Editor's note. Judith Langer is president of Langer Ascts., New York City.

Having been brought up to be polite, one of the most difficult aspects of learning to moderate focus groups for me was to cut respondents off. In polite company-business lunches, cocktail parties, dates-one doesn't (or tries not to) do this. In focus groups, however, it is often essential:

  • With a typical two-hour focus group and question-packed topic guide, time is tightly limited. There isn't the luxury to allow long-winded or digressing respondents to go on.
  • Dominant respondents are a focus group bugaboo-in fact, one of the main problems with using this rich technique. These are the people who know it all (or think they do), who invariably answer every question first, who are intent on correcting or chiding their fellow panelists, who sometimes succeed in swaying other respondents or, at least, in making others remain quiet. They distort the research process-becoming an obstacle to ascertaining individual opinions and creating a false impression of consensus.
  • Troublesome respondents are not just annoying to the moderator and back-room clients, they irritate the rest of the group. The chatterers, the bullies, the therapy patients telling their woes, the comedians often make other respondents uncomfortable; soon into the group, eyes begin to roll. If the moderator doesn't take action, the good will of the group can be lost.
  • Talkative groups are animated and far more fun than pulling-teeth groups- but it is impossible to hear what anyone is saying. The client is left out of the conversation, the group further deteriorates to elementary school level at recess, the moderator loses control, and the value of the real interactive process (responding to one another's ideas) is lost.

So, early etiquette training aside, moderators must nicely but firmly get respondents to do what they have been instructed to do at the outset-to talk one at a time, to give everyone a chance to speak.

The nice-but-firm approach is critical. Sometimes there is pressure from clients to bring respondents in line right away. Understandably, clients want to get on with the group, but it can be a big mistake for the moderator to cut respondents off too soon or too harshly.

At the beginning of the session, the moderator is working to develop rapport with and among the respondents. They should be made to feel that what they say is interesting; that they are there for a purpose, yet they can enjoy themselves. Being abrupt with a respondent during his/her first or second remark in the first half hour of the session can send the wrong message to other respondents in the group: you'd better be cautious and restrained or I'll do the same thing to you.

It's important to realize that most respondents do want the moderator's and group's approval. If they are pleasantly and subtly restrained, they tend to get the hint that their excessive talking is out of place.

There are the moderator's own feelings to be dealt with, too. Overly talkative or dominating respondents can be highly irritating. They get in the way of our doing our job (I can't find out what I need to; if I don't get this person to talk less, the client will be annoyed with me). And they can be personally annoying, reminding us of people we have difficulty with in our lives or people we would certainly avoid at a cocktail party. One of the challenges, then, is to find a way of exercising control without our personal feelings to come through.

Among the body language and verbal techniques I've found useful are these:

1. Conductor. One of the best pieces of advice I received on moderating is to use my hands as if I were conducting and orchestra. Respondents pick up on this non-verbal behavior; the hand up in a stop pose (maybe I should call this the cop, not the conductor) signals that they should wind up their comments.

2. No eye contact. Another non-verbal approach is to ignore the respondent. Since most people want to be recognized and realize they should be called on before speaking, this absence of contact bothers them. Then, as they behave more as they should, they can be rewarded with recognition.

3. A small pat. This works only with respondents seated close to the moderator. A light, reassuring pat on the arm is a warm way of saying, I'm listening, I understand, it's okay for you to stop now. (Not recommended, obviously, for business sessions.)

4. Playback. One of the reasons some respondents babble on is that they do not feel listened to in their own lives. Certain respondents have a neurotic need to present their problems, or are undergoing a crisis and are looking for moral support. (For instance, during the introductions of a group, a woman once announced that her daughter recently committed suicide.)

Letting these respondents know you've heard them and appreciate what they are feeling helps them stop. Sometimes a simple "I understand what you are saying," is sufficient. Other times it's helpful to demonstrate your comprehension through non-directive feedback-"So what you seem to be saying is..." In a recent group, a single mother went on at length about the problems her daughter had when she had no one to take her to the Father-Daughter dance at school. I told her I had been in a similar situation, that my cousin had taken me and I'd had a wonderful time. While I rarely reveal anything personal in a focus group, this comment did not bias the discussion (about women's lifestyles), but let this woman know I was sympathetic, gave her a concrete suggestion and, by implication, said it's time to get on with the discussion.

5. "Excuse me." Perhaps the most frequent approach I use when everyone is talking is to say, "Sorry, I can't hear you." The implied message is, "Help me out in doing my job, what you are saying is interesting." It is not, "You're being a pain."

6. The tape. While the focus group introduction includes informing respondents that the session is being taped and that they should speak one at a time, this is often quickly forgotten. A simple reminder that the tape is on and that it will be listened to later helps (at least for a while). With children, an incentive to speak one at a time is the promise that they can hear themselves on tape when the session is over. (With adults, humorous threats to make talkers "stay after school" and listen to the tape is another way of making the same point.)

7. "Inadvertent" interruption. Breaking in after the respondents have stated their main point and then apologizing ("Oh, I'm sorry") usually succeeds. The respondent often stops and even apologizes themselves. Such interruptions can only be used "innocently" once or twice in a session, of course.

