The Process We Use to Develop a Questioning Route
So how does one go about developing a questioning route for a focus group study? Here is a process that works well for us. Typically we do the following:
1. Brainstorming
2. Phrasing the questions
3. Sequencing the questions
4. Estimating time for questions
5. Getting feedback from others
6. Testing the questions
Step 1. Brainstorming
We invite a few people who are familiar with the purpose of the study to meet to brainstorm questions. We look for people with different backgrounds: expertise in the topic, knowledge of the organization requesting the study, experience with focus groups, and familiarity with the focus group participants. Usually the team includes us, the client, and others whom the client invites. Often we plan for four to six people and a one- to two-hour meeting. We begin by reviewing the purpose of the study and the intended audiences, and then we ask people to throw out ideas for questions that should be asked. One person records all the ideas. People are allowed to comment on questions as they come up, but we try not to get stuck talking about one question. Sometimes the ideas for questions dry up quickly. Then we will ask for ideas in different ways: What would you like to know after we are done? What kind of decision do you want to make? What kind of information would be helpful to you? At this point, we are looking for key questions-those questions that will drive the study. We don't worry too much about the other kinds of questions. After an hour or two, we typically have plenty of questions to begin the next step.
A group is great for generating ideas for questions, but a group isn't efficient for refining the questions. Therefore, we adjourn our brain-storming meeting, and then one or two people take responsibility for the next step-phrasing and sequencing the questions.
TIP
Selecting Useful Questions
After a brainstorming session, we have many more questions than could actually fit in a questioning route. Where do we start? Which ones do we include? It helps to have mental screens for the questions, like these:
Is this a "nice-to-know" or a "need-to-know" question? Nice-to-know questions often arise from curiosity but aren't crucial to the study. Need-to-know questions arise out of a need for information. We include need-to-know questions first.
A variation of the question above is, What would you do with this information if you had it? Is it going to help you move closer to your goal? We sometimes ask clients these questions to help us understand what would be useful to them. We start with questions with the most potential to provide useful information.
Step 2. Phrasing the Questions
Phrasing and sequencing really happen at the same time. The researcher examines the list of questions and begins by pulling those questions that seem key to the study and editing them (phrasing them) so they will work in a focus group (e.g., take out jargon, make them open-ended). Then, as the researcher pulls additional questions, he or she begins to build the questioning route. Let's consider the strategies for phrasing questions.
Use Open-Ended Questions
Open-ended questions allow the respondents to determine the direction of the response. The answer is not implied, and the type or manner of response is not suggested. Individuals are encouraged to respond based on their specific situation. The major advantage of the open-ended question is that it reveals what is on the interviewee's mind as opposed to what the interviewer suspects is on the interviewee's mind. For example, consider these open-ended questions: "What did you think of the program?" "How did you feel about the conference?" "Where do you get new information?" "What do you like best about the proposed program?" "What do you like least about the proposed program?"
Some questions are deceptive and appear to be open-ended but are really closed-ended questions in disguise. Questions that include phrases such as "how satisfied," "to what extent," or "how much" imply answers that fall within a specified range, such as very satisfied, to a great extent, or a great deal. Compare the questions "How satisfied were you with the services you received?" and "How did you feel about the services you received?" The more open-ended question begs for more description, more explanation.
Closed-ended questions aren't totally off-limits. They can provide very helpful information. You may want some very simple information, such as asking kids in a study of school lunch, "How many of you usually bring a bag lunch?" Or, toward the end of the group interview, it may be productive to narrow the types of responses and bring greater focus to the answers by shifting to closed-ended questions. Also, bounding the questions may be helpful to a moderator trying to regain control of a rambling discussion or in situations when the topic requires more specific insights. For example, the moderator might say, "Which of these three options do you like best?"
Ask Participants to Think Back
The "think-back" question asks participants to reflect on their personal experiences and then respond to a specific question. "Think back to when you began working at the public health service. What attracted you to the position?" Or, "Think back to the last time you registered for a course at the university. What was that experience like?" The "think-back" phrase helps establish a context for the response. These words let participants know that you want them to be specific and grounded in their experiences as opposed to "hearsay" from others or just repeating community beliefs and values. People often give great examples of their experiences.
