How to Carry Out Qualitative Research · Have an appreciation of other quali-tative research methods such as ethnography and semiotics. 10. Know when to use qualitative research and

The main aim of qualitative research is to explore the full extent of opinions

held by consumers on a given subject. Qualitative research relies on semi-

structured or even unstructured probing of respondents’ opinions in contrast to

the fully structured survey. It is inappropriate to base quantitative statements

like “Eighty percent of the respondents liked this product” on qualitative

research. Our aim is to obtain insights into consumer behaviour rather than to

draw precise conclusions.

Qualitative research techniques are among the most widely used explora -

tory procedures. The opinions gathered are typically used to generate hypothe-

ses that may be further tested and confirmed by quantitative research

procedures. While the end product of quantitative research is quantified public

opinion, the end product of qualitative research is the researcher’s reasoned

interpretations of public opinion. Qualitative techniques involve direct and

indirect questioning without a standard questionnaire as well as projective tech-

niques. Exhibit 5-1 shows some of the qualitative research techniques that will

be described in this chapter.

While qualitative research techniques are often used as a prelude to large-

scale quantitative studies, they should be distinguished from other exploratory

techniques such as secondary research. Qualitative techniques can sometimes

provide answers to questions that quantitative surveys (to be explored in detail

in later chapters) cannot provide. When the issues involved are sensitive, when

After completing this chapter, youshould be able to:

1. Understand the meaning of qualita-tive research.

2. Know how to set up a focus group.

3. Know how to organize a focus group.

4. Understand the skills needed to be agood focus group moderator.

5. Appreciate how technology-basedfocus groups work.

6. Know the essentials of analyzing andreporting focus group discussions.

7. Understand the meaning of individ-ual depth interviews.

8. Know what projective techniques areand how they fit into the qualitativeresearch framework.

9. Have an appreciation of other quali-tative research methods such asethnography and semiotics.

10. Know when to use qualitativeresearch and when not to.

How to Carry OutQualitative Research”Not everything that can be measured matters; not everything that matters

can be measured.”

C H A P T E R 5

Qualitative research elicitsinformation using semi- structured or even unstructuredprobing of the opinions ofrespondents. It also includesindirect methods such as pro-jective techniques.

EXHIBIT 5-1 Qualitative Research Techniques

A focus group involves anextended discussion between amoderator and a group ofrespondents, typically eight toten in number. The discussiongenerally focuses on a singletheme (e.g., health foods orinsurance).

146 | CHAPTER 5

the motives for buying a product are complex, when things are

judged more on an interplay of emotions than on a rational basis,

and when consumers are not aware of the reasons for buying or

not buying a product, asking consumers direct questions can be

a poor way to understand them. For instance, in a questionnaire,

we can ask, “What do you like about this coffee?” and we might

get a response such as “I like the taste.” But what does that

mean? Strong taste? Weak taste? Taste that lingers? Taste that is

different from other brands of coffee? Quantitative surveys are

not designed to explore (except in a limited way) in detail the

answers given by respondents. For these reasons, qualitative

research should be considered as more than a set of exploratory

techniques—it is a distinct way of understanding consumers.

Qualitative research provides insights while quantitative research

provides measurement. Since measurement is the ultimate goal in

marketing as in business, qualitative research is often followed

by quantitative research.

Not all researchers do qualitative research. Some researchers

specialize only in quantitative research and others in qualitative

research. A substantial amount of qualitative research is carried

out by independent moderators who assume the responsibility for

the project but use the facilities offered by others to recruit suit-

able respondents and rent focus group rooms. Most full-service

research houses have qualitative research specialists among their

consultants. Large firms that specialize essentially in qualitative

research, such as In-Sync, are uncommon.

Qualitative research provides insights while quantitative

research provides measurement.

Ethnographic/observational

FocusGroups

IndividualDepth Interviews

ProjectiveTechniques

Qualitative Research

Traditional

Online

Bulletin Board

Use of Technology

Personification

Transposed Vocabulary

Transference

Sentence Completion

Half Questions

Cartoon Tests

Photo Sort

Other Methods

Semiotics

Laddering

ZMET

Other MethodsInternet Viewing

Video Conferencing

Digital Recording

OtherTechniques

Focus GroupsFocus groups are by far the most frequently used form of qualitative research. A focusgroup involves an extended discussion between a moderator and a group of respon-dents (typically eight to ten). The discussion generally focuses on a single theme such aswhat consumers like or dislike about a brand of perfume. During the course of a focusgroup, respondents will discuss a particular issue or range of issues facilitated by themoderator whose role is to “focus” the group. Moderators are required to perform a bal-ancing act: ensure that the client’s issues are addressed and at the same time allow theconversation to flow so that unanticipated but valuable discussion can occur.

The influence of focus groups is much more pervasive than is normally recognized.For example, political parties regularly conduct focus groups to find out how peoplewould react to a particular issue or how to word an unpalatable message to make itacceptable to voters. Hollywood producers use focus groups to help assess which of twopossible endings for a movie would appeal more to moviegoers. Given the visible andinvisible roles focus groups play in our lives, it is important to understand how theywork. Exhibit 5-2 shows the different steps involved in carrying out a focus group.

HOW ARE FOCUS GROUPS SET UP?Most focus groups involve attracting a number of respondents to a central location. Theindividuals will meet a particular set of demographic, psychographic, and/or behav-ioural criteria. The moderator will discuss the issue of interest with the participantsusing a discussion guide, and will record and interpret the discussion.

THE SETTING In a typical focus group, eight to ten people whose opinions are relevantto the subject are assembled in a central location. A focus group, for instance, could aimto understand the basic motivations of consumers that relate to buying an automobile.A trained moderator facilitates their conversation for two hours or so. Although a skilledmoderator will make it appear to be a friendly conversation among a group of peopleabout a product or service, it is in fact a clever elicitation of consumer opinions on spe-cific areas that are of interest to the researcher. (A mini-group is a junior version of thefull-sized focus group. Mini-groups usually include two to six participants and lastabout an hour or so.)

The ambience of a focus group is informal because it is essential that participantsbe at ease in order to establish rapport with them. Unless there is a specific reason to dootherwise, participants in a focus group discussion are generally compatible with oneanother in terms of demographics and in sharing a common topic-related interest. Mostfocus groups are set up either in a boardroom style (participants sitting around a table)or in a living room style (participants sitting in comfortable furniture around coffeetables).

DISCUSSION GUIDE In a focus group, no set questionnaire is used to elicit opinions.Rather, the moderator works from an outline of topic areas that need to be covered,which is usually referred to as a discussion guide (see Exhibit 5-2 for a brief example ofa discussion guide). A discussion guide rep-resents an advance best guess as to how thediscussion should flow. However, it is oftenthe case that a better way of exploring thesubject emerges after groups are actuallyunderway. When this happens, ad hoc adap-tation is required. Hence, a discussion guideis not a hard and fast rule in the sense that aquestionnaire is, but more a jumping-offpoint into possibly uncharted waters. Theongoing adaptability of qualitative researchis what makes it conducive to exploring allissues, whether anticipated or not. As Langer(2001) put it, “Each focus group has its ownrhythm, and the moderator follows it upright away. To cut respondents off and moveon to another question can dissipate the energy that the group may not regain.”

A mini-group is a junior ver-sion of the full-sized focusgroup with about two to sixparticipants.

The discussion guide is anoutline of the topic areas to becovered in the focus group.Moderators generally use writ-ten discussion guides.

How to Carry Out Qualitative Research | 147

148 | CHAPTER 5

EXHIBIT 5-2 Discussion Guide (Example)

Subject Elements of customer satisfaction

Client Zapatax Insurance

Introduction Welcome

(5 min.) Orientation: the role of participants and of the moderator

Disclosures (one-way mirror, audio-taping)

Background Who is covered

(15 min.) What is covered

Employer vs. personal enrolment

Coverage with Zapatax

Switch from and to Zapatax

When first bought Insurance with Zapatax

Reasons for being with Zapatax

Reasons for switching, if relevant: referral, ad, and dissatisfaction with previous

firm

Satisfaction Setup(15 min.) Current plan vs. former plan

Motivations for buying the current plan

Was it what you wanted or just better of the two unsatisfactory choices?

What would have made it satisfactory?

How was the decision made? e.g. Superiority of the plan, spousal support

Satisfaction level with the plan. Reasons for this.

(15 min.) Communication processEase of access to representatives

Ease of accessing relevant information either through the rep or other sources

Friendliness and helpfulness of the reps

Knowledge of what is included or excluded in the plan

(15 min.) Decision variablesWhy buy insurance?

What considerations go into buying different kinds of insurance?

Would you forgo any other purchase to buy insurance?

Is the cost reasonable?

If you can afford it will you buy more?

(15 min.) Utilization satisfaction Have you ever made a claim?

Was the process satisfactory? Fair? Why, why not?

Phone servicing

Billing

(10 min) Overall plan satisfaction

(10 min.) Specific plan satisfaction

Suggestions What are the negative factors?

(15 min.) What needs improving?

Wind down Thank respondents and close session

(5 min.)

Like questionnaires, focus groups tend to follow a funnelling structure: the discus-sion starts with broader issues and becomes more and more narrowly focused as thediscussion continues. For example, in a focus group discussion about insurance, westart very generally: current plan versus former plan, motivations for buying the cur-rent plan, how good the plan is, why it is satisfactory, why it is superior, etc. Such top-ics are impersonal and easy to comment on. This ease, in turn, creates rapport amongparticipants and the moderator. Then we move along to specific topics such as com-munication, decision, and satisfaction. This is not necessarily the only way to structurea discussion guide. In fact, a good moderator would not hesitate to change the order ifthe flow of the discussion dictated it. So, a discussion guide should be seen as a list oftopics that we would like information about. The sequence is our guess of how the dis-cussion will proceed.

RECORDING Focus groups are generally recorded onto audiotapes for transcriptionand analysis. In some cases, they are videotaped as well. Typically, the camera is in a fixedposition and videotaping is controlled by an operator that is not present in the room.The interview room is equipped with a one-way mirror to facilitate client viewing of thediscussion. The participants are made aware of both the tapings and the presence ofobservers behind the one-way mirror, but for evident reasons of bias prevention, thespecific identity or connections of the observers are not divulged. They are normallyonly described as people relevant to the topic and anxious to hear the respondents’ com-ments first hand. Even when respondents do not object to such recordings, it should benoted that recordings may not be used in other contexts (such as advertising) withoutobtaining specific permission from the respondents involved.

HOW TO ORGANIZE FOCUS GROUPS Carrying out a successful focus group requires good answers to the following questions:

● What is the objective of the focus group? ● Who should be included in the group? ● Who should be avoided? ● How should participants be recruited? ● How should groups be scheduled? ● Where should discussions be conducted? ● How should the number of groups necessary to solve a given problem be

decided? ● The validity1 of the study may be affected if proper attention is not paid to

these issues.

Exhibit 5-3 shows the steps involved in conducting a focus group. Although thesesteps look linear—one step after another—they are not often carried out in the samesequence. For instance, it is quite common to select a moderator and then define theobjectives and select participants. The exposition here covers all the aspects of conduct-ing focus groups but in a somewhat different order. We believe this order is more con-ducive to understanding.

OBJECTIVES It is important to decide the objectives of the focus groups beforehand.As we saw, most focus groups use discussion guides. Since topics can be added or deleted


1 Validity refers to whether a technique measures what it purports to measure. While there are quantitative

measures of validity, in this case, we refer to validity in more general terms, things that can be construed to

affect the validity. This is often referred to as “face validity.”

until the group discussion starts, sometimes even in the middle of a focus group, itmight be tempting to neglect the primary objectives of the study. This can be an errorbecause if the objectives are unclear, even the best moderator cannot direct the discus-sion effectively. For instance, a focus group with the objective to assess the corporateimage of a company may have a discussion guide with content that is very similar to theone designed to assess the competitive structure of the market. The objective of thefocus group will dictate the order in which topics are discussed and the amount ofemphasis placed on each topic.

