Face-to-face interviews are most effective, but telephone interviews can work well, especially if there is prior notification for participants. People are likely to say more than they will write. Participants are motivated by an effective, personable interviewer. This increased rapport means that more interview questions are likely to be answered and the interview is more likely to be completed. Oral interaction provides immediate opportunities for probing, clarifying answers, and providing feedback. Too, visual or non-verbal messages such as body language can be captured.
Participants are often reluctant to participate, perhaps because of the time involved or because it appears to be challenging or may take considerable time. People are more anxious about their privacy and response rates are lower today than in the past.
Unless the interview schedule is carefully done, there are opportunities for the interviewer to lead or direct the participant to some responses rather than others. Leading may be seen in verbal comment or in body language.
Interviewing is expensive, particularly in the time and effort required by trained interviewers. However, the high cost per response must be balanced against the depth of the answers obtained.
Project management skills are needed to manage the process of interviewing several people at different times and, perhaps, places. Strategies will need to be developed to insure a good response rate. Each interview must be conducted and managed to insure good results.
The interviewer must be successful in communicating and motivating. Motivating includes effective communication skills so the two attributes are intertwined. Communication skills include personality and manner so that the interviewer is clearly seen as welcoming and non-threatening. She or he must be a person that the participant would like to spend time with. Creating a comfortable environment is crucial.
Although part of good communication, well developed listening skills are important enough to receive separate attention. Thoughtful, focused listening is essential. Without it, the interviewer will likely miss important messages or misunderstand them.
When possible, the interviewer should be similar in race, ethnic, economic, and social background as those interviewed. For an example, an African-American interviewer is more productive with African-American participants than a Hispanic one.
The interviewer must be able to control the interview and keep it on track. With challenging questions and human nature, this can be quite a challenge, especially if the interview schedule is not as well developed as it might be.
Given the importance of skill and manner, the supervision, training, and management of interviewers is a crucial aspect of success in data collection.
Many cues are available in the interview. Some will be appropriate. Others will not be. Cues should not lead; however, they may be used to reassure participants and keep the interview moving. In their training, interviewers need to become familiar with their own cues. Often, they may not be aware of them. The interview schedule and training sessions should clearly indicate when and what cues are appropriate.
Cues may be verbal and depend both on the words used and the way that they are used. More often, cues are verbal:
Observation is limited to what the interviewer sees and hears. This is the heart of the interview. Inference is a reasonable guess, a suspicion, or a probability. Inferences, when captured, may be in error. For example, a poorly groomed and dressed librarian user may well be well educated and reasonably affluent. Do not be tempted by inferences when observations are required. Capture what the participant says rather than inferring what you believe they "really" mean.
The interviewer must listen carefully and discriminate. Failure to discriminate is a problem that must be avoided. There are three common failures:
Language is always a problem in the interview. Meanings and definitions are inside people and vary from person to person. Their interpretation and response may be quite different than what you expect. Thoughtful definitions and explanations plus some use of cues and prompts can reduce the ambiguity of language
Don't assume that words have but one meaning, especially if you use problematic ones such as "use" of a library or what does it mean to "read" a book.
Feedback, especially during pilot testing, helps to clarify whether or not the right message was heard. Here, a skilled interviewer and thoughtful listening can make a notable difference.
Assume that there will be differences in interpretation and understanding and compensate for that when you develop the interview schedule.
In the research business, bias is unplanned, unwanted systematic influence that yields a response which is not valid. Usually, it is the result of poor preparation and inadequate testing. For example, poorly worded or phrased questions may "lead" the participant to certain answers [but not intentionally on your part].
Attitudes make a dramatic difference and will determine the success of the interview. The attitude toward the participant is crucial. For example, you are likely to devote more time and energy to those that you feel comfortable with. Participants are likely to devote more time and energy to those that they feel comfortable with. Both the interviewer and the interviewee will have attitudes toward face-to-face communication. Similarly, they will have attitudes towards them self or self-confidence. The interviewer is responsible, as much as is possible, for creating or facilitating positive attitudes. Obviously, it is essential that all participants are treated in the same manner, style, and approach.
While it is best to use an inconspicuous audio recorder, you should also take shorthand notes on those elements not easily captured by audio such as body language. Some notes may be taken during the interview. Other notes, a sort of debriefing, should take place as soon after the interview as possible.
Some participants will find either note-taking or audio recording invasive or threatening. Be prepared and have a back-up strategy.
Six broad aspects of the interview should be considered:
For the interview to be successful, the purpose must be clearly and directly specified so that the researcher knows exactly the information that is needed. Objectives need to be limited and specific.
Following this stage, it should be possible to create an agenda for
the
interview, a list of the kinds of evidence to be gathered, and the
priority
or sequence of the topics.
Several concerns should have been raised during the planning
process. In particular, focus on the evidence that you need as the
question is framed and reframed. Preliminary questions should be
created for each problem question. Consider the following points:
Each word must be understandable by likely participants. Consider their economic, social, and political situation as well as their educational level. Mold questions to the participant's vocabulary and depth of understanding. Pretest questions to see how they are interpreted and how appropriate the response might be. Consider linking questions as well as cues and prompts that will create an environment in which appropriate answers are likely.
Never load a question in favor of a particular response by using evaluative words. Words should be neutral of objective, especially when relating to issues that might be controversial.