There's a tendency for participants to respond to the more immediate interviewing experience-the here and now-unless you ask them to shift themselves to another timeframe. This focus on the past increases the reliability of the responses because it asks about specific experiences as opposed to current intentions or future possibilities. The question asks what the person has done as opposed to what might be done in the future. The shift is from what might be, or ought to be, to what has been. This time shift cues the respondent to speak from experience as opposed to wishes and intentions.
Avoid Asking Why
The "why" question has sharpness or pointedness to it that reminds one of interrogations. The respondent tends to feel confronted and defensive. Also, "why" questions imply a rational answer. Unfortunately, these "why" questions present problems because in real life, people make decisions based on impulse, habit, tradition, or other nonrational processes. When asked why, respondents feel like they should have a rational answer appropriate to the situation. The participant "intellectualizes" the answer and speaks from the brain and not from deeper forces that motivate behavior.
If the researcher decides to use a "why" question, it should be specific. Paul Lazarfeld (1934/1986) has called this the principle of specification. Lazarfeld's principle of specification is that "why" questions are answered in two ways. When asked why, the respondent may respond on (a) the basis of "influences" that prompted the action or (b) the basis of certain desirable "attributes." "Why" questions can be messy to analyze if participants aren't clear whether you are asking for influences or attributes.
Let's use Lazarfeld's model to examine the responses to a seemingly simple question: "Why did you go to the zoo?"
Influence answer: "Because my kids really wanted to go."
Attribute answer: "Because I wanted to see the Beluga whale."
What seems like a straightforward and simple question can really be answered on several dimensions. The first answer describes an influence, and the second answer relates to a feature or attribute of the zoo. The preferred strategy is to break the "why" question down into different questions. For example:
Influence: "What prompted (influenced, caused, made) you to go to the zoo?" Or,
Attribute: "What features of the zoo do you particularly like?"
A less direct approach is to ask people "what" or "how" they feel about the object of discussion. Often people can describe the feelings they had when they considered using a particular product or program. In addition, they can probably describe the anticipated consequences from using the product or program.
Keep Questions Simple
Beginning researchers tend to make focus group questions too complex. Simple, clear questions are essential. For example, don't ask, "What are the ingredients that are associated with healthy living?" Instead, ask, "Describe a healthy lifestyle." Think of the shortest way to ask the question clearly.
CAUTION
The Limit of Think-Back Questions
"Think-back" questions should be limited to events or experiences that are fairly recent or particularly memorable. If the participants can't readily remember the experience, the question won't work.
TIP
Make Complex Questions Visual
If you have a question that is difficult to put into few words, write it on a flip chart before the group begins and flip to it when you get to that question in the discussion. The visual cue helps people understand and remember the complex question.
The best focus group questions are stated simply. When these questions are asked, the participants immediately know what is asked for, and within seconds they are on their way to providing an answer. By contrast, avoid questions that have multiple interpretations. The participant hesitates because the question is confusing. Then, while thinking, he or she becomes distracted by the comments of other participants and forgets his or her train of thought.
Simple questions do not yield simple answers! It is often the simple question that gets the participant to bring shape and form to the discussion. It pulls out assumptions and lays bare the core principles. You can spot the simple questions because they typically have few words, no jargon or insider language, and no commas, semicolons, or hyphens. The simple question is not condescending or childish. It's a sophisticated question that gets at the core of the topic.
Perhaps the most distinctive feature of simple questions is that they are memorable. Too often, participants forget the question, in part because it is too complex. The memorable question is one that continues to "ring" in their heads. Even if one participant gets off topic, another participant remembers the question and brings the discussion back on track.
Make questions sound conversational, and use words the people in the group would use.
Sometimes we get caught up in our own jargon or the language of our profession. This doesn't work well in focus groups. Insider language is offensive and doesn't communicate well to outsiders. We always try to ask questions in a way that sounds conversational. And we try to use language that is comfortable for the types of people we are asking. If we are talking with special education teachers, we may use more technical language than if we are talking with parents or students. If we are working with health care providers, we may use more technical language than if we are talking to patients or their family members. Sometimes we imagine, OK, if I were in the backyard talking to the neighbors, how would I ask this? We don't use acronyms unless they are well known or we explain them. Our goal is not to baffle them with our vocabulary or puff up our egos but to seek to be enlightened by them.