WHOM TO INCLUDE? As with any opinion study method, qualitative respondents aremethodically drawn from the target market group. Criteria for respondent selection arenormally specified by the client, working in collaboration with the moderator, and thenpassed on to professional recruiters who specialize in assembling respondents for qual-itative studies. If the client is a manufacturer of an expensive perfume aimed at youngwomen, possible criteria for focus group participants could be: women between the agesof 25 and 39 who have a household income of at least $50 000, live in a metropolitanarea, and are receptive to the idea of using perfume. While determined recruiters canusually find respondents who fulfil almost any set of criteria, complex recruitment cri-teria should be avoided whenever possible because they prolong recruitment time,increase cost, and may result in erroneous respondent selection. It is therefore impor-tant to test each criterion and restriction for importance and relevance. The larger thesize of the defined universe being recruited from, the more successful the process is likely to be. Recruitment procedures take at least a week (or several weeks if the criteriaare complex) and this should be taken into account when deciding on the lead timerequired for focus groups.

WHOM TO AVOID? Qualitative research is not bound by the same rigorous samplingrules that govern major quantitative surveys. Respondents are not expected to be statis-tically representative of the population as they should be for a survey. However, this doesnot imply that participants can be recruited in a haphazard way. It is especially impor-tant that recruitment procedures be designed to control the introduction of unin tendedrecruitment-related bias. To ensure this, good recruitment follows a number of guide-lines.

Criteria for respondent selec-tion are normally specified bythe client, working in collabo-ration with the moderator.

150 | CHAPTER 5

EXHIBIT 5-3 Steps in Conducting a Focus Group

Recruitparticipants

Select/briefmoderator

Schedule groups

Conductgroups

Define theobjectives

● Respondents in any group should not be known to the moderator and shouldbe strangers to one another. Friends tend to act as a social unit of their own andthis can inhibit the global interaction process. Within reason, people who knoweach other may take part in back-to-back groups on the same day but not ondifferent days due to the likelihood of the first participant briefing the latter.There are exceptions, however. Bring-a-friend groups are found to work for cer-tain “chatty” issues such as cosmetics. Also, when recruiting in small communi-ties, finding participants who do not know each other might be almostimpossible.

● Respondents should not be regular participants in focus group discussions.Respondents who have participated in too many groups, sometimes with thecollaboration of an unethical recruiter, are called “professional respondents.”They become wise to the ways of moderator questioning and cease to respondas lay consumers. Some come to see themselves as amateur moderators and tryto help out by asking probing questions. The general rule is that participation ispermitted just once a year, or five times in total, and never twice on the samesubject. In Toronto, Montreal, and Vancouver, there is a computerized data bankcalled the “Central File” (see Research in Practice) to control this participationrate. Elsewhere, the ethics of recruiters must be relied upon.

There are many accepted guidelines that relate to focus group respondents. Theseinclude:

● As far as is feasible, respondents should be unaware in advance of the discussiondetails. A primed respondent is less likely to be a representative typical con-sumer. Screening questionnaires can be designed to incorporate spurious crite-ria intended to deliberately keep respondents off guard. For example, eachcandidate could be asked if he or she drinks milk, soft drinks, coffee, tea, beer,and wine for a focus group on coffee.

● Respondents should be sufficiently skilled in the language so that they do notmisunderstand the moderator and vice versa. Linguistically disadvantagedrespondents are usually nonparticipatory and can also inhibit the free flow ofdialogue.

● Respondents should not be closely connected either personally, through theirfamilies, with advertising, with marketing research, or with industries related tothe topic of discussion. The latter exclusions are both to assure that the discus-sion represents lay consumers rather than those who are particularly wellinformed on a topic and to guard against revealing confidential information tocompetitors.

● Respondents should not be recent immigrants to the country or to the citybecause until they get used to their new environment, their point of referencecontinues to be the point from which they immigrated.

● Respondents should be literate and told to bring glasses or hearing aids if theyuse them. Many qualitative techniques involve responding to stimuli, questionsin writing, or through projective drawing techniques. Respondents may also beasked to respond to visual stimuli such as concept statements or television com-mercials.

● Respondents should be articulate and comfortable speaking in a group. Manyrecruiting companies include a standard question to ensure that potentialrespondents can express themselves clearly.


Whom to avoid? (general guidance only)

● respondents who are friends with each other● professional respondents● respondents with language difficulty● recent immigrants

152 | CHAPTER 5

RESEARCH IN PRACTICE

Central Files: The Respondent Tracking SystemWhat Is Central Files?Central Files is a system of monitoring and tracking the par-ticipation of respondents in qualitative interviews and focusgroups. It can also include any undesirable respondentsreported by moderators as unco-operative or overbearing.

How Does It Work?Every month recruiting firms submit the names and phonenumbers of all respondents who attended a group or inter-view during the past month. While one recruiting firm caneasily monitor the participation of respondents in their owndatabase, it is impossible to ascertain whether a respon-dent has attended for another firm in the past month or thepast six months. When each recruiting firm submits itsattendees to Central Files they can be cross-referenced andabusers are highlighted. Central Files checks by phone num-ber and first names, even if the name is similar (i.e.,Brown/Browne). If a match is found in any of the three cat-egories, the name goes on the “Do not recruit again” list.This list is sent monthly to all recruiting firms who regularlysubmit to Central Files.

What Are the Benefits of Central Filesto Qualitative Researchers?The Central Files system represents a commitment to excel-lence and self-monitoring by qualitative researchers. It hasbeen ongoing since 1983 and is the only system of its kindin the world. The fact that it is controlled by PMRS andmaintained by an independent data management groupgives legitimacy and security to the industry. Several U.S.moderators have tried to initiate a similar system with lit-tle success so far. Recruiting firms pay $0.50 per name sub-mitted to Central Files. This money pays to maintain thedatabase and updates.

There is still the possibility of involving respondentswho misrepresent themselves to get into qualitative stud-ies, but at the very least Central Files provides a safety netto eliminate the majority of them. It is only successfulthrough the co-operative efforts of the recruiting firms inCanada.

Based on an article from Imprints written by Dawn Smith, February

2001. Reprinted with permission.

These points above are not rigid rules but guidelines to be used when no specificcriteria are set. For example, although respondents who know one another should beavoided in most cases, there are occasions in which it may be permissible. In small cities,it is all but inevitable that some respondents will know one another even if not recruitedby friend-to-friend reference.

When there is a potential for one part of the group to restrict the comments ofanother part, it may be necessary to keep subgroups separate. For example, if a teengroup were composed of respondents ranging from 13 to 18 years old, the younger onescould easily be intimidated by the older ones and defer to them. Users and nonusers mayconflict by one’s deference to the other or, unpredictably, as a result of the force of per-sonalities on each side.

Whenever the budget permits, the study structure should include at least twogroups in each project specification category. This is to guard against any one of thegroups being atypical and leading the researchers astray. This precaution is no differentfrom any other inherently small-sample research.

Mixing respondents of opposing views (e.g., in political or social matters) for thepurpose of creating a confrontation has evident theoretical appeal but may not workvery well in practice. The presence of strong and unbridgeable disagreement among agroup of strangers is likely to induce silence and discomfort rather than a spirited andconstructive debate.

HOW TO RECRUIT PARTICIPANTS Respondents for focus groups may be recruited bytwo methods: the random method or the referral method. In the random method,respondents are chosen using mechanical statistical procedures. In many ways, this is anideal procedure because it eliminates human judgmental errors that may result in bias.(Chapter 4 provides a more detailed explanation of sampling procedures in general.) Inpractice, however, this method is seldom used because it requires over-recruitment andit can be frustrating to the recruiters. Both of these drawbacks stem from the inherentskepticism in our culture. Potential respondents tend to doubt the legitimacy of theinvitation to participate and may possibly be concerned about their own security as well.The refusal rate can be very high. Moreover, a disturbing proportion of such randomrecruits may not show up or are on guard when they do. Above all, it has yet to bedemonstrated that the random method is appreciably better than the referral method towarrant the additional time and expense.

Therefore, in practice, the most commonly used method is the referral method. Itis a snowball or pyramid technique in which respondents are recruited on the basis ofbeing referred by others. Recruiters build up data banks of past participants and ofpotential references known by past participants. This technique is relatively inexpensiveand participant reliability is usually good. The technique reduces skepticism since arecruiter can tell a potential participant that someone they know has participated in astudy before and has recommended them. Potential participants can quickly and inde-pendently verify that the invitation is genuine. The referring friend will probablyencourage them to take part. Respondents tend to enjoy the experience. Many are flat-tered to know someone values hearing their opinion, learn from the interchange withothers, and like the sense of having been admitted to the inner sanctum of the market-ing world.

The PMRS Code of Conduct has recruiting specifications. According to the Code,the following Rules of Conduct should be followed unless changes to any or all of therules have been discussed and agreed to by all parties involved in the research study.

a. All respondents must meet usage/trial/ownership standards including type ofbrand, frequency of use/trial, or other time limits specified for the study plusdemographic specifications for the study including marital status, age, sex,income, occupation, and household composition, etc.

b. No respondents (nor anyone in their immediate families or households) maywork in an occupation that has anything to do with the topic area (wholesale,retail, sales, service, or consultant) nor in advertising, marketing, marketingresearch, public relations, or the media (radio, television, newspaper, film/videoproduction, etc.).

c. No respondent may be recruited who has attended, in the past two years, afocus group discussion or in-depth interview on the same general topic asdefined by the moderator.

d. No respondents should be recruited who know each other for the same study,unless they are in different groups or interviews that are scheduled back-to-back.

e. No respondent may be recruited who has attended a group discussion or in-depth interview within the past six months.

In the random method,respondents are chosen usingmechanical statistical proce-dures.

In the referral method,respondents are recruited onthe basis of being referred byothers. The referral method ismore commonly used than therandom method.


f. No respondent who has attended five or more focus groups or in-depth inter-views in the past five years may be recruited.

g. At least one-half of the respondents recruited for each group/study must neverhave attended a group discussion or in-depth interview.

h. All respondents must have been living in the specified market area for at leastthe past two years.

i. All respondents must be able to speak, read, and write in the language of thegroup or study being conducted.

HOW TO SCHEDULE GROUPS There are no hard and fast rules with regard to the sched-uling of group interviews. The following suggestions may, however, be useful.

● Customarily, two groups are conducted per evening, at 6:00 p.m. and 8:00 p.m.Daytime nonbusiness focus groups may be suitable for students and for nonem-ployed, part-time employed, or retired people.

● As many as three group discussions can be carried out in a day. More than threea day is too draining on moderators and observers alike.

● Friday evenings or evenings before a statutory holiday are a poor choice forgroup discussions; these are the worst times for no-shows. For the same reason,scheduling should take into consideration sports playoff series, major televisionevents, etc.

● Respondents are paid an incentive for participating in the group discussion.(During 2005, the standard incentives in most Canadian cities were $50 to $75for most respondents and a little less for teens. For rare and difficult-to- motivate respondents, $100 or more was common. Examples of the latter wouldbe doctors, owners of luxury cars, business executives, and so on.)

Once again, the points above describe what is commonly done in practice andshould be used only for general guidance.

