exploration actions may very well lead to Users Forum sessions exclusively designed for new users.

5. You can use this question to discuss the practice of sugging (sales under the guise of research).You could also tie this question back to Chapter 7 and the subject's right to privacy. The issue of incorporating marketing promotion within the context of customer satisfaction research is commonly done, although it can be a two-edged sword. From a marketer's perspective, you have a potential attendee on the phone and you could efficiently use his or her time by telling them about the Users Forum. From a professional researcher's perspective, such a combination use of the survey is totally inappropriate and contributes to the growing refusal rate among potential respondents. If research is accompanied by a sales presentation, potential respondents have the right to be angry when they agreed to participate only in a survey. Such a change in orientation and purpose is a violation of the subject's right to privacy.

Case: Campbell-Ewald Pumps Awareness into the American Heart Association

Abstract: You wouldnt think that an organization that does as much good as the American

Heart Association would have low awareness, but at the start of the described research program its unaided awareness level was just 16 percent. For a company reliant on contributions, low awareness is a major problem. This case profiles the research behind the American Heart Associations first-ever paid advertising campaign. www.campbell-ewald.com; www.americanheart.org1.After watching the ads on your text DVD, take the Learn and Live quiz. Take a position on the tracking of individuals taking the quiz as a measure of direct response of the ads effectiveness; defend your position. (http://www.americanheart.org)

You might find it useful to use this discussion question as a class exercise. You can divide your class into pro/con teams and have each come up with the arguments for that position.

Arguments for using involvement with the online quiz as a means of ad tracking might include:

That getting audience members to the Web site was clearly an important action objective of the ad.

Engaging visitors by involving them in the taking of a quiz is more likely to reveal their understanding of the underlying message of the ad campaign.

Arguments for not using involvement with the online quiz as a means of ad tracking might include:

That many individuals that see/hear and understand the message may not be encouraged to visit the Web siteso the communication objective may be met while the action objective may not.

Individuals that see/hear the ad and understand the message may not have access to a computer. Since computer access was not a variable in choosing the media for the campaign to exclude this portion of the relevant population when tracking ad results creates error.

2.What are the advantages and disadvantages of using secondary data in this particular case?

The primary reason Campbell-Ewald (C-E) used secondary data during their preparation of the AHAs pitch for the advertising account was to gain a better understanding of the organization and of heart disease, as well as understand what motivated giving to a non-profit organization. Using secondary data shortened the time they would spend on research. Also, they quickly discovered the scope of the potential problem facing AHA. While they wouldnt have wanted to craft a creative strategy based solely on secondary data, effective use of secondary data collected during exploration is actually expected in marketing research. It was the interview of John Paling (primary data) about how people perceived risk, however, that gave them the insight to develop the approach that won C-E the AHA business.

3.Discuss the role of Diagnostic Research in the American Heart Associations selection of the personal stories creative approach.

Diagnostic Research was hired to pretest the various creative approaches that evolved from the focus group discussions, the interview with John Paling, and the secondary data search. DR showed the animatic prepared by C-E to a sample of 1000 adults in a mall setting. The AHA animatic was positioned within a pod of animatics for other products and services, similar to the way the audience would see the real AHA commercial. Interviews that followed measured brand recall and the main idea that the audience got from the ad.

The DR methodology indicated that all three creative approaches tested would generate appropriate brand recall and message capture. That meant the C-E might need additional research to correctly select the best creative approach or they would need to use their vast experience to recommend the better approach for AHA to take at this particular time in AHA history. This is good time to mention that even great research doesnt make the decision for the manager. Rather, research provides the undecided manager with evidence to sway his judgment in one direction or another, or it substantiates a direction in which the decision-maker might be leaning.

4.Explain why both qualitative and quantitative research techniques were used to develop the personal stories campaign?

The use of more than one research technique in building an advertising campaign is fairly common in marketing research. The quantitative data derived from previous AHA surveys and from the DR animatic testing allowed C-E to quantify the risk of each campaign. It was the qualitative research, first the interview with John Paling that gave C-E its understanding of personal risk, as well as the focus groups that put a face to real, heart-disease stories that gave C-E their strategic approach to the problem of raising AHA awareness and that, ultimately, provided the winning campaign creative approach.

You might want to discuss whether the mall setting was conducive to either the viewing of the animatic or whether patrons of malls represent an appropriate group from which to draw a sample. If DR has facilities in the mall, as many such firms do, then the group from which it draws its sample might suffer from over-testing and thus inject error into the process.

Depending on how much you wish to discuss advertising, other discussion opportunities include:

Discuss how students would test which of the personal stories ads was most effective.

Discuss whether the appropriate sample should have been weighted more toward females than males.

Discuss what research could be used to determine whether an ad campaign would be more effective in raising awareness and understanding of the AHA, compared with other promotional methods.

Case: Campbell-Ewald: R-E-S-P-E-C-T Spells Loyalty

Abstract: The case describes a detailed study conducted by Campbell-Ewald, who sought to substantiate the importance of respect when dealing with customers in different business related sectors. This agency, steadfast in its success through customer satisfaction principles, establishes through its research that the construct of respect is held in high regard by customers and therefore plays a significant role, when determining, among other things, the longevity of client relationships.

1. How would you operationally define the construct of respect? Take the perspective

of each of the three client sectors: insurance, air travel, and retail.

This question is fitting for discussions concerning concepts and constructs. It prepares students to identify the similarities and differences involved with the use of both scientific terms (Chapter 3) and how dictionary meanings differ from those used in testing. Both terms are inherently similar yet the construct has specificity necessary for measurement and replication of a study. Constructs are not easily observable, created for the express purpose of testing, and generally more complex. The construct of respect though similar in various theories, assumes variations when operationally defined in the research literature and when viewed from the perspective of these three client sectors. The term respect as used in conversation has more individualized experiences attached to it and is idiosyncratic to individual speakers. Thus it is difficult to find agreement among students in this definition during a discussion. You will find that through brainstorming, students may indicate that the definitions overlap and they may point out difficulties while trying to operationally define them. When such constructs are used in pure research, measurement scales are essential to further specify meanings. The important thing for students to grasp is that through operational definitions, a person should be able to measure or conceptualize the construct, given the specific purpose of the research. Advise students that operational definitions often serve the same purpose for researchers, as common jargon does for narrow language groups. Here are some student definitions of respect after reading the case:

Insurance: tailor products for specific customer needs; maintain the individuals right to privacy when dealing with client information; show concern for unfortunate mishaps; provide contractually correct compensation in a timely fashion, refraining from unethical and dishonest conduct.