Be Cautious About Giving Examples
Examples are like mental ruts. Although they provide ideas for the type of response, they also limit the thinking of respondents. Suppose you are doing a study of customer satisfaction, and because the topic is broad, you decide to use the example of how complaints are handled. Well, handling complaints is only one facet of customer satisfaction, and because it evokes memories and vivid experiences, it can dominate the conversation and prevent other dimensions from emerging. If you do give examples, give them as probes after participants have already given their insights.
Step 3. Sequencing the Questions
Now, let's think about the sequence of the questions. Focus group questions are not just thrown together. The researcher arranges the questions with care. This question sequence is the reason we use the word focus in the name. This focused sequence is sensible to participants. It provides an opportunity for participants to anchor their opinions and then build on those views. Here is how we sequence:
General Before Specific
The most common procedure in arranging questions is to go from general to specific-that is, begin with general overview questions that lead to more specific questions of critical interest. Avoid presenting participants with key questions without first establishing the context created by more general questions. For example, suppose a series of focus group interviews will be held with young people. The purpose is to learn their perceptions of youth organizations and eventually to identify an effective means of advertising a particular organization. It would be premature to begin with questions on advertising the organization. Instead, the moderator might ask the participants to describe their favorite youth organization or to describe what they like about youth clubs. Later in the discussion, the moderator might narrow the topic to focus on a specific youth organization under investigation. Perhaps toward the end of the discussion, the moderator might solicit opinions on several different approaches that are being considered for advertising the youth group.
EXAMPLE
Moving From General to Specific Questions
An actual illustration of the general to specific technique of focusing questions comes from Hawaii. To gain insights into how consumers use Kona coffee, the moderator began with questions about gourmet foods, then asked about gourmet beverages. When a participant suggested Kona coffee, the moderator then encouraged discussion of how and when this type of coffee was used.
The funnel analogy is helpful because it presents the researcher with a visual guide for arranging questions. The funneling concept is used to move the discussion from broad to narrow, from general to specific, or from abstract to specific. The funneling begins with fairly broad discussion and is followed by a series of narrower, more focused questions. Just how broad should the beginning questions be? Part of that depends on the number of questions you have and the amount of time scheduled for the focus group.
Positive Questions Before Negative Ones
If you want to ask a negative question, first ask the question phrased in a positive way. For example, if you want to ask, "What don't you like about eating in the cafeteria?" first ask, "What do you like about eating in the cafeteria?" Give participants the chance to comment on both positive and negative experiences or observations. This strategy usually works better when the first request is for positive items. Perhaps it was our mothers' exhortations that we shouldn't say something bad unless we've first said something good.
The benefit of using both positive and negative questions is that it allows participants to comment on both sides of the issue, and in some situations, this is particularly important. At times, focus group participants get in a rut and become excessively critical. It's reasonable for employees of an organization, students in an educational setting, or military personnel to launch into criticism of those who have control and power. In situations when participants begin with negative features and tend to dwell on the undesirable factors, there is value in "turning the tables" and asking for opposite views. "So what are the benefits of working around here?" "What's positive about being a student here?" and so on.
Often the transition from positive to negative aspects is smooth and comfortable, but care is needed so that it isn't premature. One rather predictable scenario is that although positive attributes are being discussed, a participant might disagree with the positive statement and want to offer a contrary point of view. This can easily lead into the discussion of negative attributes without further exploration of the positive features. Here the moderator will need to exercise mild control and encourage the group to complete the discussion of the positive attributes before shifting to the less desirable features.
Uncued Questions Before Cued Questions
The rule of thumb is to ask the uncued question first and then follow up with cues to prompt additional discussion. For example, a moderator could ask, "What are the needs in this neighborhood?" as an uncued question. After people discuss this question, the moderator could list categories that help spur additional thoughts (e.g., children, teenagers, young families, older families, the elderly, or safety, health, child care, jobs) and ask, "When you think of these categories, do any other needs come to mind?"
If the researcher is particularly interested in the needs of teenagers but those needs don't come up in the discussion, the researcher has no way of knowing whether needs of teenagers aren't really important in that neighborhood or if they were just overlooked in the discussion. The researcher has to ask. The cues themselves require some thought. They are developed before the focus group. They are limited in number yet also reasonably exhaustive.
