Volume 2, No. 2, Art. 7 – May 2001
Phone Interviewing as a Means of Data Collection:Lessons Learned and Practical Recommendations
Lisa A. Burke & Monica K. Miller
Abstract: This paper presents specific and useful suggestions for successfully using phone interviewing as a means of data collection in research studies. The lessons and insights are intended for anyone conducting research with phone interviews, particularly the novice researcher. The authors speak directly to the reader and offer numerous practical recommendations, generally absent from the traditional research methods literature.
Key words: phone interviewing, qualitative research methods, data collection
Table of Contents
2. Lessons Learned in Conducting Phone Interviews
2.1 Pre-interview phase
2.2 During the interview phase
2.3 Post-interview phase
When embarking on a piece of qualitative research to explore the topic of intuitive decision making, we felt armed with the research methods and knowledge necessary to conduct a sound study. Since then, we have concluded that the "nuts and bolts" of conducting qualitative research are often short shrifted in the research design literature and in methods-based coursework. Consequently, one of the more important outcomes from our qualitative study was the accumulation of knowledge gained about a particular data collection method that we had employed in this study—phone interviews. 
Indeed, students in research design courses, in particular academics in their doctoral education, are often only exposed to traditional methods textbooks that discuss field research in a very sterile fashion (e.g., CAMPBELL & STANLEY, 1963; COOK & CAMPBELL, 1979; JUDD, SMITH, & KIDDER, 1991; SCHMITT & KLIMOSKI, 1991). While established research method texts are extremely useful and necessary in helping the novice understand research methodology and basic concepts such as validity, generalizability, and reliability, they are often insufficient for preparing one to actually conduct the study. 
However, comprehensive phone interviews are being increasingly used in multi-stage research studies (e.g., MITCHELL & ZMUD, 1999; WILLIAMS, 1993). As such, what follows in this piece stems from the multitude of lessons learned in conducting a qualitative investigation using phone interviews as the primary data collection method. The points discussed in this paper are critical, especially for the novice researcher, to be aware of and to proactively address before initiating data collection. Next, we present the numerous straightforward and practical suggestions to promote the successful execution of phone interviewing data collection techniques. 
2. Lessons Learned in Conducting Phone Interviews
The insights we have to share in the present paper cluster into three logical categories: before, during, and after the use of a phone-based data collection. We have summarized the highlights of each of these three phases in Table 1, and discuss the details of each phase in turn.
Before the interview
During the interview
After the interview
Table 1: Phone interviewing as a means of data collection 
2.1 Pre-interview phase
The pre-interview phase is principally concerned with those things that need to be dealt with in advance of conducting the study. Planning, pre-testing, and organizational skills are most heavily used in this stage of conducting a study. 
As in any high-quality research plan, a researcher should pre-test the interview protocol, or list of interview questions, before collecting data for the main study. In other words, first conduct a pilot study of your list of interview questions with a group of persons who are demographically similar to your ultimate sample profile. This helps to determine the most logical and smooth-flowing order of the questions. It will also identify wording issues that need to be addressed for clarity sake, which will enhance the integrity of your data. Lastly, a pre-test will shed important light on the amount of time it will take to conduct the interview, which is one of the first questions you will be asked by potential participants. (We will elaborate upon this issue in a later section.) 
We found it useful to communicate our interview questions ahead of time to participants, along with a general introductory letter about our study. This would be especially relevant if you are researching a topic that is abstract, such as intuitive decision-making. Participants need time to reflect and think about their responses, and we found that this padding of time ultimately yielded more thick, rich descriptive data from participants. 
If you want to audiotape the phone conversations, prepare well in advance for exactly how you are going to do it (e.g., what equipment you will use, where you will obtain the equipment, and so on). There are multiple ways one could audiotape the participant conversations, depending upon the sophistication of the equipment at your disposal. For example, you could use a speakerphone and then tape record with some type of portable taping machine. Or you could use a computer software program (e.g., MegaphoneTM), along with a microphone, to dial and speak to the participant, taping the conversation coming through your computer's speakers. 
The speakerphone/tape recording method was used in our study. A couple notable problems were encountered during pilot testing, including low-end speakerphones that tended to "cut out" after a minimal length of time (e.g., 15 minutes). When this would happen, the interviewer had to pick up the phone receiver to continue talking with the person, while at the same time missing the tape recording of the conversation. Before conducting our main study, we therefore obtained a high-powered telephone from the telecommunications department at our school to solve this problem. Whichever technology you choose, conduct several "test" conversations ahead of time with your setup to ensure that: the person on the other end can hear you, the phone does not cut out, you can hear the interviewee, and the equipment can endure the length of your interviews. Also, ensure that background noise is minimal to avoid unwanted distractions during the interview. 