WHAT TO CONSIDER ABOUT LOCATIONS AND OBSERVERS All professional focus groupfacilities offer clients the opportunity to observe groups through one-way mirrors.Clients can observe the focus group live through the mirror as well as listen to the dis-cussion, which is transmitted through microphones. Participants are told about theclients’ participation, but once the discussion gets going, they do not seem to be affectedby it.

It is generally desirable for clients to observe the focus group sessions as they hap-pen. In fact, this is one of the main spin-off benefits of qualitative research. The oppor-tunity for marketing people to play “fly on the wall” and to learn first-hand what theircustomers (or those of the competition) are thinking can provide invaluable insight.This is why special focus group facilities are equipped with one-way mirrors. Having theclient present is also invaluable to the moderator, at least for the first few sessions, inorder to ensure that she or he is working to the ends the client desires. Frequently, pro-cedural modifications are made as a result of client observation.

A good focus group location should have a neutral and relaxing decor. There shouldbe a host or hostess to control comings and goings, look after hospitality, attend to therecording equipment, and ensure that all respondents are on the attendance list.Sometimes, respondents will try to bring along friends. Of course, only those specificallyrecruited for the group are allowed into the room.

The facility may offer an open living-room-style seating arrangement or a boardroom -style table. The layout is not too important although some moderators may prefer one overthe other. Boardroom arrangements are by far the more common of the two types.

154 | CHAPTER 5

It is desirable to keep the observers limited to people with a direct reason for beingpresent. Experience has shown that the more observers there are, the more they interactamong themselves and the less attentive they are to the job of observing. .

The observers should view the proceedings with a completely open mind and notbe quick to judge the participants or the moderator. The respondents may not say whatthe observers expect them to say, and the moderator may not ask the questionsobservers think should be asked. Nothing can be learned by being critical of the partic-ipants or of the moderator. Observers should make it a point to be respectful of the par-ticipants (as they may be drawn from a very different social stratum than the observers)and of the moderator (who may have deviated from the discussion script for reasons notobvious to the observers). If the participants say what the observers want to hear, and ifthe moderator asks the questions in the way the observers want him or her to ask them,there can be little additional learning.

In Canada, there are observation facilities in all major cities such as Toronto,Halifax, Quebec City, Montreal, Ottawa, London, Winnipeg, Calgary, Edmonton, andVancouver. In towns where there are no observation facilities, a suite in a major, cen-trally located hotel is used instead. A living room and adjoining bedroom with separateentrances is one format that is often used under these conditions. Also, two adjoiningmeeting rooms can be used where the group is conducted in one boardroom andclients observe the group in the room next door over a closed-circuit camera/TV con-nection. Clients can listen to the proceedings via audio hookup in the adjoining room.Video hookup can be used as well. A passive, fixed camera usually adds little enhance-ment. The camera can be panned and zoomed by an operator to produce a more inter-esting video. While initially this might be distracting for some respondents, theyquickly become accustomed to the operator. The camera itself is no longer much of aninhibitor because the proliferation of camcorders has made video equipment fairlycommonplace now.

If it is impossible to have adjoining rooms, and depending on the arrangement ofthe hotel space, one or two observers can sit off to the side in the group room. It is veryimportant that such observers remain quiet and unobtrusive and do nothing to inter-fere with the moderator’s control of the group. The moderator customarily introducessuch observers as research colleagues who collaborate on the eventual report even if the“colleague” is the president of the client company.

VIDEO CONFERENCING AND INTERNET VIEWING A high-tech version of focus groupsis video conferencing. In video conferencing, the focus group discussion is transmittedlive to another location across the city or across the country. This arrangement permitsthe client to transmit extra questions or seek clarifications in addition to observing thediscussion. However, video conferencing is not widely used due to the cost as well as theavailability of suitable facilities.

A more popular version of video conferencing is internet viewing, a far less expen-sive and far more flexible method. Here, the interview is made available live over theinternet, enabling the client to watch the focus group from anywhere, chat with associ-ates, and communicate with the moderator in real time. Although still not widely useddue to technological limitations, as advanced technology becomes commonplace, inter-net viewing will likely become a standard method of viewing focus groups.

HOW MANY GROUPS TO CONDUCT It is important to determine the appropriate num-ber of groups (or individual interviews) that should be used for a given study.Obviously, the greater the diversity of a market, the larger the number of groups needed. The best test for determining the number of groups required is to identify how


In video conferencing, thefocus group discussion is trans-mitted live to another locationacross the city or country.

In internet viewing, the inter-view is made available live overthe internet, enabling theclient to watch the focus groupfrom anywhere.

many segments are necessary in order to elicit all major motivational perspectivesrelated to the subject. The law of diminishing returns also applies. Increasing the num-ber of groups increases the chances of finding new material that is relevant to the prod-uct category. However, increasing the number of groups beyond a certain point iscounterproductive.

Is there a magical number (of groups) beyond which the insight generated byrespondents is no longer new? There is no generalized answer to this question. Thenumber of groups required depends on answers to a number of questions, such as:

● Are there major regional variations in attitudes toward the product?● Do different socio-economic groups buy the product for different reasons?● Are there distinct age segments in the target group with different needs?● Is the target market highly heterogeneous?● Are the motivations of segments incompatible?● Are heavy users different from light users?● Is it important to deal with users of your product and that of your competitors

separately?● Is the budget being overextended?● Is there a danger that so many respondents over the many groups might lead

some people in the client organization to begin interpreting the findings quan-titatively rather than qualitatively?

The way in which these questions can be answered will indicate whether a small orlarge number of groups is needed. For example, if younger people respond to yourproduct differently than older people, if men respond differently than women, and ifconsumers in the West respondent differently than consumers in the East, you willneed many groups to understand each one better than would be the case if all con-sumers responded to your product more or less similarly. The number of groups canreally mushroom if users and nonusers must be dealt with separately and if there areseveral strong regional differences.

In addition to the above, there are other considerations regarding the smooth func-tioning of groups. There are many unwritten rules and even these differ from modera-tor to moderator. Here are some common considerations gathered by informally talkingwith many moderators:

● Age Range: In order to assure a sense of participating among one’s peers, it maybe a good idea to use separate groups for younger and older people. Amongteenagers, an age span of two years is advisable; among those under 30, five toten years; among older adults, 20 to 25 years. Most studies are confined to peo-ple less than 50 years of age, but this is due to marketing target definition not toany inherent age-related impediment. When their input is pertinent, groups ofolder people can be just as productive as younger segments.

● Demographics: It is best to keep white/blue collar, higher/lower educated,upscale/downscale groups apart. People do not tend to open up in the presence oftheir social opposites when affluence or sophistication is involved in the subject.

● Gender: It is better not to mix males and females in the same group, although itcan be done if budgetary constraints deem it necessary. Mixed groups tend tobe less relaxed and less productive. Moreover, gender-related issues will usuallynot emerge because both genders will avoid the risk of possibly demeaning orstereotyping the other (though such stereotypes can provide highly valuableinsight). These are, of course, general suggestions. Whether to mix genders ornot will depend on the context and content of the focus group.

156 | CHAPTER 5

WHAT THE MODERATOR NEEDS TO KNOWIn qualitative research, there is no set questionnaire and it is difficult to determine inadvance the course that group discussions may actually take. It is therefore imperativethat the moderator be thoroughly briefed beforehand. This allows him or her to makead hoc adaptations intelligently as required and to sense unexpected but potentiallyvaluable additional avenues to follow. The poorly briefed moderator will be ill-equipped to adapt. Naturally, the more familiar the moderator is with the category, theless general briefing she or he will need. Briefing should take into account the follow-ing aspects:

● The marketing environment of the product or service being investigated;● The range of subjects and issues to be investigated;● The format, kind, finish, etc. of display materials if they are to be used during

the interview. Examples of display materials are samples of existing products,prototypes of new ones, advertisements and commercials, concept boards, etc.(The moderator may have some suggestions as to the suitability of these variousforms of display material.)

When the moderator is being briefed, clients should avoid revealing their prefer-ences regarding favoured concepts, political affiliations, or desired results. This may putpressure on the moderator’s essential role of impartiality. Another temptation to resistis the tendency to include every aspect of the subject, especially aspects that are onlymarginally relevant and time-consuming to explore; worse are add-on subjects not rel-evant to the actual topic. The time available is usually limited to two hours due to sched-uling needs and respondent patience. Marginal topics reduce the amount of timeavailable for discussing important ones.

WRITING EXERCISESDuring the course of the discussion, a moderator can make use of writing exercises ortheir substitutes. Writing exercises are used when the moderator would like to know theuninfluenced opinions of each member of the group on a pivotal point. For example,after viewing a commercial, each respondent may be asked to complete a simple ques-tionnaire before the floor is turned over to open discussion. There is no intent to tabu-late such data. The questionnaires are collected and scanned by the moderator as anopinion snapshot of the group and they provide guidance on where to proceed withquestioning.

When less precaution is needed, a close equivalent of a show-of-hands may be used.The round-robin question is a similar technique in which each participant’s bottom-lineresponse is obtained serially (e.g., for a yes or no, for ratings from one to ten, etc.).Although serial questioning tends to make the respondents passive, a few round-robinsdone quickly cause no harm and aid the goal of equal participation.

USE OF DISPLAY MATERIALSMany discussions involve display materials such as package or ad designs, product sam-ples, and concept statements. When the goal is to assess the relative impact of thesematerials, they are presented simultaneously. When relative impact is not important butmaximum depth of response to each design piece is, serial presentation is more appro-priate. However, in serial presentation, sequence bias may be involved. This should beminimized by careful a priori sequencing of the materials and by proper rotation pro-cedures between groups.

Writing exercises are usedwhen the moderator would liketo know the uninfluenced opin-ions of each member of thegroup on a pivotal point.


158 | CHAPTER 5

What type of qualitativetechniques do you use on anongoing basis?

Most frequently, the qualitative

research that we conduct is in

the form of focus groups (both

traditionally sized groups and

smaller formats like triads or

mini-groups). When we are addressing a potentially sensi-

tive subject or require more detailed feedback from respon-

dents, we use in-depth one-on-one interviews. We have

recently begun to use in-home anthropological interviews

and online qualitative methodologies.

For what type of problems do you use qualitative research?

We put qualitative research to its best use when we use it

for true exploration at the beginning of a project. Through

qualitative research, we are able to identify the scope of an

issue, appropriate consumer language, and our information

gaps for exploration in further research, including quanti-

tative methods.

We also use qualitative research for screening con-

sumer communications (advertising, packaging, etc.) at the

concept stage. This methodology allows us to get detailed

feedback from consumers about these communications and,

more importantly, lets us delve into consumers’ associations

and emotional responses.

What type of insights do you gain from qualitative studies?The most important insights we gain from qualitativeresearch are those that allow us to make a connection witha consumer on an emotional, non-rational level. The abilityto delve into consumers’ deeply held associations can pro-vide rich information to improve the products we make andto identify new areas of focus.

Are qualitative studies misused in any way? Are there any other pitfalls?It is my observation that misuse of qualitative research iswidespread among buyers. Because of the quick turnaroundtime and relatively low cost, it is often tempting to usefocus groups to address issues that require a rigorous quan-titative technique. Qualitative research should not be usedto make business decisions, for example, to choose an ad orpackage. By its very nature, results are not projectile. Themost serious limitation of qualitative research is the extentto which it is misused by those who commission it.

While one advantage of qualitative research is that itallows clients to observe the proceedings in progress andwatch consumers react to and speak about the issue underconsideration, it is also a pitfall. Observers can give undueweight to an offhand comment from a participant, “tuneout” for the rest of the group, and take away a distortedview of what really unfolded. And it is always possible tofind one quote that supports a particular finding or recom-mendation, so observers with a point of view can focus onthe one participant whose responses back up their position.