Air Travel: strive to meet expectations in check-in, boarding, onboard services, debarking, and luggage handling; provide appropriate reimbursements or compensation to customers when necessary; show concern for safety and on-time schedules.

Retail: maintain a polite and cordial demeanor when dealing with the public; listen to customers specific needs and match your product accordingly; remain professional when faced with unpleasant circumstances, regardless of who is at fault.

2. Map the overall design of the research described here.

The nature of this question gets students into the habit of identifying the different types of designs and the steps involved when conducting research. It allows students to become more familiar with the various methodologies utilized and is an opportune time to explore data collection techniques and the constraints researchers sometimes encounter when utilizing such techniques. As such, this question is quite applicable to Chapter 8, which discusses design strategies as well as 11, which explores data collection methods.

The first step involved Campbell-Ewalds exploratory attempts to discover the reasons for the disconnect between respect bonds and its customers. Campbell-Ewald discovered firms were accustomed to using Customer Relationship Management (CRM) software, as a mechanism for tracking satisfaction behavior among customers. However, research revealed that investment in technology and software to model likely behavior was not achieving its desired goal. Among the reasons for Campbell-Ewald interest was the apparent connection between customer satisfaction and customer loyalty and purchase behavior. This stage also tapped into feedback from Gartner Group and Accenture, relationship experts such as Dr. Phil and Steven Covey; and information garnered from existing literature on the dynamics of people relationships.

This is an opportunity to ask students what other exploratory avenues that Campbell-Ewald might have pursued.

This is also an opportunity to discuss the difference between pure and applied research. The Campbell-Ewald clients in the three sectors did not have specific relationship management issues they were researching when they signed on as part of the respect research initiative.

The second stage of the research(data collection(revealed the use of two levels of primary data. Through partnership with Synovate, Campbell-Ewald first utilized Synovates omnibus Telenation (nationally representative telephone survey) to ascertain whether the public viewed respect as an issue when they dealt with companies. This data was used to justify Campbell-Ewalds belief that respect was a salient issue to customers. Here you can discuss the implications involved with using an omnibus study and how the process of asking a few questions to a national probability study can prove helpful to a researcher. Campbell-Ewalds decision to explore this data may have yielded significant findings, which further validated their conclusion that respect truly matters. You can ask students their view of how credible the use of an omnibus study data is and why.

To establish the varying perspectives of respect concerning the different sectors (insurance, airline and retail) Synovate utilized a sample of 12 focus groups, four each in the above named sectors. This stage explored the scope of the research question and seeded the subsequent 4-page questionnaire. To better allow for a representative sample, the focus groups included men and women from both Chicago and Detroit. This is an apt time to discuss the advantages and disadvantages of using focus groups. Considering the sample usedfour each in three sectorsraises questions about accuracy and representative results from two market areas. Feedback from the focus groups was used to design measurement questions for the mail survey sent to Campbell-Ewalds clients customer lists from each sector. The focus groups were used fairly traditionally in this case to see a more detailed quantitative study.

The final data collection phase of the research was a national probability mail survey conducted by Synovate. One possible discussion question is: Why did Campbell-Ewald choose NOT to use the Synovate panel to increase the response rate to the survey. [Campbell-Ewald feared it wouldnt be representative of the three sectors.] Another might relate to the non-response error inherent in a mail study and how it is handled. With only a small response rate (less than 9%) could Campbell-Ewald have concluded that respect matters to only a small portion of the relevant population and the remainder simply didnt care? Generally this is a good time to discuss the strengths and weaknesses of mail surveys.

A.What types of studies were involved in Campbell-Ewalds respect initiative?

There were three study types involved in Campbell-Ewalds respect initiative. You can use this time to ask students how they would relate the case to other types of studies. You can also discuss the characteristics of these studies and the purposes each serve in research. Given the outlines of each study, you can ask the students to indicate its strengths and weaknesses.

Exploratory study. Campbell-Ewald used secondary data searches to define the construct (including studies published by Gartner Group and Accenture, books by authors such as Dr. Phil and Steven Covey, and information from other existing literature), and focus groups to help refine attitudinal statements to be used in later quantitative research. The twelve focus groups study completed the exploratory phase of the research and segued into the survey.

Descriptive Study. Campbell-Ewald utilized data from Synovates Telenation telephone survey, which served as a preliminary step in the large quantitative study which followed.

Explanatory study. Here Campbell-Ewalds research team attempted to isolate the components of respect and examine its relationship to satisfaction and loyalty. They were interested in discovering how customers viewed the issue of respect when dealing with companies. They aimed to ascertain what caused customers to remain loyal to businesses and in so doing "identify the tenets of strong personal relationships." This effort aided in the creation of its five "People Principles." With its extensive clientele, which included customers in different sectors (insurance, travel, retail and government) Campbell-Ewald sought to discover, based on their five principles, how respect was viewed in each sector. To build on this, Campbell-Ewald capitalized on Synovates experience in conducting large scale mail surveys, which was a significant source in the data collection process. Its sophisticated analytical techniques, namely its proprietary brand propensity model called The Momentum Engine aided in providing more detailed information in analyzing relationships with sales growth. Given its varying client list, therefore its sample size, Campbell-Ewald desired a more comprehensive feedback.

B. What are the strengths and weaknesses of the various methodologies?

This question is applicable to Parts 2 and 3 of the text where the methods of primary data collection are presented.

Mail Surveys (Self Administered Surveys): The use of this methodology, though often cost effective in research, can be problematic when it comes to response rates. Mail surveys are typically inexpensive, however it is not always a guarantee that responses will be substantial. Five thousand surveys were distributed, and the amount returned (5% 9 %) opens the discussion to nonresponse bias. Response rates for mail surveys are typically low, but steps can be taken to adjust this. This is an apt time to discuss the pros and cons of using different survey methods and the challenges researchers sometimes encounter. Was the survey too lengthy at four pages? Did that contribute to the low response rate. Were participants just not interested in the whole issue of respect from the companies they dealt with? The size of the dataset for each sector (200-500 cases) obscures the fact that while the data are sufficient for statistical analysis, representativeness is in question. You can ask the student:

What steps could Synovate take to increase response rates?

How can a researcher minimize nonresponse error?

How effective would incentives be given sample size and cost?

