Focus groups are discussions that last one to two hours, usually done with 6-12 people. The discussion is led by a moderator asking open-ended questions. In usability research for web sites, focus groups are used to explore perceptions and gather opinions of your interface and patron needs.

Focus groups can be used to discover people's general reactions to an interface or services. A wide range of information can be gathered in a relatively short time span, and it is a good device to find out what users want from your site. For example, if you are testing your web site with current users, and they all express the concern that the interlibrary loan feature is very difficult to navigate, that is concrete information.

Focus groups are particularly helpful when used in conjunction with surveys. Your library can follow up a survey with focus groups to clarify the issues revealed, and perhaps hear surprising new ideas or concerns. Unlike doing polls or asking a listserv, the strength of this technique is the interaction between participants. With a skilled moderator, the conversation can go beyond "like it, don't like it" and allow new views to surface. Since there is no pressure to reach a consensus, all views can be encouraged and aired.

The "con" side of focus groups is the results cannot be generalized or treated statistically. A focus group analysis is qualitative. Jacob Nielsen cautions that "Since there are often major differences between what people say and what they do, direct observation of one user at a time always needs to be done to supplement focus groups." Some believe that focus groups are not a good way to gather information. Malcolm Gladwell, author of Blink, is quoted in Advertising Age January 24, 2005, "There is very little psychological justification for the notion that you can find out what people think about an idea -- particularly a revolutionary new idea -- by asking them."

However, Microsoft and other high tech companies use focus groups, but they don't just pick random users. They carefully choose the membership of the groups. They find advanced customers by checking message boards and Web sites to see who is answering technical questions about their products. High tech companies seek input from advanced customers to improve current products and products in development, and find their input valuable.

To apply this to a library setting, you may want invite different constituencies to separate focus groups, for example, asking graduate students writing their thesis as one group and computer science majors as another, and library staff as a third. However, recruiting participants is often difficult, involving lots of soliciting and following up with calls and e-mails.

To prepare for a focus group, the steps are to make sure you understand what information you are trying to gather, develop relevant questions, schedule and facilitate the meeting, and then analyze the results. Details of these steps follow.

The first step is to understand what you want to learn. Focus groups are not polls or surveys; they are in-depth, qualitative interviews with a small number of carefully selected people, to help you develop an idea or specific service.

To prepare, you may want to first consider your budget. This is a labor intensive project; time may be the most expensive item. It takes considerable time to plan the sessions, recruit volunteers, and develop your questions. If you hire a professional moderator, that could be a considerable expense, other optional things can add to the cost, including honorariums, refreshments and video taping.

Much of the literature refers to a "trained moderator" leading the discussion. Unless you hire an independent consultant to do this, you need to build the skills to run a focus group or ask an experienced neutral colleague. Practice will help improve skills, so build in a "test" focus group. The moderator cannot be too close to the project under discussion. If it is a general library focus group, it is probably best to not have a person who works at the library conduct the session, since the participants may not want to offend, and the results from a librarian could have bias.

What skills are needed? A skilled moderator has excellent interpersonal skills and keeps the focus! They keep the discussion on track so that none of the material that is intended to be covered is omitted, and they encourage discussion without influencing the answers given. A good moderator ensures that each of the people in the group participates and interacts with the others and that the discussion is not dominated by one or two individuals.

They remain neutral, and encourage responses with occasional nods of the head, and phrases like "uh huh". The moderator steers the discussion, using phrases like, "we've been talking about desktop delivery, and now I'd like to move on to the format of the documents."

The moderator keeps control of the interview and maintains momentum. If the conversation gets off topic, he or she can use phrases like, "that is an interesting point, but let's stay focused on..." If someone is not participating, the moderator can directly ask their opinion.

The next step is membership. Who do you invite to participate? Again, this depends on what you are trying to learn. When developing new features, some gaming companies recruit avid players. Whether you are thinking about offering a new service or building a new building, your must carefully choose your membership, and perhaps plan several groups. If you are gathering data on your library web site, you may want to plan several groups -- one for staff, another for freshman, one for graduate students, and one for teaching faculty. You can plan a practice session with librarians or student workers to help you refine your strategies.

When you choose your target members, you can solicit them in any way that works locally -- send them and e-mail or give them a call, and explain the time commitment and if there is any remuneration. Speaking of this, should you give an honorarium to participants? This it depends on your budget and your situation. You may consider a voucher for the bookstore or ten to fifteen dollars an hour, or you could hold the session over lunch and provide a meal.

The next step is developing your questions. Experts recommend using only about six or seven questions in a focus group, so this step takes some planning. The questions will center on what you need in-depth information about. The questions will be different if you trying to learn if a new service or idea will work, or if you want to understand where a web site is confusing.

Guidelines for the questions include: make questions as neutral as possible, that is, don't use disparaging language or describe something in enthusiastic terms. For example, if you are discussing a service, just name the service and describe it if needed, without using terms like, "popular" or "costly". Make the questions open-ended (i.e. if it can be answered with a yes or no, don't use it). When listening to answers, don't allow one participant to finish another's answer.

Avoid "why" questions. The answer to a "why" question may conclude a false cause-effect relationship. Also, be careful asking "why" in following up with a participant, in case the respondent feels challenged or defensive. If an answer surprises the moderator, instead of asking "Why do you think..." a phrase like, "Can you tell me more about that?" might get a more nuanced response.

Plan the sequence of questions. To get everyone involved, first ask some factual questions, before asking for opinions. For the last question, ask respondents if they have any other comments (try a round robin for this) and if they have any feedback on the process of the focus group itself.

Set the mood: a focus group is an "event" for your participants. Hold sessions in a pleasant room, free of distractions. Configure chairs in a circle or around a table so that all members can see each other. Provide name tags and refreshments.

To facilitate, after introducing yourself and the observer (if used) first review the purpose of the meeting, and how the results will be used. Address terms of confidentiality, who will get access to their answers and how their answers will be analyzed. If you will be using direct quotes, or if your organization requires consent forms for human testing, have participants sign the necessary forms.

Tell the group how long the meeting will last. Most focus groups run one to one to two hours long (overestimate by 15 minutes and do not run over). Go over ground rules, such as, we are here to gather opinions, let everyone participate, and if participants will receive remuneration how that process will work. Explain the means to record the session (note taking, audio or video recorder). Ask participants if they have any questions before getting started with the interview.

At the end of the session, thank the members, remind them their input was helpful, and restate how the day's results will be used. Tell them how to get in touch with you later if they have further thoughts.

Should the event be recorded? It adds cost and complexity to use a video camera, but it gives others of the team a chance to view the results later without sitting in the room, or behind a mirror (possibly influencing the participants). It can also help the moderator build their skills ...did they raise their eyebrows, or frown at a comment? If you do not record the event, it is important to have an non-participating observer in the room (or behind a mirror) unobtrusively take notes.

After the session, write a short summary, grouping results into specific ideas. Include anything that was "not what you expected" and relevant quotes to demonstrate concepts.

What about alternatives such as using electronic discussion lists or group chat for a version of a focus group? You can gather very valid input this way, but it does not include the advantages of being able to observe non-verbal reactions and you can lose or change the group dynamic. Also, the confidential nature of a focus group is lost. Jacob Nielson says, "Posting questions to a newsgroup with an interest in the issues can generate considerable discussion. A disadvantage is that online discussions are difficult (or impossible) to keep confidential unless they take place on an intranet, behind a firewall. Although online forum discussions are unlikely to reflect the average user's concerns, they can be a good way of getting in touch with "power users." These users have needs that will sometimes surface later for the average user. Thus, addressing the power users' needs may be a way of getting a head start on future usability work."