PRACTISING RESEARCHER INTERVIEWS

Qualitative Research: A Client’s PerspectiveAn Interview with Nancy Kramarich, Consumer Insight Manager, Pfizer Consumer Healthcare

WHAT SKILLS SHOULD A MODERATOR POSSESS? Because of the nature of qualitativeresearch, the skills that are required to be a good moderator involve a high degree of sen-sitivity to others, lateral thinking, and the ability to be a catalyst rather than an influ-ence. The aim of this chapter is to familiarize you with some of the tools that areavailable to moderators without being exhaustive. Almost all the skills described here arealso relevant to individual depth interviews. In fact, some of the skills (e.g., creating rap-port using matching and mirroring) are much more effective in individual depth inter-views than in focus groups.

A good moderator should be

● sensitive to others

● able to think laterally● a catalyst

Putting Respondents at Ease The first priority of the moderator is to put the parti -cipants at ease. This is done by first explaining the purpose of the discussion in an infor-mal and nontechnical way and then by establishing a relaxed and permissive climate.The moderator talks to the respondents at their own level to aid in establishing rapportand should be dressed accordingly. Moderating a group of blue-collar workers whilewearing a business suit may send the signal that the moderator wants the respondentsto know he or she is not one of them. This may imperil the creation of rapport.

Skilled moderators follow several explicit and implicit rules of conduct except whenthey deliberately stray from them to achieve a certain goal. In qualitative research, muchmore than in quantitative research, it is critical for the moderator to develop rapportquickly and effectively with the respondents. Most good qualitative researchers do thisintuitively, even subconsciously. Recent developments in the fields of socio-linguistics,hypnotic patterns, psychology, and neuro-linguistic programming (NLP) have identi-fied explicit patterns that can help the researcher establish rapport with the respondents.

Controlling Group Dynamics Whenever two or more people interact, two conflictingmotives propel their interaction: power and solidarity. One may want to be heard(power) but is socialized to make room for others’ views (solidarity). Different peoplehave different balances of power and solidarity. Most people balance the two well,although an imperfect balance is to be expected. When a group has one person with anextreme power orientation and seven others with high solidarity orientation, the power-oriented person will tend to dominate the discussion. The main danger in this case isnot that person’s domination per se but the acquiescence of the remaining participantsfor reasons of solidarity. In other words, they may simply agree with the dominant per-son although they have their own opinions on the subject.

The challenge of moderators is to control the dominant person(s) without appear-ing to do so and, at the same time, enhance their communication links with the others.In fact, whether power-solidarity conflict is a problem or not, it is important for moder-ators to keep all their links open and always make it clear that conflicting views are notonly welcome but also essential to their purpose in learning all facets of the subject.

Creating Rapport Creating a feeling of rapport requires the ability to enter someoneelse’s world and to communicate in such a way as to make the other person feel that sheor he is understood and that his or her views (mainstream or not) are important to themoderator. This leads the person to trust the moderator, which in turn results in anuninhibited flow of information. When rapport is established, the respondent becomesless self-conscious and guarded and thus is in a position to more openly provide valu-able information. However, it is important to note that maintaining rapport does notmean surrendering control of the group to the respondents. In fact, if this happens, themoderator will have failed in one of the most important tasks.

Understanding Surface Messages and Metamessages When people are together ingroups, two types of messages are sent out: surface messages and metamessages.Suppose you ask your colleague, “Have you finished the report?” The surface message isthe information that is sought in the literal sense: whether the report is finished or not.The metamessage might be, “I don’t think you have,” implying incompetence.

Metamessages are conveyed through the intonation, pitch, and facial and bodilyexpressions that accompany the words. They carry or are perceived to carry an implied


message within the surface message. Because metamessages are so important in com-munication, it is quite common to observe that people almost always react to metames-sages. For example, if you ask your colleague if he has finished the report, he might notsay no but might respond, “It’s almost finished,” or “I would have finished it by now butthe computer crashed for a couple of hours this morning.” A person who buys a notablynonprestigious brand may offer a defensive explanation of her or his purchase withouteven being asked.

Metamessages are important in all forms of human communication. They becomeparticularly critical in face-to-face situations between strangers as is mostly the case withfocus groups and depth interviews. A straightforward question such as, “What wine doyou drink?” can send out any of the following metamessages to the respondent:

● I expect any normal person to drink wine.● The wine you drink will tell me how sophisticated you are.● You look like a person who would drink wine.● I’ll bet you drink cheap wines.

The respondent’s answer will vary depending on the metamessage he or shethought was received. The metamessage would depend, among other things, on the toneand context in which the question was asked. The moderator should be aware of themetamessages he or she might be sending by the way a question is asked and the con-text in which it is asked. There may be no more important skill for a moderator to cul-tivate than how to ask a question without simultaneously seeming to have made anunintended statement or to have inadvertently suggested a desired avenue of response.

Maintaining Rapport Socio-linguists, psychologists, hypnotists, and NLP specialistshave identified many ways in which rapport is created between people and have offereda number of guidelines for creating and maintaining it in the shortest possible time.

One major premise in creating rapport is that we have a greater affinity with peo-ple who are similar to us. Similarity can be created through three main methods: phys-ical similarity, speech patterns, and framing. The techniques described here can be usedeffectively in focus groups.

Physical Similarity: Matching and Mirroring If we observe people in a social setting,we find that close friends tend to mirror each other. If one of them leans forward, theother tends to lean forward as well; if one tilts his head, so does the other. This processcan be used deliberately to create rapport. Thus, if you want to create rapport withsomeone, you can mirror the body posture, breathing pattern, tonality of voice, etc. ofthe other person. If the person you are dealing with is a slow talker, talking to him fastis not conducive to establishing rapport.

Rapport is created to make the respondents feel that the moderator is one of them.In terms of physiology, oneness is created by the moderator mirroring the physicalbehaviour of the participants. If the participants are relaxed, a relaxed moderator has abetter chance of creating rapport than a tense moderator. If most participants lean for-ward, a moderator who leans back in his or her chair is not likely to create rapport. Ofcourse, there are situations in which there is no consistent pattern in the group. In suchcases, the moderator may not be able to use his or her physiology to create rapport.However, there are other means of creating rapport.

Neuro-linguistic programming (NLP) holds that we use a dominant mode when wecommunicate. The dominant mode may be visual, auditory, kinesthetic, or even olfac-tory or gustatory. Greater rapport is created when we follow the speaker’s modality. Letus review the following conversation:

160 | CHAPTER 5

Participant: I don’t see how this new concept fits into the whole picture.Moderator: Listen to this carefully. I was saying...Participant: It’s still fuzzy to me.The participant is coming from a visual perspective, picturing a mental image and

trying to make sense of that. The moderator, however, is coming from an auditory per-spective and he merely repeats himself, which at this point is like speaking to the par-ticipant in a foreign language. Moreover, in this example, there is a concealedmetamessage that the respondent is either imperceptive or not very bright. NLPers holdthat such mismatch is destructive to maintaining rapport. The moderator could havematched the participant in the following way:

Participant: I don’t see how this new concept fits into the whole picture.Moderator: Let us look at it this way. If you observe...Then perhaps the participant might have responded:Participant: Oh, I see what you mean.The second conversation, in which the moderator matched the participant’s domi-

nant modality, would have created a powerful rapport between the moderator and theparticipant.

When the conversation is not one-on-one as described above, the moderator mayhave to use the dominant modality of the group while also reiterating his talk in othermodalities.

Even when participants do not use modality-specific words (such as see, hear, per-spective, etc.), it is possible to know how they are processing information. Our eye move-ments correlate with the mode in which we operate. For example, when we try picturingsomething, our eyes go up; when we feel angry, we look down to the right. By observingthe eye movements of the participant, it is possible to know how she or he has processedthe information.

Speech Patterns The second method of creating rapport is through following thespeech patterns of the participants. Speech patterns vary in different dimensions:

● speed● tempo● tonality● words

To a slow speaker, a fast speaker may appear rude and self-centred. A fast speaker inturn may perceive a slow speaker as less intelligent. Whether a given tone is perceived tobe firm and friendly or bossy and offensive depends on the background of the partici-pants. What is bossy and offensive to people from one part of the country may simplybe normal to people from another part of the country. Extensive use of colloquialexpressions may be perceived as coarse by members of some socio-economic groups,while absence of colloquial expressions may be perceived as stuffy and pompous byother groups. A good moderator is aware (consciously or subconsciously) of suchspeech patterns and adjusts his or her speech to suit that of the group.

When the groups are distinct, it is sufficient to identify the suitable speech patterns.However, a good sensory awareness is needed to assess the physiological and languagepatterns when the groups are not very distinct. Some rapport almost always developswithin the group and with the moderator (assuming that she or he is competent). Thereal challenge to the moderator is to elevate the level of rapport as much as possible bybeing perceptive of the subconscious signals generated by participants through thecourse of the discussion.


Framing and Reframing Framing is a way of creating a context. This is done to guideparticipants to change the topic or to give up their current orientation while maintain-ing rapport. Sometimes, framing can also be used to create rapport. Consider a situa-tion in which the discussion has moved into areas that are outside the scope of the studyand the moderator’s attempts to steer the conversation have not been very successful.The moderator wants to know which of the two products under consideration the par-ticipants are likely to buy, but for some reason the discussion animatedly moves on tothe colour of the packages. In such cases, reframing the mindset of the participants canbe accomplished by the moderator by saying something like, “That is very interesting.Now suppose you are in a store and you have only $5. Which brand of Product X wouldyou buy? Assume that the colours of the packages are switched. Now which brand wouldyou buy?” These new frames are specific and therefore require the participants to aban-don their earlier frame. Yet it is nonoffensive as it does not ask them to stop the originaldiscussion. However, the indirect result is that the participants have no logical way ofreturning to their earlier frame.

Reframing can also be used to change someone’s orientation while maintainingrapport. Consider a situation in which a participant believes that he has superior knowl-edge and is bent on exhibiting it by straying into areas that are of no consequence eitherto the moderator or to the other participants. (This is one of the key reasons for screen-ing respondents with inside industry knowledge.) The moderator may reframe the par-ticipant’s orientation by saying something like: “That is a very sophisticated view. Wemay even have another group discussion on that subject. Now, if you were to deal withthe issue at a more common level what would you say about...?” Here, by acknowledg-ing the participant as sophisticated, the moderator maintains rapport by patting therespondent on the back while making it clear that another frame is required. Therespondent cannot ignore the fact that his previous frame of reference and that of themoderator are at odds.

Breaking Rapport While it is absolutely critical for a moderator to create and maintainrapport, she or he may also want to break the rapport from time to time by not follow-ing some of the procedures described above. For example, the group rapport may be sohigh that participants start interacting with exuberance. To regain order, the moderatormay temporarily break rapport with the group by reframing. One must also rememberthat the group is being recorded and observed. When the proceedings break down intocrosscurrents of babble, the conversation will become unintelligible and the observerswill be lost. The moderator’s challenge is to maintain rapport with participants withoutlosing control to the group itself or becoming personally involved in the discussion.

Specific Moderating Skills In the previous section, we discussed rapport and how tocreate and maintain it. In addition to creating and maintaining rapport, the moderatorsshould be:

● capable of maintaining firm but quiet control;● noncritical and nonjudgmental of the opinions expressed and able to keep their

own biases out of the discussion entirely;● alert and interested without showing emotion;● even-handed in seeking equal participation;● encouraging respondents, especially the reticent who may need special attention

and stroking to be forthcoming;● able to keep the topic reasonably on track;

Framing is putting an idea or asituation in the context chosenby the moderator.

Reframing is removing an ideaor a situation from the currentcontext and putting it in thecontext chosen by the modera-tor.