Campbell-Ewalds decision to disguise their clients, by sending surveys under Synovates letterhead could have increased or decreased response rates. By disguising the surveys sponsor participants may be less inclined to complete the survey. Many might have been unfamiliar with Synovate, further reducing their likelihood of response. Some participation in any survey is dependent on participants recognizing the value they bring and on the rapport between participant and researcher; participants may be more likely to participate in the survey when they are familiar with the organization or feel some sense of loyalty to it. At the same time, disguising the name may have proven helpful, as this would give respondents the opportunity to give an honest opinion of how they felt Campbell-Ewalds clients treated them, or how they viewed the issue of respect.

Focus Groups: The use of focus groups as a way of developing measurement questions that would clarify the core motivational drivers in categories being studied was a useful way of defining the parameters of respect. However, clients being actively involved in this process of creating the measures and the subsequent development of the survey could potentially create some bias. As such, measurement questions could be created that only define how specific clients may conceptualize respect and not necessarily be applicable to all member firms in a sector. Since a communication agency rarely represents more than one firm in a category, having the measurement questions apply more to one firm (e.g., Continental) in a category rather than all other firms in the sector is really not problematic. The use of mainly top customers as the sample frame, begs the question as to how representative that sample may be for pure research.

Telephone Omnibus Survey: An omnibus survey has a very short turn-around while it limits the number of questions contributed by any one sponsor. Therefore, the omnibus can sometimes represent a bundle of questions that have little direct connection. This leap from topic to topic can sometimes confuse a participant and, as a result, the sponsor may get less than the clarity hoped for. For an undefined construct like respect, the omnibus survey should have been a comfortable respite from more product/store/issue specific questions. The lower cost and shorter turn-around would have prevented the larger study from being slowed.

C. How do the methodologies complement each other?

The use of the focus groups for ideation and creating measurement questions is a useful way of getting first hand knowledge as to the type of issues that customers were most concerned about in each sector when it came to respect. Using feedback from the focus groups, Synovate was able to create and appropriately tailor the measurement questions that participants could relate to. Campbell-Ewalds decision to actively engage its clients in this process further enabled them to seed the survey questions. Additionally, the incorporation of a 5-point scale complements the attitudinal nature of the investigation.

3.Analyze the use of the 5-point scale for measurement of respect dimensions. What other statements would you add to the sample provided in Exhibit C-E-1.

This type scale as a measurement tool is a simple and useful way of recording responses. With the attitude statements connected to a five-point strongly agree to strongly disagree scale, comparisons between sectors is more efficient for data analysis of attitudes. Given the length of survey, this is an easy way to allow participants to indicate their opinion, while providing them with a middle ground, in the event they neither agree or disagree.

Though the scale channels a response within a specific context, participants responding to a complex construct such as respect may either add to measurement error by responding to statements with which they have no experience or desire to make additional written clarification.

Using some open-ended questions could have been advantageous. A useful comparison is the evaluation surveys that students complete at the close of each semester, where they rate their professors and the course. To generate a discussion you can ask students how likely they are to complete open-ended questions on such surveys, and why? You can also ask students to suggest statements they would include in the Campbell-Ewald study as opened-ended questions.

Campbell-Ewald and its partners used correlation, multiple regressions, and other multivariate analyses in an effort to identify customers with the greatest potential to drive sales growth. Nonmetric scales would have presented severe data analysis problems since they lack the power to extract the information required by the research question.

Exhibit C-E-1 Sample of Attitudinal Statements

Researchers asked participants their degree of agreement with several respect validation statements. A sample of these statements, developed from focus group discussions, appears below.

StronglyAgreeNeitherAgree norStrongly

Agree

DisagreeDisagree

General Respect Initiatives, across categories

Is interested in listening to what is important to me as a customerPlaces their own interests above those of the customer

Notifies me in advance of sales and special promotions

Rewards me for repeat business

Honor commitments/promises theyve made to me

Specific Respect Initiatives, with categoryAccepts returns without a hassle (retail)

Flights take off and land on schedule (airline travel)

Handles claims in a timely manner (insurance)

4. Discuss the pros and cons of the sampling plan for the mail survey.

Campbell-Ewald received names of customers from their clients. These clients provided names mostly from their top customers. Synovate distributed five thousand surveys to each sector: insurance, air travel and retail. However, the case did not indicate as to how customers were chosen as part of the sample to determine if this was done in a randomized manner. It also did not indicate if criteria were set to choose customers, such as length of time with a particular client or company, their gender, or age groups.

Considering that clients provided names mostly from their top customers, establishing whether or not this is a representative sample of their customer population is pertinent information. Subsequently, with the target population being mostly top customers, this potentially limits the ability to make inferences about the general population, and inevitably raises questions as to the accuracy of the findings. Given the nature of the research, and the fact that five thousand surveys were sent out to each sector with a small return, it is safe to assume that not every customer relevant sampling element is represented. This is a good time to discuss with students, the pros and cons of using probability versus nonprobability sampling. Students may argue that this was the most cost effective method to take, while others may argue that the sample produces much bias and therefore questions the accuracy of the results.

Case: Can Research Rescue the Red Cross?

Abstract: The American Red Cross, known for its worldwide efforts in providing disaster relief and assistance to the sick and needy, comes under intense scrutiny and attack for the misappropriation of post September 11 donations. The case describes the concerns expressed by board members of the Red Cross with regard to how such problems could be prevented in future fundraising efforts. It also outlines how this organization sought to address its existing dilemma of changing the publics perception of the way in which funds are managed and whether public donations were being used appropriately.

This case is quite applicable to Chapters 1 through 4, where students are introduced to the research process and the different types of studies that the Red Cross could have undertaken to avoid their existing dilemma. The case also ties in to Chapter 5 and prepares students for the creation of the management-research question hierarchy (Exhibit 5-4, Management-Research Question Hierarchy). Additionally, it allows for discussions on the decisions made by researchers when designing samples (Chapters 16 and 17). By describing and discussing the dilemma faced by the Red Cross, students will be able to use Exhibit 5-6, Formulating the Research Question, to identify management dilemmas and, in so doing, formulate appropriate research questions that the Red Cross could consider for discovering better ways of funds management. Used with Chapters 13 and 14, students can discuss the types of measurement scales. In addition, the case ties into the data collection methods described in Chapters 11 and 15).