8. The gavel. When everyone animatedly talking at once, I've sometimes joked about wishing I'd brought my gavel with me. When the hubbub reaches a point where such small jokes can't be heard, I've resorted to improvising a gavel and banging the table with a coffee mug, preferably empty. (Ashtrays used to serve this purpose but now we don't allow smoking in sessions.)

9. "Yoo-hoo." Another hubbub breaking technique is simply to wave and say (loudly), "You remember me," "Gentlemen/Ladies!" Usually this gets a laugh and brings respondents back to attention.

10. Total group involvement. Reminders that "we want to give everyone a chance to talk" signal to the dominant members that they have to share the platform; to the quieter members that their comments are welcomed and, in fact, expected. At the same time, these reminders do not necessarily seem to be addressed to anyone in particular. The comments avoid embarrassing individual respondents and reinforce the need for the group to work together.

Somewhat more pointed are comments to the talkative members about "letting other people have a chance to talk." Most people do not want to appear rude and inconsiderate, so they take the hint (at least for a while when the reminder may need to be repeated).

11. Yes, I remember it well. On occasion one or even several respondents will start reiterating views they have already expressed. This repetition wastes time, of course, and bores other members of the session. A brief, seemingly off hand, "Yes, you mentioned that before," allows the conversation to move ahead.

12. Delayed response. To prevent the always eager respondent answering every question first, the moderator can simply say, "Let's have someone else go first," or "George, please hold back a moment on this one," or call on another respondent. The eager talker will be given an opportunity to speak, but not to take over.

13. Reseating. This is sometimes necessary with children's groups-two (or more) children chat together or, worse, argue and fight. If several warnings about being quiet don't work, having them change seats literally rearranges the relationships in the group. Occasionally there are adults who repeatedly engage in side conversations. If necessary, the moderator can take a break and tell respondents that everyone is going to take new seats, assigning places. This avoids singling out the troublesome respondent.

14. Flattery. A potentially serious problem in focus groups occurs when levels of knowledge vary sharply. Typically, the more informed respondents talk more because they have more to say or because they want to show off. Less informed respondents feel stupid, embarrassed about revealing that they know little about the subject (or, in some cases, that they have never given it much thought). They have to be assured that it's important for us to know what people feel and what they do or don't know, that whatever impressions they have, however vague or incorrect, are helpful to us.

One of the best ways of holding back the "experts" (who may, in fact, be misinformed) is to recognize their greater familiarity with the subject, asking that they wait until others have answered before responding to questions. ("Mary, you know a lot about this, so let me ask you to hold back for a while." ) Seeing themselves on the moderator's side, they will cooperate.

15. Cooling off. Heated debated occasionally occur in a focus group These can be over a controversial issue (one took place prior to the January 16th deadline about whether or not the U.S. should go to war in the Gulf). A respondent might also see another's comment as an implied insult-you can't be a good parent if you feed your children that product. Less expectedly, they can arise on what is seemingly innocuous subject, such as the color of a package. Sometimes, for whatever reason, one respondent takes an immediate dislike to another, using him/her as a scapegoat, mocking or disagreeing with virtually any comment the other person makes.

Whatever the source of the disagreement, I don't believe in allowing it to go on long. If it is (at least on the surface) about the issue at hand, it may be useful to hear out both sides for a short while. There are moderators and clients who see such confrontations as useful in understanding both sides of an issue. My own feeling, however, is that the argument gets in the way of the group, preventing others from talking, as well as making them uncomfortable.

Among the ways to stop these head-to-head conflicts are: "Obviously there are a lot of different views on this subject and it's important that we here these. Now, let's move on [or] give someone else a chance," or "Hold on, let's hear what [the picked on respondent] has to say."

16. Exasperation. This is a technique which should be reserved for repeatedly rambunctious groups. Several years ago a focus group of Detroit men talking about ceiling tile became exceedingly talkative, wandering off the subject. Having tried a number of times to restrain the group, I finally exclaimed, "Could we talk about the subject just for the hell of it." This directness seemed to shock respondents into remembering why they were at the group.

17. Confrontation. The moderator can be forced to deal very directly with a respondent's disruptions if the other approaches haven't worked. This is particularly true of hostile respondents who hate the client's products or marketing in general, or who argue with the moderator. On the few (thank goodness) occasions when this happened, I've found it sufficient to say that I understand their views but that they are getting in the way of the discussion and, therefore, I'd like them to hold back unless they can contribute.

18. Ejection. This is most definitely the technique of last resort. Some clients are impatient to use this approach; on a number of occasions I've heard them relate stories of other moderators who threw a respondent out, clearly impressed by the show of authority. Besides my own personal reluctance to take such a drastic step (unless a respondent reveals a close relationship with the client's competitor), my concern, again, is about the effect on the rest of the group, arousing their anxieties about being kicked out.

There are, of course, subtle ways of handling this; most commonly, a client or the moderator on a break will ask the host or hostess to come in and call the respondent out of the room. (We did this once in the case of a drunk respondent.) Only, if absolutely necessary, should a moderator directly ask the respondent to leave. After the respondent has left, the moderator needs to briefly explain this was needed so that the group could go.

The nice-but-firm approach can help to reduce the problems of focus groups while maintaining the rapport which is essential to their success. With all its rigors and challenges, the talkative group or respondent is definitely to be preferred to the non-communicative one. It's easier and ultimately more productive to channel this energy than to create it. "Shut up" is better than "speak up"- a subject for another day.