162 | CHAPTER 5

● credibly naive so as to induce discussion of key fundamentals that respondentsmay otherwise assume are self-evident;

● sensitive to undercurrents and disagreements;● flexible, adaptable, and able to improvise;● observant of cues from body language;● able to interact, probe, and bring out feelings and emotions.

When a moderator asks a question, he or she does not settle for the first answer butprobes over and over again to get at deeper levels of perception. The ability to uncoversuch material is one of the most important differences between qualitative and quanti-tative research. The moderator achieves this in several different ways: by encouragingthe respondents to expound on their views in greater detail, by feigning noncompre-hension, by deliberately playing the devil’s advocate, by reframing, by questioning bodylanguage, and by questioning without words (with looks, gestures, etc.). A good moder-ator never takes anything for granted. Moderators have to think on their feet. Unlikesurvey researchers, they do not have the luxury of pre-testing questions to refine theireffectiveness. Every moderator miscue or error may potentially affect the respondentsand influence results.

The moderators’ toolbox consists of a wide variety of tools that enable them toachieve their objectives. We shall examine some of these tools in the following sections.

Hypothetical Scenarios. If the discussion is about alcoholic beverages, the moder-ator may introduce several scenarios such as dinner at home, dinner in a restaurant,relaxing in the afternoon, etc. Such scenarios put the participants in theoretical situa-tions that can make the discussion more realistic and shed light on how different prod-ucts are perceived under different usage circumstances.

Regression. When participants get stuck, removing them from the present and put-ting them in a different context in the past might help the discussion to move forward.A question such as, “What did you think about smoking cigarettes when you wereyoung?” may also be used to gather additional information about participants’ attitudestoward a product.

Silence. In Western cultures, whenever people interact, a level of discomfort quicklydevelops if there is a long pause in conversation, as often happens. The moderator canuse this pause to ask another question, but someone in the group is bound to talk even-tually, simply to fill the gap. Forcing respondents to fill the vacuum when the groupseems to be running out of steam is one way to induce a fresh approach. However, themoderator should be careful not to let the silence go on too long, lest someone say, “Iguess we’re done, then,” when this is not the case.

Negative Questions. Using negative questions is a way to elicit less apparent dis-tinctions that exist in the participant’s mind. Suppose that a participant states she nevergoes to supermarkets. The moderator might simply ask, “What would happen if youdid?” This type of negative questioning is likely to focus the participant’s thinking on thespecific reasons for a given behaviour.

HOW DO TECHNOLOGY-BASED GROUPS WORK?It was inevitable that technology would affect the conduct of focus groups as it has everyother area of research. The current technology explosion has led to a number of technology-based variations of focus groups such as online focus groups, online bulletinboards, video conferencing, and internet viewing. The number of research houses offer-ing high-tech focus groups has been steadily increasing in the past few years and willcontinue to do so.

Hypothetical scenarios putthe participants in particularsituations that can make thediscussion more realistic andshed light on how differentproducts are perceived underdifferent usage circumstances.

Regression refers to mentallyplacing the respondents in anearlier period in time (such aschildhood).

Negative questions deal withcontrary scenarios and focusthinking on the specific rea-sons for a given behaviour.


ONLINE FOCUS GROUPS Online focus groups are conducted at a website set up tofacilitate the conduct of group discussions through personal computers. Online focusgroups have several advantages: participants from a wide variety of backgrounds can berecruited quickly, low incidence respondents can be located faster, geographically dis-persed respondents can be included in the same group, some respondents may be lessintimidated by others in the group, and the full (unedited) transcripts can be madeavailable instantaneously.

There are disadvantages as well. Unless careful checks are made, respondents caneasily falsify their identity including their age, gender, and place of residence. The samerespondent can easily participate with different identities in a number of focus groups,sometimes even with the same moderator on the same subject on the same day. Theresponses in online focus groups tend to be shorter—most people find it easier to talkthan to write. Writing also tends to reduce interaction among participants because writ-ing requires greater effort for most people than speaking. Even more importantly, it isnot possible for the moderator to get nonverbal clues—such as voice modulation orbody language—from the participants. The moderator has no way of detecting subtledifferences in meaning; when the respondent types “pretty good,” the moderator has noway of knowing whether it was “pretty good” or “pretty good.” A moderator also cannotknow whether a respondent is engaged fully in the discussion.

ONLINE BULLETIN BOARDS Here, the respondent is simply asked to check the bulletinboard periodically. The respondent reads the question, thinks about it, and posts ananswer. It is not an online focus group where all participants are present at the sametime. There is no interaction among participants or even any meaningful interactionbetween the moderator and the participants.

Conceptually, online bulletin board discussions come closest to individual depthinterviews. Despite these not being true focus groups, online bulletin boards have somedistinct advantages. Since respondents have time to think about a problem and not beinfluenced by other participants, their answers will often be more detailed and wellthought out. For these reasons, although they cannot be considered as a substitute foronline focus groups, online bulletin boards can be a useful qualitative technique underspecific circumstances.

For an example of how an online focus group looks from the moderator and theparticipant perspectives, see in Exhibits 5-4 and 5-5.

HOW TO ANALYZE, INTERPRET, AND REPORT THE FINDINGSFocus group moderators generally analyze the discussion, interpret it, and produce awritten report. Focus groups are analyzed using a wide variety of techniques and tactics.Either the moderator or a trained analyst listens to the tapes or a transcript is made.Moderators usually also make analytical use of the notes they took during the group. Inaddition, their own recall of the dynamics and nonverbal aspects of the group plays animportant part. The two common modes of analysis are classification and pattern iden-tification.

Classification. The moderator examines every quotation and analytical note andclassifies them under various categories derived from the study objectives. A study outline may yield 10 or more main classifications each divided into subclassifications orthere may be dozens, depending on how fine the breakdowns are and the complexity ofthe subject matter.

Pattern Identification. In this critical phase, the moderator looks for patternsamong groups and tests the data with the following questions:

● Which groups reacted one way, which another? Why?

Online focus groups are con-ducted at one of the websitesset up to facilitate the conductof group discussions throughpersonal computers.

164 | CHAPTER 5

● Were there correlations by group structure or not? Why?● On what did the groups agree or disagree? Why?

From this process, which seeks not simply to describe the responses but to explainthem, come the hypotheses that qualitative research generates. The experienced moder-ator does not interpret the results in isolation but judiciously takes into account pastexperience in the category as well.


RESEARCH TRENDS

Two Technologies Vie for Piece ofGrowing Focus Group Marketby Steve Jarvis, Marketing News, Staff Writer

Until recently, the only way a qualitative researcher couldremotely view a live focus group was through videoconfer-encing, and that required a special focus group facility andIntegrated Services Digital Network (ISDN) phone lines car-rying the video feed to the researcher’s office or anothersuitably equipped place for viewing.

But in just three years, Internet video streaming (alsocalled webcasting or Internet video broadcasting) hasemerged as a practical alternative that is growing rapidly inacceptance. Webcasting now comprises about 15% of thetotal market for focus group video broadcasts—up fromzero only three years ago—and is growing steadily, sayservice providers and qualitative research executives.

Such early enthusiasm for the Net technology has theindustry wondering if it will be the method of choice forbroadcast focus groups, but that outcome does not yetseem imminent. Both technologies are growing in popular-ity—especially in the post 9-11, travel-wary business envi-ronment—and the new Internet video streamingtechnology still has some serious drawbacks.

“I think they both (webcasting and traditional video-conferencing) will grow and I think there is a place forboth,” says John Houlahan, founder and CEO of Stamford,Conn.-based FocusVision Worldwide Inc. His company hasoffered traditional videoconferencing for more than 10years and began offering Internet video services only abouttwo years ago.

Videoconferencing is familiar to most researchers. Atits most basic, a video camera records the focus group’s dis-cussion, and the picture is sent over ISDN lines to a moni-

tor watched by the client’s executives. Although they arenot on-site, with videoconferencing, remote viewers stillmay be able to communicate with moderators by audio orvideo during or after the session.

Internet video streaming is much more an option withthe widespread deployment of broadband technology. Thepicture’s infamous herky-jerky qualities are largely elimi-nated, although the size of the picture—seen via comput-er software such as Microsoft Windows Media Player or RealPlayer—is only about 4 inches square. Also, remote viewerstypically can communicate with the moderator only viatwo-way text chat.

Whereas videoconferencing can accommodate apotentially unlimited number of viewers, Net video stream-ing can be viewed by a maximum of 15 to 20 viewers loggedin at one time. But the biggest disadvantage of videocon-ferencing—and one of Net video streaming’s strengths—isgeography; videoconferencing has to happen at a speciallyequipped facility (that usually is stationary), but Net videocan be observed from any computer with a broadband net-work. (This advantage usually disappears when theresearchers are on the road; broadband Internet connec-tions still are rare in most hotel rooms and airports, and inthose cases executives can only access the session’s audio.)

The Internet option also lets researchers review thefocus group sessions online at any time without the needfor videotapes and VCRs, important advantages that evenlongtime purveyors of videoconferencing services cannotignore.Reprinted with permission from Marketing News, May 27, 2002, page 4.

166 | CHAPTER 5

EXHIBIT 5-4 Online Bulletin Board Focus Group—The Participant View

EXHIBIT 5-5 Online Bulletin Board Focus Group—The Moderator View

ANALYSIS AND INTERPRETATION Although there is no generally agreed-upon formatfor reporting qualitative research findings, Judith Langer (2001) suggests that differentlevels of understanding and the research report should reflect this. She suggests the fol-lowing levels of understanding:

Reporting (what happened). At this level, the analyst looks for what people actuallysaid, often using their verbatim comments. This can be useful in understanding the “lan-guage of the consumer” as well as establishing the difficulty people may have relating to thelanguage used by marketers and advertisers in communicating their ideas. Does an averageinvestor really know what a “fixed-income instrument” is? At this level, the researcher maysummarize the salient points. At this point, as Langer points out, many researchers stopelaborating on their “what happened” level of reporting by simply stating the salient pointsof the discussion with supporting verbatim comments from the participants.

Analysis (what it seems to mean). Here, the analyst goes beyond what was said ver-bally by interpreting the body language of the respondents, commenting on areas ofcontradiction, and offering insights on deeper motivations that manifested themselvesduring the discussion. For instance, was the participant who simply smiled and com-mented that it was good when shown an advertisement being polite or did she really likethe ad? Did the body language of the respondent who agreed with others consistentlyindicate that he was doing so only to be agreeable? Did participants’ positive commentsindicate simple acceptance or great enthusiasm? Which words or phrases did a respon-dent emphasize? Did the participant simply feed back a cliché or genuinely comment onthe subject at hand? Was the participant saying things simply because they were sociallyacceptable? A trained and experienced moderator’s comments in these areas can beextremely valuable to the client.

Targeting (identifying market segments). In well-conducted focus groups, themoderator will be able to identify the breadth of buying patterns and attitudes held byparticipants. Such behaviour and attitudes are indicators of underlying market seg-ments. Each segment could be a group of consumers who may respond to a specificmessage and be motivated by a specific benefit.

Trend Context (what it means today). Since focus groups are open to the views ofthe participants, similar comments offered by respondents in a number of focus groupsmay reveal a trend. For instance, after the events of September 11, 2001, if “personalsafety” was voluntarily mentioned in many focus groups, that dimension may be worthconsidering when designing services such as travel.

Implications (so what?). A good moderator does not simply lay out the problemsbut offers solutions as well. To do this, the moderator needs some familiarity with mar-keting and communication objectives so that the solutions do not sound unrealistic. Infact, the more a moderator knows about the context of the problem, the more effectivehe or she will be in outlining solutions.