1. If you had been McLaughlin or Decker, what research would you want done?

You can begin discussions of this question by having students suggest how McLaughlin or Decker might define the dilemma. Some may suggest that the dilemma faced by the Red Cross lies in how to better manage and appropriately make use of public donations. Others may see the dilemma as how to overcome the negative perception that donors may have of the Red Crosss fund raising activities and regaining public trust. Others might see the dilemma as one of education of the public and how best to do such activities. Students may also suggest that the Red Cross needs to find ways to reassure donors that their donations will be distributed appropriately. Both McLaughlin and Decker could benefit from exploring how current funds are pooled and distributed, how informed donors are of the ways in which the Red Cross manages and distributes donations. Management could also explore if its decision to use a portion of the Liberty Fund as reserve for future disaster was consistent with its mission and previous fundraising strategies.

2. Create the Management-Research Question hierarchy for the research you think might help the Red Cross make decisions related to public relations efforts and future advertising soliciting donations.

Exhibit 5-4 and 5-5 are useful as a guide for this discussion:

We suggest using the management dilemma facing the Red Cross as the publics perception of its mismanagement of donations. The Red Cross must change how it is perceived by donors who may question whether their donations are being used for the causes for which the Red Cross was perceived to be soliciting funds. The Red Cross must be able to provide services based on the contributions it receives from public donors. If it were to continue to be seen as engaging in activities inconsistent with its mission, the Red Cross could face a significant decrease in its funding sources. This may also place a negative stigma on the nonprofit charitable sector and its future fundraising activities. Therefore, the Red Cross must regain and maintain its credibility in the publics eyes, so as to be able to ensure the continued receipt of donations.

Some of the options students might generate are:

Management QuestionsResearch QuestionsInvestigative Questions

All management questions are not equally addressed by research. However, considering the gravity of the existing dilemma, how should the Red Cross make donors clearly aware of their policies with regard to how contributions are expended?How should the Red Cross change the way it advertises its policies with respect to donation allocation?

What is donor understanding of the existing policies relating to how donations are raised and spent?

Should the Red Cross consider revising its policies relating to how donors contributions are pooled and distributed?

Should the Red Cross use separate fund-raising activities for each disaster, or should it solicit funds only for the general fund that may be allocated to any disaster where needed? What practice (specific fund or general fund) is more likely to generate more largesse among donors?

What level of trust do donors have that the Red Cross will spend its donations wisely?

What are the donation patterns: primarily to specific pleas for help or have donors simply donated to further the general mission of the organization?

How can the Red Cross better manage the funds it receives from fundraising activities, specifically funds raised in excess of the amount needed for a specific disaster?Should the Red Cross utilize a predetermined donation goal, redirecting funds to the general fund after that goal for donations is received, and communicate this goal to the potential donors? How is a dollar estimate of a donation need level determined?

What is the potential donors understanding of this estimation process?

Will a dollar donation goal for a particular disaster relief fund discourage donations once that goal is reached?

How would potential donors want to learn about progress toward goal achievement?

3.If you created a RFP, what would it contain?Chapter 6 introduces the RFP, which is supplemented by the sample on your text DVD, and provides additional insight as to the contents and assessment of a Request for Proposal. The proposal would contain:

Definition of the problem facing the Red Cross (the management dilemma)

Identifying the limitations involved

Providing a description of the policies relating to Red Crosss fundraising

Guidelines

4.What considerations should influence sampling decisions in any research the Red Cross would do on this issue?

Given that the American Red Cross is a national organization and its reputation was at stake, the selection of an appropriate probability sampling technique is indicated.

The larger question is whether sample units should be stratified in some way. Students should be encouraged to brainstorm all the possible subgroups within the donor population. Many donors may have made contributions to the Red Cross via its Web site, where they were given the option of contributing to a particular cause. To allow for accuracy and effectively address the issue of the existing policies relating to how donations are raised and spent, a sample could be drawn from persons who have been previous donors to a particular fund. Persons who visit the Web site for general information and not for donation purposes could also be targeted as possible sampling units.

Students should also be encouraged to identify the sample frames available to the Red Cross due to its past fundraising practices. Make sure that students dont just focus on Web site contributors as a sample frame when making sampling decisions, as the Red Crosss population of donors extends beyond Web site contributors.

Once students have focused on a type of probability sample, they should also consider sampling procedures for drawing that sample. For example, if they have decided that a stratified sample is appropriate, and that the larger "former non-profit donor population" is the basis for sample unit selection, how might such donors be screened to determine whether they have donated previously to the Red Cross general or disaster-specific funds.

5. If a survey is used, what scales would be most appropriate?

You can use this question as a group activity where students can formulate or use the class-generated investigative questions to argue for specific types of measurement questions. One debate to expect is the use of ranking versus rating scales, or for specific types of ranking or rating scales, for example Likert versus paired comparison scale. In Chapter 14, the section on Selecting a Measurement Scale along with Exhibits 14-2 and 14-9 provides the context for decision making.

Students should also be challenged to ask for the appropriate type of survey. Whether the students choose phone, intercept, or Web survey should influence the types of scales used.

Also, given the large size of a national probability sample of donors, simplicity and practicality should be considerations in selecting appropriate scales. Students will likely offer that numerical and Likert scales meet such criteria. And both also allow for the production of ordinal and interval data, suitable for extensive analysis. With respect to the likelihood of donors making donations to a particular cause or a general fund, or having them indicate their preference to where their funds should be spent, the Constant Sum or Forced Ranking scales may also be offered and would be appropriate.

Case: Donatos: Finding the New Pizza

Abstract: The case describes a multi-stage study conducted by Donatos, an independent, premium pizza restaurant chain, aimed at tracking interest and response rates to a newly introduced NO DOUGH pizza concept. The research conducted by Donatos is an attempt to test and subsequently meet the needs of its low-carb diet market. In so doing, it also aims to measure customer satisfaction with the new pizza concept.

1.Map the research design used by Donatos for new product development.

This question provides an appropriate introduction for students to the various descriptors of research design discussed in Chapter 8, Exhibit 8-2 Descriptors of Research Design. Here the framework and procedures for research activity are outlined. Students will be able to discuss the exploratory steps taken by Donatos, such as their monthly WASSUP Meetings, e-mail comments from customers and monitoring of eating trends from different sources. The case also allows students to distinguish between different data collection methods, such as the self-administered intercept survey and call-back phone survey employed by Donatos. Secondary data was also used during the exploratory stages from sources that documented and monitored eating trends. Therefore, students can discuss the pros and cons of using this type of data. The experimental design also utilized will allow for discussions relating to the taste tests conducted among employees and how Donatos decision to exclude its "special ingredients" from the Web site may have affected the yielded responses.