Each question must be asked exactly the same way for every participant. Do not paraphrase or summarize. Use a script or an interview schedule to be certain.
Determine if the needed response can best be obtained via a direct or an indirect question. A direct question might be: "is your department head an effective communicator"? while an indirect question might be "how would most members of your department evaluate your supervisor as a communicator"?
Using a variety of question types makes the interview more interesting and a better experience for the participant. However, too much variety can be confusing. Too, several questions of the same type can lead participants to skip questions or become bored. Be absolutely certain that the participant has completed one response before asking another.
Open-ended questions have several benefits:
There are also liabilities:
These questions are more common because they require less effort by the interviewer and the participant. They may also have better response rates. Here are some benefits:
Close-ended questions may be placed in four families. Paired comparisons require a choice between two alternatives. For example, do you favor or oppose a constitutional amendment to ban "gay marriage"?
Rating scale questions require a choice between a series of ordinal values. For example, do you strongly agree, agree, somewhat agree ... that tuition should be capped according to the CPI?
Rank order questions require several items to be ranked according to some sort of priority. Most participants cannot validly or reliably rank more than three to five values.
Semantic differential questions require a choice between a series of bipolar adjectives that indicate relative degrees of feeling such as your supervisor is competent or incompetent.
The first few minutes are crucial, especially the introduction. Begin with a honest, welcoming greeting that builds rapport. This includes a self-introduction and positive body language. While conscious of time enough time should be allowed for the participant to become comfortable. This may involve informal, banal exchange about the weather, traffic, or sports. Appropriate humor and personal inquiry often work well. Still, move quickly to the point.
It is important to motivate the participant. These are useful strategies:
Provide an orientation to the interview with some emphasis on these aspects:
The body of the interview is the heart of the matter. With structured questions, the interview schedule [a script for the interview] is followed word by word. With the open-ended or "non-directive approach," there are fewer questions, but the interviewer must be much more skilled.
The inverted funnel begins with specific questions and then moves to the more general. This encourages consideration of specifics before general reaction or conclusion. It may encourage response with specific, easier to answer questions. It may also prime the participant's memory.
Transitions can be difficult to plan and execute, but are most
important. Sometimes it is useful to summarize where you have been and
link that to the next topic. Transitions can also be helpful for
terminating discussion on one question and moving on.
Conclusions must also be considered in the planning process. Each interview must be brought to a comfortable close. Each participant should have a sense of accomplishment and feel appreciated for her participation. The close must also clearly indicate what will happen as a result of the interview.
The physical setting makes a substantial difference. It should be comfortable and informal. It must allow privacy and quiet. Since the setting is often determined by where the participant lives or works, there may be little opportunity to change the setting.
Begin with the assumption that each participant is different and that no one is wholly predictable. Be flexible enough to adapt to situations as they develop.
It is important to establish and maintain a productive climate. The general mood should be comfortable. The interviewer needs to dissipate tension and show genuine concern for the participant. Adopt an appropriate level of formality. Don't be phony or pretend that you are what you are not. Don't forget the continuing importance of body language. Don't make judgmental comments
Allow adequate time for the interview so neither you nor the participant will feel rushed. You must be able to listen analytically. Combine listening, analyzing, and asking questions. Remember that listening is not the same as hearing. Concentrate, be self-disciplined, and resist distractions. Avoid interrupting unless absolutely necessary. Maintain good eye contact.
With open-ended questions and some closed-end ones, you may need to probe. Some responses will be inadequate or incomplete. Problems include:
You will need to consider why responses are inadequate or inappropriate. The participant may be unwilling to respond because of loss of face or similar. The participant may be confused about what is wanted. The participant my lack knowledge or have a poor memory. Avoid a confrontation, but intervene is allowed to do so. Silence can be used to give participants time to think. Neutral phrases such as "I see" or "hmmm" may encourage continued response. It may be useful to repeat or summarize what the participant has said and ask if that is correct. This allows the participant to verify accuracy, demonstrates that you are a good listener, and builds the transition to another topic.
Create, evaluate and test your interview schedule to insure that it is effective and takes no longer than it must. Stay on track. Know when to stop. Recognize the participant's needs. Be as flexible as allowed when problems are encountered. Build upon the cues from the interviewee. Attempt to distinguish between fact, opinion, hearsay, and feelings.
Motivate continuously. Clearly and visibly care about the interview, the study, and the participant. Recognize and value the participant. Be congenial and watch your body language, especially late in the day.
Effective interviewers must be educated and trained. Even those with experience should receive training with new studies and their particulars. The interviewer must be familiar with the interview schedule, the sampling process, and the research study itself. Larger studies will include a data collection handbook for the interviewers with information on:
Supervision is an important part of interview management. Interviewers must be supervised and evaluated. Responses must be checked for legibility and completeness. Normally, in large scale studies, the supervisor would contact about ten percent of those interviewed to verify that the interview took place.
Although not as popular as before with the advent of cell phones and do not call lists, telephone interviewing is allowed for research purposes and may work well. There are assets:
There are also liabilities:
The opening or introduction is much more important since it is so easy for participants to hang up. Do not use the word "survey" since it has so many negative connotations with marketing. Use simple words and phrases. Be as brief as possible. If rating scales are used, they must have few values. Select a sample size that is about three times larger than what you need. Interviewers must have a good telephone voice. Be prepared for some rude behavior. Advance letters and the use of appointments will increase the response rate.