There are also computer programs that mechanically analyze the text (Dey, 2000).These programs are not widely used in marketing research.

THE REPORT Even when the client attends all focus groups and has videotape of thediscussions, it is still a near-universal practice to write a report. Because many peoplewho have not attended the focus group will use the findings (e.g., people in the client’sadvertising agency), it is important to have a written report. The standard report runsfrom 50 to 100 pages and covers all aspects discussed in the previous section. It may takeas long as four to five weeks to complete a comprehensive report, which includes verba-tim comments of participants and the moderator’s interpretation. In many cases,because a full report takes time, a client may prefer to receive a topline report or a sum-mary report (a longer version of topline report). These reports may range from two to


twenty pages and may be produced within three days or three weeks. These reports arenot substitutes for comprehensive reports since topline and summary reports typicallydo not contain verbatim comments by participants. Another trend is the request for apresentation style report. Here, the report consists of a series of presentation slides suchas the ones produced using PowerPoint. The findings are presented in point form. Thesereports are easier and quicker to write but are unlikely to capture the nuances of the dis-cussion as well as comprehensive reports. A comprehensive report may have the follow-ing format:

1. Introductiona. Background and purposeb. Detailed objectives

2. Methoda. General methodb. Where and how conducted

3. Executive Summary4. Discussion. This part may contain several sections and deal with different

aspects such as brand perceptions, reactions to advertising, price perceptions,reactions to proposed concepts, and so on. (A quantitative statement such as,“Ninety percent liked the product” is meaningless and misleading in the focusgroup context and should be avoided.) The report will typically include manyverbatim comments from the respondents.

5. Summary and conclusions6. Additional material. The report may also include other supporting materials

such as discussion guides. When the audio is transcribed, the transcript is oftenprovided to the client as a separate document.

168 | CHAPTER 5

Individual Depth InterviewsThe individual depth interview, another common qualitative research tool, is concep-tually similar to a focus group. Here, the moderator talks to one respondent at a timerather than to a group of people. The interviews tend to be more structured than withgroups and have a greater need for moderator intervention. They last about 20 to 40minutes, longer if necessary.

In some instances, it is preferable to have individual interviews instead of focusgroups. Such instances would include occasions where a person’s opinion can easily becoloured by others in the group or when certain opinions are not likely to be expressedin a group because they are too personal, sensitive, or embarrassing in nature. In addi-tion, when it is as important to learn what people do not know about a subject as whatthey do know, the individual interview is the best approach. In a group setting, moreknowledgeable respondents quickly teach less knowledgeable ones, making it very diffi-cult to explore areas of ignorance or misperception. In other instances, participants maybe reminded of issues that they normally do not have because someone in the groupexpressed strong concerns. Again, in a group setting, some respondents may offer com-ments to impress other people. Whenever we suspect factors like these may distort focusgroup findings, we can use individual depth interviews. Sometimes, there is no choicebut to conduct individual interviews. An example of this is when executives from competing firms are interviewed. They will not open up in front of each other and mayeven intentionally make misleading statements.

Although there are differences between focus groups and individual depth inter-views, the skill set required of the moderator is common for both techniques.

In the individual depth inter-view, the moderator talks toone respondent at a timerather than to a group of people.


RESEARCH TRENDS

Qualitative Research: A Case Studyby Susan Bardwell, Partner, In-Sync

Fairmont Hotels & Resorts is a North American–based chainof luxury properties, boasting landmark locations such asThe Plaza in New York City and The Royal York in Toronto.In response to competitive hotels branching out to includespas, Fairmont decided they too needed to include a spafacility. The strategic challenge was to create a distinct spapositioning that balanced Fairmont’s existing image of his-toric and authentic elegance with the needs and desires ofspa users. Fairmont approached In-Sync of Toronto, a cus-tomer insight company, to uncover consumer insights thatcould craft a compelling strategic solution.

In-Sync began by identifying the target market as twodistinct groups: “Spa Enthusiasts,” busy women from 35 to50 years old who visit urban spas regularly with friends oralone, but not as part of a vacation; and “Luxury ResortGoers,” couples in their 40s and 50s, with two or more chil-dren, who work in demanding jobs and prefer a luxuryresort as the perfect getaway.

With the segments identified, In-Sync employed theirMultiple Lens approach which leverages a diverse set ofangles and methodologies that work in tandem to unearthpotent consumer insights. This encompassed a sequentialdesign of Anthropological Cultural Imprints, Hot Buttons 1

and Innovative Qualitative Focus Groups.The purpose of Anthropological Cultural Imprints analy-

sis is to expose the hidden cultural forces that shapehuman behaviour and beliefs about health and beauty. Thetwo components of this phase were: Anthropological Digand Walk in Their Shoes. The Anthropological Dig constitutedsearching out the wealth of untapped studies and literaturethat deal directly with health, beauty, and the current cul-tural trends of people improving their minds, bodies, andsouls. The Walk in Their Shoes entailed a cultural anthro-pologist spending two to three hours with individualrespondents in their homes, observing and exploring theirlifestyle and surroundings, and discussing their needs andaspirations as it pertained to luxury resorts/spas. A total ofnine Walk in Their Shoes were conducted, three in each ofToronto, New York, and Santa Monica.

The researchers collaborated internally to generate aset of Hot Buttons, individual statements relating to spasthat captured strong consumer sentiments. They were to be

used in the groups to provoke discussion and contribute tothe strategic learning. The anthropology insights concen-trated largely on gender and language, gender and bodyimage, humans and nurturing, and body and mind, whichhelped to produce four themes for the Hot Buttons: HigherEmotional Purpose; The Experience Itself; Comfort Level;and Environment, Atmosphere, and Process. A couple of HotButtons examples are: “I love switching off when I go to aspa,” and “Spas are not about setting and achieving goals.”

The next step was initiating the Innovative QualitativeFocus Groups. A total of six two-hour focus groups with sixrespondents in each were conducted in Toronto, New York,and Santa Monica. The “Spa Enthusiasts” were separated intotwo categories, women and pairs of female friends, whereasthe “Luxury Resort Goers” consisted of men and couples.

Completed by respondents prior to the group, theHomework was to bring in two items: one concerning howthey personally feel before going to a spa, and one regard-ing how they feel afterwards. This proved powerful inrevealing consumers’ inner feelings, the stark contrastbetween before and after, and the vocabulary around theexperience.

While waiting for the groups to begin, respondentscompleted a Lobby Exercise in which they wrote down threethings that they aspired to change about themselves ortheir lives. Revelations were in line with the Anthropologyfindings, which centred on losing weight and surface- oriented aspects.

The research presented Fairmont with various keyinsights into how they should launch their new spas. Fourcore Cultural Imprints2 surfaced out of the anthropologicalresearch. The first revealed that men focus on depth, as inmuscles, versus women who focus on surface needs like skinand beauty. The second contrasted the language thatwomen use when talking about the spa, such as pamperingand indulgence, with how men avoid using this languagealtogether because, historically, masculine cultural valuesnever permitted them to. The next imprint emphasizedthat, while the need for and appeal of touch is deeplyingrained in humans since birth, the adult world lacks thisnurturing even though it is hugely valued. Lastly, it wasshown that Western culture deconstructs the body into

Projective TechniquesIn addition to the techniques described, a moderator can use a wide variety of othertechniques. These are used either as a part of focus groups/individual depth interviewsor independently.

WHAT ARE PROJECTIVE TECHNIQUES? Projective techniques are an indirect way of gleaning insights into human behaviour.These techniques are based on the work of psychoanalysts who claim that we attributeour less socially acceptable feelings, emotions, and motives to others. For example, wemay be too embarrassed to admit (or may not even know at the conscious level) that weare highly concerned about our appearance. Yet, if we are asked to write a story basedon a photograph, we may freely project our concerns onto the person in the story andstate that he or she seems highly concerned with his/her appearance.

PERSONIFICATION Personification is a commonly used method. In this technique,respondents are asked to assume that a product is a person and have the respondentdescribe this person. For example, if a Toyota and a Jaguar were people, what humancharacteristics would you attribute to them? Personification is relatively easy to interpretand to understand. Due to its reasonably common reference points, it is widely used.However, it only lends itself to subjects in which it is reasonable to assume that respon-dents will be able to carry the task out because the product has a well-established anddistinctive image. Trying to personify such products as two obscure brands of allergyremedy would be asking too much. The test for the appropriateness of any projectivetechnique is for the researchers to ask themselves whether or not they could respondfairly to the question. A question that leaves researchers uncertain about how to answeris bound to leave respondents scratching their heads.

TRANSPOSED VOCABULARY In the transposed vocabulary technique, which is simi-lar to personification, respondents are asked to assume that a given product is actually

170 | CHAPTER 5

pieces rather than embracing it as an organic whole, andconsequently the body and mind become disconnected.

The qualitative work built on the anthropologicalinsights revealed that consumers see a significant role forspas outside the realm of mere beautification and indul-gence. Ideas that safely promote the caring of mind andbody surfaced as having the greatest potential. The terms“nurture” and “well being” resonated with spa usersbecause the words themselves suggested warmth, caring,and the reunification of the body and mind.

Based on the insights culled from the anthropologyand groups, the researchers recommended the followingpositioning platform, which was adopted by Fairmont in the2001 launch of their spas Willow Stream: Fairmont spas are

a great place to nurture your sense of well being. Theinsights also impacted the presentation of the actual envi-ronment, the packaging and delivery of services, and thelanguage used to describe the women’s versus men’s prod-ucts. For example, all spa services, no matter what aspectof the body they focus on, finish with a touch to otherparts of the body to leave customers with a reconnectedwhole feeling rather than a deconstructed one.

1 Anthropological Cultural Imprints, Hot Buttons are trademarks of

In-Sync.2 Cultural Imprints is a trademark of In-Sync.

Permission granted by The Fairmont and In-Sync Research.

Projective techniques are anindirect way of gleaninginsights into human behaviour.They are based on the idea thatpeople will subconsciouslyattribute to others what theyfind unacceptable in them-selves.

Personification is a techniquein which respondents are askedto assume that a product orservice is a person and toattribute human characteristicsto that person.

In the transposed vocabularytechnique, respondents areasked to assume that a givenproduct or service under con-sideration is actually a differ-ent product.

a different product. How would the products be described? For example, what if IvoryLiquid and Palmolive Liquid were not detergents but clothing stores? What would theylook like? What lines would they carry? How would they arrange their displays? Whowould shop there? How would they advertise? Such transposition is expected to removethe respondents’ mental blocks and release them from rational and linear thinkingthereby providing valuable insights to the researcher.

TRANSFERENCE Similar to projections with reference to its psychological roots, thetransference technique is particularly suitable for socially sensitive issues. An exampleof the use of this technique would be: “That is very interesting, but I bet you know peo-ple who wouldn’t share your viewpoints. What do you know of the way others thinkabout this?” Thus, by removing from the respondent the responsibility for the opinionsexpressed, the researcher may be able to obtain a less inhibited view of the underlyingmotives behind the consumer’s behaviour. There is a danger here: it may be impossibleto know whether the respondent shares the transferred views or is merely being a goodreporter.

SENTENCE COMPLETION Here, the respondent is given an incomplete sentence and isasked to complete it. The theory is that, in so doing, the respondent will reveal her or hisbiases, preferences, and hidden motives. An example of the sentence completion tech-nique is: “From what I’ve seen, people who only fly with Air Canada are....” (The respon-dent is expected to complete the sentence.) There are several variations of thistechnique. Respondents may be quick to recognize this for the manipulation it is. Moresophisticated respondents may refuse to play along by pointedly doing no more thanreiterating what they have said before in an effort to dissuade the moderator from try-ing again. They may also become annoyed. However, these drawbacks can be true of anyprojective technique.