The case indicates that Donatos has monitored different sources, such as the syndicated data, NPD Eating Trends, which provides indications of changes in eating habits. They have also garnered feedback from comments customers send via e-mail from their Web site, in addition to the WASSUP meetings held on a monthly basis, where employees were asked to provide feedback on existing social and cultural trends. These steps then establish the exploratory stages of their research. Through this stage, Donatos was able to discover that there was an existing interest in low-carb eating plans.

Stage 2 of the study, the research-based product development phase, involved development of the product prototype. Here data was collected from employee taste testing; in restaurant tests, where participants would complete self-administered intercept surveys; call-back phone surveys for customers who were serviced through delivery and concept screen activities where participants were shown photographs of food products and then questioned. You can use this time to discuss:

Why a phone survey and self-administered intercept survey were used.

The pros and cons of using the self-administered intercept and phone survey using Exhibit 11-5, Comparison of Communication Approaches as a guide.

How reliable would the feedback be from employee taste testing activities?

Stage 3 involved choosing a marketable name for the new pizza concept, here Donatos tested three different names using a weekend omnibus phone survey. The case indicates that their typical new-product development research would routinely take 12-14 months to complete. Given the time constraints that Donatos faced (the possible short-term nature of the low-carb trend), you can ask students to discuss the implications of the findings yielded from the phone survey completed over a weekend period.

Stage 4, the final stages of the design, addressed the tracking of response rates where Donatos employed ongoing telephone tracking studies, in restaurant comment cards and e-mails from customers through its Web site. You can ask students to discuss the usefulness of this stage of the research. Some may indicate that for the purposes of future research, Donatos can benefit from the findings in terms of strategies that were or were not employed or other data collection techniques (personal interviews or mail surveys) and sampling methods.

2. Evaluate the wassup meetings as an exploratory methodology to help define the research question.

The WASSUP meetings undertaken by Donatos served as an exploratory step to further identifying the dilemma faced, that being how to address increasing interest in low-carb diets. These meetings not only supplemented reviews done of other sources, namely the feedback via e-mails and monitoring of eating trends, but it provided additional insight into the dynamics of other social and cultural trends. The WASSUP meeting is essential to Donatos, as it is able to gather useful information about the market from employees, and create a synthesis of this information to best determine how its decision will affect not only the general public, but also persons of different cultures. This feedback from the meeting will also be helpful when tailoring research questions, as through exploring knowledge from different cultures (likes and dislikes). Donatos can then establish a central focus and design specific research questions. You can have students suggest possible research questions that would be applicable to the study, which Donatos could take to solve the existing problem or address concerns about the low-carb diet market. Chapter 4 (Exhibit 4-2, Formulating the Research Question for MindWriter) can be used as a guide. Donatos may suggest the following examples as research questions:

Should we introduce a new pizza concept to satisfy the low-carb diet market?

Should the crust-free pizza concept be modified?

Should the soy crisp recipe be used or should we introduce another protein rich alternative?

3.Evaluate the test market Donatos used. What were its pros and cons?

Evaluation of the test market relates well to Chapters 16 and 17, when discussing the intricacies of sampling and Chapter 12 on Experiments and Test Markets. This question also becomes useful when preparing students for data collection in research (Part 4) and discussing ethical implications involved when dealing with participants (Chapter 7). The test market (customer reactions to the product in two stores in one market) used by Donatos raises important reliability questions.

First, one city is rarely used in a test market as researchers have found distinct eating preferences by geography.

Second, while the test was conducted over several days in the two stores, unlike the introduction of many new food products, no promotion was done to attract customers to the restaurant for the purpose of purchasing the NO DOUGH pizza. This might be indicative of a sample of customers who did not match the profile of the potential customer: those individuals following a low-carb diet. As the test market continued over time, a screening question about whether a customer came to Donatos specifically to purchase the NO DOUGH pizza could have been used to distinguish these desired sample units from those who were likely not the primary target market segment for the low-carb pizza.

Third, the ideal situation for a true experiment would be test products prepared under controlled conditions. Most test markets, however, use actual market conditions which limits the availability of control. The case indicates that 16-17 year old employees were preparing the NO DOUGH pizzas that were used during the test market; this would be typical in most Donatos restaurants. A discussion of the type of experiment being conducted is appropriate here.

There are some advantages as well as disadvantages that this test market brings to the study. As a class activity, you can ask students to suggest possible pros and cons of the test market and instrument used; these may include the following:

In restaurant tests-

Pros:

Reduced research cost by virtue of having its own current customers as testers.

The ability to capitalize on a greater speed of data collection.

Requires less supervision and manpower as test was conducted on site.

Eliminates the need for higher-cost product-testing facilities.

Cons:

Walk-in customers may not be representative of the target market for low-carb pizza as pizza was not at the time considered a low-carb option.Call-back phone survey-

Pros:

Allows for feedback from a wide range of customers.

Delivery customers may provide useful responses, given that the call must be initiated by them.

Ability to reach customers who otherwise would be inaccessible.

Customers can remain anonymous if they choose to express negative views about the taste of the pizza.

Cons:

Donatos cannot guarantee that customers will be willing to take the time to participate in a call-back phone survey, thus non-response error could be very large. The study did not indicate some form of incentive for customers upon making the call. Failure to do this may not encourage customers to want to participate. Customer-initiated calls creates a self-selection sample which might not be indicative of the desired sample unit.

Given that interviewers are needed to facilitate the survey, Donatos may incur additional costs with this type of instrument.4.What measurement scales would you have used on the survey used as part of the in-restaurant product tests?

You can begin by discussing the different types of rating and ranking scales that may be appropriate for the in-restaurant tests, using Exhibit 14-2, Sample Rating Scales, as an initial guide. You can have students debate the usefulness of rating scales, such as the Likert or Numerical scales, where the likes and dislikes of a product can be easily rated. Some may argue that utilizing ranking or paired-comparison scales would be most effective, as Donatos could provide other types of pizzas or different low-carb ingredients and ask customers to make comparisons and choose their preferences. To begin the discussion you can have students brainstorm the different types of pizzas Donatos would sell and have them compare the types of ingredients/toppings that would most appeal to low-carb eaters. You can also have them rate and/or rank their suggestions. This can serve as a preface to discussions relating to ranking and rating scales.

For the in-restaurant tests, a numerical scale or Likert scale would be the most appropriate measurement scales. Both would allow participants the opportunity to indicate their liking or preference attitude toward the product. The Likert scale would allow Donatos to compare a customers preference to others, while the numerical scale provides the flexibility of using both ordinal and interval data for the purposes of analysis. Intention to purchase is another dimension that can be easily measured by the Likert or numerical scale. Another important advantage to the use of these scales, is the simplicity involved in administering them.