When using uncued and cued questions, it may also be helpful to include an "all-things-considered" question, described earlier. In this question, the participants are asked to identify the one factor (need, concern, etc.) that they consider to be the most important (critical, necessary to address, etc.). Responses to this question greatly aid the analysis. An analysis error sometimes made in focus groups is to assume that what is most frequently mentioned is also most important.
Step 4. Estimating Time for Questions
Another common error of beginning focus group researchers is to try to ask too many questions. We have seen questioning routes with thirty questions. In a two-hour group, that is less than four minutes per question-too little time to expect in-depth discussion. Researchers tend to get superficial, top-of-the-mind information if they don't allow enough time for discussion.
Focus groups are typically two hours long. Successful groups have been conducted in less time, particularly with children or teenagers or on narrowly focused studies. We also occasionally hear of focus groups lasting more than two hours. The two-hour time limit, however, is a physical and psychological limit for most people. Don't go beyond the two-hour maximum unless there is a special event or circumstance that makes it comfortable for participants, such as providing lunch or dinner.
Once we have a draft-questioning route, we estimate how much time we should spend on each question, typically five, ten, fifteen, or twenty minutes. When planning a two-hour group, we allow a little flexible time at the front end, about fifteen minutes. If everyone is there on time, we start right away. But occasionally, we will have to wait until enough people arrive to begin the group. We wait about fifteen minutes and then begin with whoever is there. We also allow time at the end of the group for ending questions and summarizing (at least fifteen minutes). This leaves about ninety minutes for the opening, introductory, transition, and key questions. We add up the time we have assigned the questions and decide if we need to add or delete questions.
When estimating time for the questions, consider the following:
The complexity of the question. Some questions can be answered in a matter of seconds. For example, we occasionally ask very simple questions such as, "How many of you usually eat school lunch? Raise your hand if you usually eat school lunch." Other questions demand more time and discussion, such as, "What do you think of the school cafeteria?"
The category of questions. Opening and introductory questions typically don't take much time. Allow the most time for key questions.
The level of participants' expertise. A group of experts will have more to say than a group of nonexperts on any topic. So experts need more time per question. One way to deal with this is to limit the number of questions to be asked of experts. For example, we may include fourteen questions in a discussion with nonusers of a program but include only ten questions in the questioning route for users of the program.
* The size of the focus group. A group of nine participants will usually discuss each question longer than a group of six participants.
* The level of discussion you want related to the question. If you don't want in-depth information about a particular question, allow less time for it. If you want in-depth data or insights, allow enough time for participants to wrestle with the question.
Step 5. Getting Feedback From Others
Once the questioning route is completed, it is time to send it back to the team that brainstormed the questions for their review. Usually it isn't necessary to physically get back together to review the questions. Often, e-mail or fax works well for getting the draft-questioning route back to people. Ask people to think about the following:
* Are these the right questions? Will they get the type of information you need?
* Do you understand the questions? Are any of them confusing to you?
* Do the questions seem to flow from one topic to another?
* Are these the words that people in the groups would use to talk about the issue?
* What have we missed?
It is not unusual to run through several drafts before the team feels comfortable with the questions. If you get to a point when people start to nitpick, it is definitely time to pilot test the questions.
Step 6. Testing the Questions
Before using the questions in a group, we test them. Sometimes it is as simple as finding a few people who fit the focus group screen and asking them the questions. At this point, we ask the questions as if we were conducting an individual interview. We pay attention to two things:
How easy it is to ask the question? Do the words flow smoothly, or do we stumble when we ask it?
TIP
Put the Date on Each Draft to Minimize Confusion
With multiple revisions, it is easy to get confused about which draft is the latest. Using different colored paper and numbering the draft versions are helpful, but it also helps to date each draft.
A question that seemed simple when we wrote it on the page may be awkward when we ask it aloud. If we stumble, we rephrase it to make it easier to say-more conversational.
* Does the question seem confusing to the participants? Do they hesitate too long? Do they look confused? Do they give an answer that shows the question is confusing? Do they ask for clarification? If so, we ask them to tell us about what is confusing and ask for their help in making the question simpler.
After we have tested the questions with a few people, we hold the first focus group. We don't pilot test the questions in the group. It is so time and labor intensive to set up a group that we want to be able to use the results from the discussion, rather than consider it a pilot. If a question doesn't work in the first group, we revise it before the second group. At the end of the first group, we may ask the participants to help us revise a question that seemed awkward or confusing.