HALF-QUESTIONS The half-questions technique is a variation of the sentence com-pletion technique—except that the respondents do not know they are filling in a sen-tence or a question. For example, the researcher may state, “If I understand youcorrectly, this commercial is getting at...,” and then seems to draw a blank. As theresearcher fumbles for words, the respondent tries to help out with a suitable comple-tion without realizing that the researcher had no intention of finishing the sentence inthe first place. This approach is much less likely to be problematic with sophisticatedrespondents who may even think their wisdom has come to the rescue. Here, too, dis-cretion is called for. Doing this more than once or twice in a group can convey themetamessage that the moderator might be daydreaming or has trouble understandingwhat is being said, to the detriment of rapport.

CARTOON TEST In a typical cartoon test, two characters are shown. One of them maysay something like, “I heard that John bought a Ferrari” (see Exhibit 5-6). The balloonfor the other character is empty and the respondent is asked to fill it in. The responsethus elicited (such as, “John loves fast cars,” “John is always trying to show-off,” or “Johncould have got a good sports car for a lot less money”) is expected to provide clues tounderlying motivations.

PHOTOSORT In the photosort technique, a deck containing photographs of people(e.g., a businessperson, student, homemaker, and manual worker) are given to

The transference technique isa projective technique thatdeliberately asks the respon-dent for opinions that might beheld by others.

In the sentence completiontechnique, the respondent isgiven an incomplete sentencewith regard to the product orservice and is asked to com-plete it.

The half-questions techniqueis a technique in which theinterviewer deliberately stopsin the middle of a question ora statement expecting therespondent to complete it.

In a cartoon test, one of thecharacters is depicted as sayingsomething about the product orservice. The balloon for theother character is empty andthe respondent is asked to fillit in.

In the photosort technique, adeck containing a number ofphotographs of people is givento respondents and they areasked to sort these as users ofdifferent products or differentbrands of a product.


respondents and they are asked to sort these as users of different products or differentbrands of a product. The photographs associated with different brands are expected toreveal the brand image held by customers. As an example, in the photo below, a numberof images of people are shown. Respondents may be given a number of photographs likethese and asked which make/model of automobile each of these people is likely to drive.

A photosort can also be done with more abstract photographs (a tightrope walker,an orchid, a hammer) to elicit respondents’ emotional, nonrational responses to the sub-ject at hand. It is critical, in this case, to probe respondents to understand the associationseach photo holds for them. For example, the moderator might assume that the life of a

tightrope walker represented a complex and challenginglife. However, for a respondent, the tightrope walker couldconjure up happy childhood memories of a circus visit.

OTHER PROJECTIVE TECHNIQUES In projective tech-niques, the respondent projects his or her own ideaswhen presented with half-questions, incomplete sen-tences, or personification. There are several other relatedtechniques, although they are less frequently used.(Projective techniques as a whole are in fact only usedwhen merited, not as a regular part of focus groups.)

Some techniques may be more helpful with certaingroups of people than with others. For example, tech-niques based on drawings and cartoons can be helpfulwhen dealing with children. When respondents arelocked into some form of linear thinking, thus providingno new insights into the problem, fantasy or hypotheticalscenarios may be helpful to reorganize their perceptions.

It may be productive to try role-playing techniqueswith teenagers who can be very shy and hard to motivate.

Role-playing can be helpful with adults as well, especially when dealing with inherentlyinteractive situations. For example, it may be an ideal way to probe customer-salesper-son relations.

172 | CHAPTER 5

EXHIBIT 5-6 Cartoon Test

Photo by Charlotte Morrison-Reed.

Free-choice rankings of products can also provide much insight. Here, the respon-dents are presented with an assortment of related products that are to be arrangedaccording to any criterion chosen by a given respondent; subsequent rankings by othersare to be based on different criteria. One can even use masking tape to create a productsegmentation grid on the discussion table and allow the group to place actual productson it as a reflection of how the products are perceived.

USE PROJECTIVE TECHNIQUES WITH CAUTION In general, projective techniques areused in conjunction with focus groups. There are several things that one should bear inmind when using projective techniques. First, the interpretation of what the respondentsays is not always clear-cut. The statements may be open to more than one interpreta-tion and there is a possibility that the moderator may project his or her own biases ontothe respondent. Second, some respondents may not want to respond to projective tech-nique type questions; they may even be dissuaded from further participation. Third,projective techniques leave room for client observers to overlay their own biases in theway they interpret the responses, resulting in conflict later when they try to decide whatit all meant. Notwithstanding all these limitations, projective techniques can be veryeffective if used properly.

Other Qualitative Research MethodsIn addition to the above, there are many other qualitative research techniques. Theyare either derived from other disciplines (such as ethnography) or a specific configu-ration of known principles and techniques (such as Zaltman Metaphor ElicitationTechnique).

ETHNOGRAPHIC, ANTHROPOLOGICAL, AND OBSERVATIONAL TECHNIQUESInstead of inviting consumers to come to a central location (as is done in traditionalfocus group interviews), the researcher could actually go directly to locations where con-sumers use or buy the product in their everyday lives. This is called ethnographicresearch. For example, instead of asking consumers how they buy their drinks, we canactually go to a bar and observe how people order their drinks and how the group theyare in influences them. This general approach is called “ethnographic,” “anthropological,”or “observational” or more mundanely as “onsite” or “in-home” studies.

In observational research, the researcher watches consumers in a careful anddefined way as they use or buy a product. The process of observation can be confined toone usage/buy cycle or can extend over several. The study can be purely observational;the observer makes sure that his or her involvement in no way influences the behaviourof the consumer or participant. In participant observation, the observer (researcher)interacts with the consumer.

Like all other methods, these techniques have both advantages and disadvantages.Their main advantage is that the observations take place in the “real world” and areunprompted. The observer can notice things that even the consumer may not be awareof. Their main disadvantage is that buying and usage behaviour may be subject to cuesthat the buyer, but not necessarily the observer, may be aware of.

SEMIOTICSSemiotics is the study of signs, symbols, and communication. Semiotics is based on theconcept that all human communication is text that needs decoding. For instance, whatdoes a brand name mean to a consumer? How do the different elements on the package

Semiotics is the study ofsigns, symbols, and communi-cation based on the conceptthat all human communicationis made to texts that needdecoding.


(e.g., font, colour) fit with the brand? How does the music in a commercial relate to thevisuals? The underlying premise of semiotics is that there is an encoding process thatunderlies communication processes. Different elements of a product (e.g., packaging,claims) develop a message that contains the information or meaning the sender hopesto convey. To better understand the symbolic meaning that might be conveyed in a com-munication, some qualitative researchers have begun focusing attention on semiotics.From a semiotic perspective, every marketing message has three basic components: anobject, a sign or symbol, and an interpreter. The product is the object and it is the focusof the message (e.g., canned peas). The sign is the sensory imagery that is used to rep-resent the intended meaning of the object (e.g., Jolly Green Giant). The interpretationis the meaning derived (e.g., fresh, cheerful). The message must be put into a commu-nicable fashion that is appropriate for the medium being used. Qualitative researcherswho specialize in semiotics aim to help us understand what is being communicated tothe consumer.

LADDERINGLaddering is based on means-end theory (Gutman, 1982) and holds that the attributesof a product or a service can be viewed as means to an end. For instance, one of theattributes of education is that it enables a person to get a good job (means) that leads toa feeling of security (end). Every product or service has several attributes. The objectiveof laddering is to connect the various attributes of a product or a service to the under-lying motivators or “ends.” In other words, means-end theory does not settle for the firstreason given by the consumer for using a product; instead it tries to explore the under-lying motives or the ends. (See Exhibit 5-7.)

In laddering techniques, the respondent is probed until he or she comes up with the end value.

Such probing leads the participant from surface attitudes to consequences to end values (called

ACV model). The moderator probes the respondent until end values are arrived at.

I like this brand of shampoo. (A—Attitude)

Why do you like this brand of shampoo?

Because it is of high quality. (C—Consequence)

What does a high-quality shampoo do?

It makes my hair look good and others notice it. (C—Consequence)

Why is that important?

It makes me feel sophisticated. (C—Consequence)

So... ?

I feel good about myself. (V—Value)

In the above example, the interviewer has led the participant from an attitude to var-ious consequences to an end value (self-esteem). The data collected are analyzed usingseveral techniques. Standard analysis results in a visual display such as tables and charts.

Step 1: Content analysis. This involves going through the interview transcripts andidentifying attitudes, values, and consequences.

Step 2: Constructing the implication matrix. In this step, the analyst constructs amatrix that shows how frequently each construct (be it an attribute, consequence, orvalue) is associated with every other construct.

Step 3: Constructing the hierarchical value map. This is a visual representation of theimplications matrix. A typical hierarchical map shows how attributes, consequences,and values are related.

Step 4: Determining the dominant perceptual orientations. In this step, the analyst

174 | CHAPTER 5

Laddering holds that theattributes of a product or aservice can be viewed as meansto an end. The aim of the tech-nique is to identify the under-lying or the “end” motives ofconsumers.

tries to understand the pathways that lead from the bottom to the top. A formal numerictable that explains the dominant perceptual orientations can be created at this stage.

ZALTMAN METAPHOR ELICITATION TECHNIQUE (ZMET)Zaltman Metaphor Elicitation Technique (ZMET) was developed by Gerald Zaltman(Harvard University) and Robin Higie Coulter (University of Connecticut). ZMET isdesigned to elicit the mental models that drive consumer thoughts and behaviour. Theobjective of ZMET is to explicate these models in actionable contexts using consumers’metaphors.

In a typical ZMET study, a small group of consumers (about 20) are recruited to par-ticipate. The study involves consumers taking photographs and collecting pictures fromsources such as magazines, newspapers, and books to show what a brand means to them.The respondents come for a two-hour, one-on-one interview. The interview, known as“guided conversation,” involves several tasks such as storytelling (describe the content ofeach picture), missed images (pictures that the consumer could not find along with anexplanation of their relevance), sorting (sort pictures into meaningful groups and labeleach group), construct elicitation (produce basic constructs and their interrelationshipsusing Kelly repertory grid and Laddering techniques), most representative picture (pic-ture that represents the brand), opposite images (picture that is opposite of the brand),sensory images (what does and does not describe the concept in terms of colour, emo-tion, sound, smell, taste, and touch), mental map (consumers are asked to create a mapor causal model that connects all the constructs elicited in the process of the guided con-versation), summary image (consumers create a summary image representing the coreissues), and vignette (consumers put together a vignette or short video to help commu-nicate important issues). Exhibit 5-8 shows an example of a ZMET chart.

HOW DOES BIAS ARISE IN QUALITATIVE RESEARCH?Bias in qualitative research is of three types: study-related, moderator-related, andclient-related. Study-related biases arise out of incorrect problem definition, improper

ZMET is designed to elicit themental models that drive con-sumer thoughts and behaviour.The objective of ZMET is toexplicate these models inactionable contexts using con-sumers’ metaphors.


EXHIBIT 5-7 Laddering Example

Self-Esteem

Impress Others

Efficient

Easy to Deal With

Less Paperwork

Belonging

Socialize

No Rush

Cheerful

They Know Me

Clean Environment

Respect

Courteous

Brief Waiting

Pleasant Wait

Knowledgeable Staff

Extended Hours

Success

Matches Needs

Good Advice

Efficient

Reasonable Charges

Flexible

Study-related biases arise outof incorrect problem definition,improper recruitment, poor dis-play material, restrictions onproper rotation of displaymaterial, an inappropriateresearch location, excessivesubject matter, and unrealistictiming provisions.

recruitment, poor display material, restrictions on proper rotation of display material,an inappropriate research location, excessive subject matter, and unrealistic timing provisions. Moderator-related biases arise out of factors such as the moderator’s inabil-ity to remain dispassionate, moderator incompetence, or inability to control and insen-sitivity to cross-currents and subtleties. Client-related biases arise when the moderatoris inadequately or incorrectly briefed by the client or when the client introduces his orher expectations and biases into the briefing. When biases are detected, they can some-times be adjusted for, provided they are not very serious.