Case: HeroBuilders.com

Abstract: This case describes a study conducted by the president of HeroBuilders.com, whose entrepreneurial zeal led to the creation of an e-commerce toy company. Owing to the increased demands for the creation of hero dolls, research was conducted to ascertain the viability of marketing action figure dolls to the then competitive commercial environment.

This case relates well to Chapter 1, where students will be introduced to the role of research in marketing, and how the application of different types of studies aids in furthering effective research strategies. It also exposes students to the mechanisms involved in exploratory studies and the benefits associated with its use (Chapters 5 & 9).

1.Which of the four types of studies are presented in this case.

You can begin by discussing the usefulness of Vicales exploratory findings and ask students to suggest other types of exploration they would pursue before launching such a business; and, how he benefited from completing this stage of his research. You can ask students to create a list of the information needs they would prepare before starting an e-commerce business to sell dolls made in the image of modern day heroes. The most appropriate study, as presented in this case, is an exploratory one. Here the researcher attempts to gain additional insight as to the viability of marketing an action figure. This type of study gives the researcher the opportunity to explore the commercial market of action figure dolls to determine the pros and cons of such a venture. It also presents other prospects where Vicale chose to conduct qualitative research, where he consulted with a lawyer to determine the legal implications involved with marketing dolls in the image of living heroes or political figures. Uncertain of the potential of starting an action figure business, Vicale also explored retail stores that marketed action figure dolls, to further determine the possibility of products gaining distribution. Here he benefited from knowing that an extensive action figure market did exist, and identifying the companies involved in the production of action figure dolls.

If you use this case during discussions of Chapters 5 and 9, you can also use this question as a preface to discussions of other data collection methods.

2.Evaluate the research that HeroBuilders.com conducted prior to launching its hero and villain action figures using the criteria in Exhibit 1-6.

The research conducted by HeroBuilders.com prior to launching its hero and villain action figures presents some limitations based on the criteria listed in the Exhibit.

Purpose clearly defined:

The purpose of the research was clearly defined. Here Vicale aimed to explore the potential of the action figure business. However, he failed to indicate the scope of the research with regard to the creation of villain action figures. The case indicates that his initial survey of the market primarily involved only hero action figures.

Research process detailed:

The case indicates that Vicale explored the e-commerce business to sell dolls by reviewing the Internet sites of other companies involved in the marketing of action figure dolls; he also looked at the retail environment, where he further discovered the extent of the action figure market. In addition, he sought legal advice to determine the implications of creating prototypes of political figures. There was no indication of an interview guide or a questionnaire used to guide data collection, nor was there evidence of an observation checklist used in his discovery trips to retail toy stores. This is a good time to discuss the inconsistencies that might be introduced in research studieseven exploratory onesif the researcher doesnt have a mechanism for collecting comparable data. You might ask students to develop a list of questions that Vicale might have used when, for example, he visited the action figure sections of toy or general merchandise stores.

Research design thoroughly planned:

The research designs plan also lacked important detail. It was not clear if considerable thought was given to how the research would be executed. Vicale conducted the research himself, which raises questions as to the objectivity of his results and whether his personal bias did not play a role in its reliability. The assessment of the retail environment (toy stores) that he explored did not make clear how he chose the types of stores visited, what process was used, or if location was a factor in his consideration.

High ethical standards applied:

Given that Vicale conducted his research independently provided him much flexibility in executing this study, it would therefore be difficult to determine if high standards of ethics were applied.

The case did not indicate whether Vicale encountered limitations during his exploratory research. Students should be asked to detail what limitations Vicale might have encountered. They should be able to list such limitations as a limited sample of stores that this busy executive might have visited, or the failure to check out different types of stores that also might be used for distribution, other than toy stores, or his obvious potential bias toward the idea of producing such dolls.

HeroBuilders.com proceeded with the e-commerce business of making action figure dolls, which indicates that the mode of analysis (not revealed) applied to the data received may have been favorable. However, given that the research process was highly subjective as it was carried out by the owner of the company, coupled with the limitations evident in the research process and design, one may question the conclusions drawn from this study.

Some students know instinctively that not all research employs every criteria listed, and many studies possess flaws, although not crippling to the result. Moreover, some researchers may be limited by cost and time factors that consequently affect the usefulness of their research.

3. What issues other than those Vicale chose to evaluate, would you have included in your research plan for HeroBuilders.com?The decision made by Vicale, to capitalize on the action figure business with the incorporation of his modern day hero concept, was apparently given much thought. However, an across-the-board study of the costs involved in such a venture could have been beneficial, considering the competitive nature of the existing market. In addition to conducting research on the marketability of dolls in the image of heroes, it may also prove helpful to explore how the general market would respond to dolls patterned after villains such as Osama Bin Laden and Saddam Hussein. Despite the numerous requests for villain dolls, this introduction could be damaging to the sale of dolls depicted in a hero like image. Considering the delicate nature of 911 issues, patriotic Americans may find it distasteful that villain images such as those introduced by HeroBuilders.com are marketed. Despite the apparent profit motive involved, this market should be further explored prior to the introduction of such dolls, to evaluate long-term implications.

As outlined, a majority of the responses that Vicale received regarding the creation of the product came from friends and acquaintances. As such, it may have been worthwhile to perform a broader survey to establish other marketing factors (e.g., how much persons would be willing to pay for a such a doll, their primary motivation for purchasing such a figure and for whom they would be purchasing the doll, or how they would like to learn about the doll), but also how the general market of doll buyers felt about villain dolls. Vicale could have also taken his research of the retail market a step further by exploring the cost/profit factors with individual stores and Web sites that marketed action figures and toys.

Case: Inquiring Minds Want to KnowNow!

Abstract: This case describes a multi-stage, communication study undertaken by the research department of Penton Media, a publisher of business trade magazines, to determine the long-term viability of a reader and advertiser service, the reader service card, a post-card sized device used by readers to request additional information from a particular advertiser.

The discussion questions guide the student through the research process. As such, the case may be used throughout the courseespecially through Chapter 15simply by assigning different discussion questions as exercises during different points in the course.

1. We suggest using Exhibit 5-4 for this discussion; it is helpful if you can project the graphic on the screen in front of the class.

A. The management dilemma is a declining number of reader service cards returned, which is causing a smaller number of inquiries and thus a smaller number of sales leads for Penton advertisers. Students won't have too much difficulty in identifying the management dilemma, but you might use this opportunity to discuss that research can be proactiveactually preceding the development of a problem. If Penton saw this as a value enhancing service for advertisers, shouldn't they have been tracking the change in rate of response card return at the very least? Penton also should have been tracking advertisers' continued interest in the reader service card program--after all, the reader service card program has a cost associated with each issue (card printing, return postage, and card handling and forwarding) that needs to be offset by perceived value on the part of the advertiser. Students are usually quick to point out that advertisers don't always track the success of their advertising (not as true an observation for business-to-business advertising as it is for consumer advertising) because not all advertising asks for a specific action. But in the case of advertising designed to generate leads, Penton clearly had the ability to track reader service card activity (a research activity in its own right).

B. Management question: What should be done to ensure that advertisers and readers alike are getting the information they need and want?

C. Research Question: Should the reader service card program be maintained, discontinued or modified? Rather than three alternatives, Penton may have only looked at maintaining or discontinuing the reader service card. Nothing in the questionnaire indicates that they were exploring other unspecified options.

D. Investigative questions: What means do advertisers offer to obtain information about advertisers' products and services? What influences a trade magazine reader in their choice of response method? What types of information are most frequently sought? What is the time frame in which information is needed? Do purchasing agents have different needs than supervisors? How many advertisers offer Web access to information? How many readers use company Web sites for information? Are readers changing their methods of response in the last 5 years? 5-10 years? Do shifts in response methods parallel the decline in reader service card returns? Does a reader's gender, age, or job experience affect their choice of information retrieval method? To what degree do advertisers value the reader service card program?

E. Measurement questions: see instrument in text.

2. Using Exhibit 7-1 is appropriate for this exercise. Since Penton is conducting this research internally, most of the sponsor's rights are covered. Subject's rights are much bigger issues here.

Avoiding Subject Deception: Subjects are all subscribers to Penton's trade magazines, and the names are drawn from the magazine's subscriber lists (the sample frame). The cover letter states that the study is being conducted to help "companies better understand and respond to your request for information." While it doesn't clearly state the reader service card is being considered for elimination, modification or replacement, any reader who has been reading the magazine for any length of time will know that the reader service card is one response option likely to be evaluated. This small deception is unlikely to offend or distort data from a reader responding to the survey.

Subject's right of informed consent (implied): The cover letter invites readers to participate, so they have the right to choose.

Students may want to discuss the right of privacy, believing that the magazine has used information extracted from reader service cards and subscriber records in ways that the reader never intended.

Subject's right to confidentiality. Penton offered an inducement--a drawing for a hand-held color TV(to participate, but that is not likely to alter the information that respondents are willing to share. But to obtain that entry right, the respondent must provide their contact information ("To ensure a correct entry in the random drawing for the held-held color TV, please make any necessary changes to your mailing label."), which affects their right of confidentiality.

3. The sampling plan called for using the subscriber database (1.7 million) as the sample frame. The case clearly specifies that they used a stratified disproportionate random sample, sending out 4000 surveys to obtain the 710 completed surveys of which 676 were considered useable (came from purchasing decision-makers). You can also discuss the sampling plan for the stage 2 study of ad content. And students might want to discuss why they didn't draw a sample of advertisers to participate in a parallel study.

In terms of sample size, Penton mailed 4000 magazine subscribers and received 676 usable out of 710 completed surveys. Penton chose a stratified sample in order to check response patterns in different subsets (42 in all) of the business-to-business market. You can use this opportunity to discuss several sampling issues:

The effect of such stratification on sample size.

The effects of self-selection within a mail survey on the quality of the data (and non-response error).

The importance of data preparation and why certain returned surveys may not be used. In Penton's case, the discarded instruments were not completed by purchase decision-makers, the only qualified respondent from the viewpoint of the advertisers.

The effect of error as it is introduced by the quality of the sample frame.

4. Penton had started their project with exploratory research that revealed that advertisers perceived they were getting fewer viable sales leads with the advertising in 1998 than they were in earlier years. Penton could have been conducting monitoring research with their advertisers or might have been receiving increasing complaints or comments in normal dealings with them.

Stage 2 involved an observation study of 1330 past advertising placed in Penton magazines in two years, 1992 (648) and 1997 (690). This did show a change in advertiser behavior. What was missing at this stage was an indication of change in respondent inquiry behavior. This would lead us to believe that as Penton did not data mine for this information, that it was not tracking the ads that generated response card inquiries. Having the response cards in their possession, it would have been quite easy to do a tally on a periodic basis for each issue's reader service cards.

Stage 3 involved a mail survey of subscribers, which involved an inducement to participate(a hand-held, color television giveaway. The mail survey was pretested in two ways: by phone, and then by mail. This is a perfect time to discuss:

Various types of pretesting using Appendix 15b on pretesting.

Why the mail survey was tested twice.

Why a mail survey was preliminarily tested by a means other than that chosen for the survey itself.

When such inducements might cause error, when they might inject ethical issues, and what typical inducements are used: money, merchandise, and coupons for discounts are common.

Stage 4 involved 40 personal interviews "to gain a deeper understanding of their behavior and attitudes." This is the perfect time to discuss the limitations of a survey for collecting the nuances of attitudes, and the strengths of the personal interview communication methodology.

5. The easiest way to conduct this exercise is for the students, armed with the issues list from their text, to critique the data collection instrument as an out-of-class assignment. You could have them critique the whole instrument, or you might divide the class up into teams, with several teams dealing with a question-level critique and others dealing with the instrument as a whole. The following checklist might be used for such a critique.

Question Level Critique

Should this question be asked?

Issue 1:Purposeful vs. Interesting

Is the question of proper scope and coverage?

Issue 2:Incomplete or unfocused

Issue 3:Multiple Questions

Issue 4:Precision

Can the respondent answer adequately?

Issue 5:Time for thought

Issue 6:Participation at expense of accuracy

Issue 7:Presumed knowledge

Issue 8:Recall and memory decay

Issue 9:Balance (general vs. specific)

Issue 10: Objectivity

Will the respondents answer willingly?