When to Use Qualitative ResearchBy now it should be obvious that qualitative research can be a powerful tool if usedappropriately and carried out by a competent researcher. However, qualitative researchis often misused. Many neophyte research buyers tend to think that qualitative researchtells them all they want or need to know. Another tendency is to increase the number ofgroups unnecessarily and use them as a substitute for quantitative research. Someresearch buyers have an inadequate understanding of quantitative research because ofits technical nature and feel comfortable with qualitative research even when it is not

176 | CHAPTER 5

EXHIBIT 5-8 ZMET Chart

CostlyAdd-ons

Too manyOptions

SleekImpresses

OthersAdmirationPrestige

Automobile ExpenseSocial

Acceptance

GoodControls

Ease ofDriving

LeisureDriving

ComfortOriginatorConstruct

ConnectorConstruct

DestinationConstruct

Carefree

Enjoyment

Freedom Respect Comfort

Varietyof Design

ProjectSelf-image

Self-Esteem

Moderator-related biases ariseout of factors such as the mod-erator’s inability to remain dis-passionate, moderatorincompetence, or inability tocontrol and insensitivity tocross-currents and subtleties.

Client-related biases arisewhen the moderator is inade-quately or incorrectly briefedby the client or when the client introduces his or herexpectations and biases intothe briefing.


quite appropriate or adequate. All such misuse eventually diminishes the usefulness ofqualitative research by forcing it to do things that it is not designed to do.

Qualitative research should be used when it is the correct approach to the problemat hand. Although relatively inexpensive and quick compared to quantitative methods,qualitative research should not be used for time or budgetary reasons when a survey isreally called for. Some of the specific applications to which qualitative research is wellsuited include:

● brand or service positioning exploration;● new product or advertising concept screening and refinement;● assessment of brand or corporate imagery or general competitive environment

assessment;● packaging and graphic design development.

Generally, qualitative research works when:

1. there is an information vacuum that needs to be filled before the topic can bemeaningfully and effectively researched by quantitative methods;

2. there is a need to understand the basis of consumer motivations, priorities,anxieties, hidden agendas, patterns of perception, desires, and so forth that maynot emerge fully when investigated quantitatively;

3. several initial possibilities are boiled down to the more workable and promis-ing possibilities;

4. it is used as input to the creative process;5. the subject matter is too subtle to create efficient direct questions for use in

standard survey research techniques.

Qualitative research is unsuitable when our objective is to obtain a quantitative estimate of any parameter such as market share of a brand, the extent to which a brandis liked or disliked, or the extent to which consumers have changed their perceptions.The results of qualitative research cannot be meaningfully generalized to the populationbecause we cannot know whether the people in a focus group represent anyone otherthan themselves since the sample size is too small and they are not chosen using any sci-entific means (see Chapter 4). While qualitative research may provide many valuableinsights, its results cannot typically be generalized to the population without supportfrom quantitative research. It is very important to bear this in mind because errors inthis regard can prove to be expensive for the company commissioning the research.

Exhibit 5-9 lists instances where qualitative research is appropriate and instanceswhere it is not.

EXHIBIT 5-9 When and When Not to Use Qualitative Research

EXAMPLES OF APPROPRIATE ISSUES FOR QUALITATIVE RESEARCH

● We’re not sure who our market is. Who uses our product and why?● We don’t know how to speak to our market. What language and images can we use?● We don’t know what consumers think of our new product idea. What are the selling

points? How can it be improved?● We don’t know why they’ve stopped buying our product. We want to uncover the real

reasons beyond the pat answers of time and price.● We know we need to do something to boost sales, and we don’t know what it is.● We need to create new products, but we don’t know what they should be.● We need new ideas for an ad campaign.● We need to understand what our advertising communicates.● We want to know if a new program we are developing appeals to our users and how we

can improve it.● We need new ideas for packaging. We want to know if consumers can use our packaging.● We want to know what our brand image is vs. our competitors’.● We want to know how we market our audience to our advertisers (for magazines and

newspapers).● We want to know about our customers’ lifestyles, who they are and what they care

about.

EXAMPLES OF ISSUES NOT APPROPRIATE FOR QUALITATIVE RESEARCH:

● We need to make a multimillion-dollar decision based only on this research.● We want a definitive answer on the top-three reasons why people buy our product.● We need to find out the demographics of our customers.● We need to know the size of different market segments.

Source: Judith Langer (2001). The Mirrored Window. NY: Paramount Market Publishing

178 | CHAPTER 5


1. The main aim of qualitative research is to explorein depth the full extent of opinions held by con-sumers on a given subject. Qualitative researchrelies on semi-structured or even unstructuredprobing of respondents’ opinions in contrast tothe fully structured survey.

2. Focus groups are by far the most frequently usedform of qualitative research. They involve anextended discussion between a moderator and agroup of respondents, typically eight to ten innumber.

3. In a focus group, no set questionnaire is used toelicit opinions. Rather, the moderator works froman outline of the topic areas that need to be cov-ered, usually referred to as a discussion guide.Focus groups are generally recorded onto audio-tapes for later transcription and analysis. In somecases, they are videotaped as well.

4. Criteria for respondent selection are normallyspecified by the client working in collaborationwith the moderator and then passed on to profes-sional recruiters who specialize in assemblingrespondents to take part in qualitative studies.Recruitment procedures are designed to controlthe introduction of unintended recruitment-related bias.

5. Respondents for focus groups may be recruited bytwo methods: the random method or the referralmethod. In the random method, respondents arechosen using mechanical statistical procedures. Inpractice, the most commonly used method is thereferral method. It is a snowball or pyramid tech-nique in which respondents are recruited throughreferrals by others.

6. There are no hard and fast rules with regard to thescheduling of group interviews, although mostconsumer focus groups take place at 6 p.m. or later.

7. It is generally desirable for clients to observe thefocus group sessions as they happen. In fact, this isone of the main spin-off benefits of qualitativeresearch. One-way mirrors and microphonesfacilitate such observations. It is desirable to keepthe observers limited to people with a direct rea-son for being present.

8. A high-tech version of focus groups is video con-ferencing. In video conferencing, the focus groupdiscussion is transmitted live to another locationacross the city or country. A more popular versionof video conferencing is internet viewing wherethe interview is made available live over the inter-net, enabling the client to watch the focus groupfrom anywhere, chat with associates, and commu-nicate with the moderator in real time.

9. The greater the diversity of a market, the larger thenumber of groups needed. The best test for deter-mining the number of groups required is to iden-tify how many segments are necessary in order toelicit all major motivational perspectives relatedto the subject.

10. In qualitative research, there is no set question-naire. Therefore, it is difficult to determine inadvance the course that group discussions mayactually take. It is therefore imperative that themoderator be thoroughly briefed beforehand.

11. In addition to specific moderating skills, focusgroup moderators need a combination of generalskills that include the ability to put respondents atease, control the direction and content of the dis-cussion, and create, maintain, and break rapport.

12. Technology-based focus groups such as onlinefocus groups and online bulletin boards are beingincreasingly used.

13. Qualitative research also includes the use of pro-jective techniques, which can be used along withfocus groups. Projective techniques are based onthe principle that people are not usually aware oftheir motives but would readily attribute them toothers (projection) when provided a proper con-text. Projective techniques include personifica-tion, transposed vocabulary, transference,sentence completion, half-questions, cartoontests, and photosort. Because the interpretation ofprojective techniques is not always clear-cut, theyshould be used with caution.

14. There are also techniques that are developed incommercial contexts. These techniques generallyuse known or assumed psychological principles.Two examples are laddering and ZMET.

SUMMARY

15. Although qualitative research is a powerful tool, itis not appropriate in all contexts. It should not beused as a substitute for more rigorous and statisti-cally defensible quantitative research. Qualitativeresearch is best seen as a complement to quantita-

tive research because it explores areas that cannotbe explored by quantitative research. In dis -criminate use of qualitative research should be avoided.

180 | CHAPTER 5

Key Termscartoon test, p. 171client-related biases, p. 176discussion guide, p. 147focus group, p. 146framing, p. 162half-questions technique, p. 171hypothetical scenarios, p. 163individual depth interviews,

p. 168internet viewing, p. 155laddering, p. 174mini-group, p. 147

moderator-related biases, p. 176negative questions, p. 163online focus groups, p. 164personification, p. 170photosort technique, p. 172projective techniques, p. 170qualitative research, p. 145 random method, p. 153referral method, p. 153reframing, p. 162regression, p. 163respondent selection, p. 150

semiotics, p. 173 sentence completion technique,

p. 171study-related biases, p. 176surface message, p. 159transference technique, p. 171transposed vocabulary technique,

p. 170video conferencing, p. 155writing exercises, p. 157ZMET, p. 175

Review Questions1. What is the essential main purpose of qualitative

research and how does it differ from the essentialpurpose of quantitative research?

2. You are writing a focus group report and you notethat eight out of ten respondents agree with hav-ing traffic cameras at major intersections in yourcity. Why should you not report that 80 percent ofrespondents felt that way?

3. What is the danger from having one focus groupon a topic?

4. What is the basis of requirements (c, e, f, and g) ofthe Rules of Conduct for focus groups in thePMRS Code of Conduct?

5. What are three projective techniques? Give anexample of each projective technique and whenthey should be used.

6. Suppose that you will be the moderator of a focus

group comprised of eight members of your mar-keting research course. The topic is how to bestpromote your academic program to new students.Develop a set of three objectives for the focusgroup discussion and write a discussion guide. Setout the agenda of topics and the number of min-utes for each topic so that the full two hours oftime is allocated among the topics.

7. Find the qualitative research division (QRD) ofthe Professional Marketing Research Society onthe internet. Fully investigate the QRD and theirQualitative Central function. Write a one-pagedescription of the QRD and emphasize the impor-tance of Qualitative Central.

8. Find QRCA on the internet. Describe in a one-page summary the organization’s purpose, bene-fits of membership, and the reasons for using atrained professional moderator for qualitativeresearch.


MINI CASE

Homer’s Supermarket Conducts Focus Groups

Homer’s Supermarket is a retail grocery supermarket chain in your area. They have sixstores with one more under construction. Homer has requested your marketing researchfirm to hold two focus groups in the catchment area that is roughly 20 km in diameter.The broad purpose is to understand where residents shop, their perceptions and prefer-ences for the stores in the area, why they shop at their current store(s), and what wouldget them to switch to a different store. Your job is to organize the focus groups. Homerwants your plan for organizing the two groups on his desk by 2:00 P.M. tomorrow. Developa detailed plan for organizing all of the aspects of the focus groups in point form. Yourplan must include:

1. Your schedule or timeline for the groups.

2. The location for the groups, including a description of the focus group facilities andthe food and refreshments that will be served, etc. (If there are formal focus groupfacilities in your area, you should be able to find them on the internet or in theYellow Pages.)

3. A description of how the participants will be recruited, the recruitment question-naire or qualifications, and the incentives that will be paid.

4. The rough draft of the discussion guide with time parameters.

5. A profile of the focus group moderator.

6. A list of the project deliverables.

Documents

How to Carry Out Qualitative Research · Have an appreciation of other quali-tative research methods such as ethnography and semiotics. 10. Know when to use qualitative research and