Issue 11:Sensitive information

Issue 12: Shared vocabulary

Issue 13:Unsupported Assumptions

Issue 14:Frame of Reference

Issue 15:Biased Wording

Issue 16:Personalization

Issue 17:Adequate Answers

Issue 18:Objective of the Study

Issue 19: Thoroughness of Prior Thought

Issue 20:Communication Skill

Issue 21:Respondent Motivation

Instrument Level Critique

Introduction and screening

Instructions

Order/question sequencing

Transitions between sections

Conclusion and disposition of instrument

On an instrument level, students should compare their investigative questions developed in Discussion Question 1 to identify whether the boundaries of information request methodologies have been adequately covered. It appears the researcher was very thorough in this regard. Although the layout has been modified to fit the page format of the text, students should look at the layout of the instrument in terms of structure. The layout of the scaled response strategy is clear, and questions clearly delineate (bold text) the time frame of the intention or actual behavior.

Students should be asked to address the instrument's scope: Does it include all necessary questions? Are questions included that seem extraneous? Students might observe that the instrument scope goes farther than might be technically necessary by asking the desired information response that advertiser's could provide. This question (3) fulfills the "what's in it for me?" query that most respondents ask prior to participating in a survey. As a result, it might be better placed earlier in the instrument. Others might ask why the respondent is not asked directly how they would feel about eliminating the reader service card.This is the opportune time to ask whether the communication method (mail) would have required a different order or instructions if the survey had been done by phone. While early placement of the screening question (1) would serve a purpose in a telephone survey, in this instrument it serves as classification data and could better be placed in the end. Also, an interviewer instruction sheet, for repeating response scales, offering skip directions, probing for reasons for chosen method of response would be necessary if the survey was done by phone.

Concerning instructions and transitions, if a respondent were to check the 4th column in question 4 for every option, it doesn't tell them to discontinue. Skip directions also could have been offered in question 6a, to allow the respondent to skip question 6b. Students should be asked about the abrupt change between the core target questions and the classification questions. The purpose of a better transition between the target questions and the classification questions is to indicate why or how the personal information will be used, in order to gain full participation in these questions used for measuring association. This instrument loses that opportunity by its abrupt transition.

Finally, the students should discuss the end of the survey. Both a conclusion and disposition instructions should follow the last question, telling the respondent, again, to return the completed survey in the postage-paid envelope and thanking them for their participation. It wouldn't hurt to repeat the request to correct the address label to insure an adequate entry for the television drawing. The information that is captured with the respondent's identity will provide other crucial association variables.

At the question level, ask your students if they feel the chosen response strategies (multiple choice-single response, checklist, free-response) are appropriate, and why or why not. Additionally, ask your students if any operational definitions are missing. Some students, for example, may not know what "fax-on-demand" is. You might ask them to come up with operational definitions for those terms they feel the respondent might not know or which might cause confusion. Students should be able to determine if the correct concept and construct has been measured in each question (increase/decrease in activity for question 4, use/expected use of Internet in question 7, etc.).

6. Have your students code the survey for analysis, identifying the number of variables, the numerical codes for each likely response, and the variable labels. This is also a great opportunity for discussing the coding of free response questions as questions 2, 4b, 5a, 5b, 7b, 10, and 11 use the free response strategy.

Q# VDataQ#VDataQ# VDataQ#VData

17Nominal4b15Ordinal6b1Nominal91Nominal

245Nominal5a1Nominal7a8Nominal101Ratio

316Ordinal5bNominal7b8Nominal111Ratio

4a15Ordinal6a1Nominal81Nominal

7. This question asks the student to build a preliminary analysis plan then fulfill it. It also allows you to discuss whether it is the researcher's role to provide a recommendation or merely to report the findings. You can also use this question to compare textual presentation (as is used by the case) with tabular and graphical formats. Finally, ask this question to determine what should be done with those original 42 subsets of interest, and what might have been used as the factor of incidence for this study. You might want to ask students to prepare both tabular and graphic depictions of the data and compare the results. Without a direct question about eliminating the reader service card any recommendation is problematic given the data presented. However, one statistic will stand out as revealing to the student: between 69-71% responded by mail during the last year when they didn't have an immediate need. This might explain why advertisers don't perceive reader service card respondents as good sales leads(these inquires may be primarily from potential purchasers whose need is not clearly defined or whose purchase is too far into the future to connect with that early reader service card inquiry. Is this enough to make a "continue the reader service card' recommendation? See below.8. Every survey has limitations based on scope and methodology. This is a perfect time to discuss sources of error in this survey. The limitation of most concern is the non-response error. That same 71% who had used a reader service card in the past year looks very different when you look at the total original mailing. The 494 respondents translates to only 12.3% of the original sample. 710 of the original 4000 responded, 17.75%. One could hypothesize that those who did not respond feel that they have sufficient ways to reach advertisers about their products or services, therefore they are not interested in preserving one method vs. another. Could those who responded have some ulterior motive for wanting to keep the reader service card? You should also raise the concern about weighting the responses based on the disproportionate stratified sample that was drawn. Does this create or solve a limitation?

9. Ask your students if a decision could be made based on the information provided. This is an opportunity to discuss the risk associated with decision making in the absence of perfect information and the value of a decision. Penton obviously thought there was some risk to this decision or they wouldn't have undertaken such a comprehensive study. Some students might suggest an experiment at this point. Create a split run of an upcoming issue, one group gets ads tied to a reader service card while a second group gets ads not tied to a reader service card. Be sure to bring out additional cost (two different ads created for each advertiser, shorter production runs, more time before making a decision on the issue at hand) of this experiment during the discussion. Other students might suggest tracking the returned reader service card use during the next several months (another descriptive study). Others could argue that the appropriate sample for the original study was incorrect, that it should have been advertisers; if advertisers thought the reader service card was no longer of value, then, and only then, should the device be eliminated.

Case: Mastering Teacher Leadership

Abstract: A multi-stage, communication study of teachers by Wittenberg University's Department of Education to determine the viability of starting a Master of Education program for Ohio-certified teachers working within school districts serving a five-county area.

The discussion questions guide the student through the research process. As such, the case may be used throughout the courseespecially through Case 15simply by assigning different discussion questions as exercises during different points in the course. With the comprehensive data set available on the DVD, this case can also be used for the data analysis chapters.

Not all research projects are well designed, and this one has some very obvious flaws that will be revealed during the discussion of the seven discussion questions. This is a good case to use to discuss whether questionable data helps reduce the risk of poor decision making. You can also tie this case back to the concerns in Chapter 1 about research being done by those untrained in research. Those in charge of this project were all highly educated, and they had been schooled in their doctoral programs to do research for their dissertations. But they were clearly not trained in marketing research methods as the research design, sampling and survey instrument demonstrate.

1.We suggest using Exhibit 5-4 for this