In terms of logistics, you should think ahead about your needs during the data analysis phase. For example, we decided to create for each interviewee a numeric code that we placed on the interviewee's specific protocol sheet (i.e., written list of questions and responses). In addition, we recorded that same numeric code onto the audiotape before we began taping the particular interviewee's responses. This provided a method for matching the written record with the audio taped record of each interview. 
You will need to conduct phone interviews when it is convenient for participants, so that you get as much information as they have to give you. We found participants more responsive to setting up interview appointments via the phone, compared to email. But this process can also become a scheduling quagmire that takes as much time as conducting the interviews themselves. Stress in your communication with participants that you will be making a special effort to re-connect with them at the agreed upon time. Then, pleasantly repeat the agreed upon time and suggest that they record the appointment in their personal scheduler. Be extremely organized in setting up these phone call appointments, given the inevitable re-scheduling that will occur. Keep a daily log of who you call and the result of the phone call (e.g., left a message, set up an interview, and so on). 
If you have used an introductory letter (sent out before making phone calls), expect that many people will not read it, or possibly say they never received the letter (of course, the latter may be true or not). Regardless, your introduction in the phone call is critical because you cannot assume in these situations that the person even knows who you are or why you are calling. Have a prepared script for the introduction that is brief and positive and that emphasizes the importance and usefulness of their participation. Also, realize that if you mention in the first breath that the interview is voluntary, many potential respondents will automatically bow out of the study. We found the following introductory approach most effective:
introduce yourself (prestigious titles help),
identify the sponsor of your study,
give the general topic of the study,
relay the confidentiality of their responses,
explain how the information will be used, and
give an estimate of the interview's length. 
Tell interviewees that you are taping the conversation for data-recording accuracy purposes, but that their comments will be held confidential. By guaranteeing confidentiality, you will not identify participants specifically with anything they have said during their interview. Relay to them that the information will be used only in aggregate when compiling and documenting your research findings. This is a vital piece of information, which if not communicated effectively can easily sway persons not to participate. 
We found it sufficient to tell participants about the need for tape recording right before we started asking the interview questions because at this point the person was already somewhat vested in the interview. Nonetheless, some interviewees continuously gave very short, non-specific answers, which may be due to the act of tape-recording in and of itself. And a couple respondents seemed apprehensive about getting into the specifics of anything, so it was important to keep asking these individuals to elaborate upon their answers. 
We also chose to take notes while the participant spoke, in addition to tape recording the conversation. We found both sources of data to be very useful (i.e., the audiotapes and the interview notes). For example, if the speakerphone malfunctioned in any way, the interviewer could still hand-record information. And the audio taped version was what we used to retrieve exact quotes we wanted to place in our manuscript, using only the participant's title as an identifier. 
Finally, think about whether you are going to promise ahead of time to those who participate some kind of summary of the results. When are you going to mention this to them in the interview? If you do plan to share results with those who participate in your study, you will need to communicate exactly what you will be sharing and the estimated timeframe for generating this report. Get the interviewee's address at the end of the conversation, and be sure to follow-through on your initial promises to participants. 
2.2 During the interview phase
The lessons learned during the interview were also numerous. While some of this information we found covered generally in research methods books, other finer nuances of this process are discussed below. 
The interviewer's style should be friendly, courteous, conversational, and unbiased. As such, interviewers should read questions in a conversational tone and avoid awkward pauses between questions. The researcher must know the interview protocol inside and out so that he/she conveys a professional tone. 
Do not show surprise or disapproval with a person's responses. Given this, it is useful to presume the role of a reporter trying to keep the interview on focus—not that of an evangelist, debater, curiosity seeker (JUDD et al., 1991). Finally, ask every question with the same exact wording, and in the prescribed order, for each interview. This will maintain data reliability and integrity. 
During the phone interview, it is important to get the interviewee to talk as much as possible. Probe any vague and general answers by saying things like "That's interesting ... could you explain that a little more" or "Let's see, you said ... just how do you mean that?" This type of rejoinder entices the participant to expand upon his/her thoughts, without biasing him/her as to what you think of the information provided thus far. Obviously, you do not want to influence any participant's response. In other words, do not suggest a possible answer and also avoid agreeing with a position any interviewee takes. At the same time, you need to get participants to elaborate upon things that you know are interesting (JUDD et al., 1991). 
Ensure you have a mix of open-ended and close-ended questions. It is helpful to have some questions where people respond, for example, in a specific Likert scale fashion (i.e., close-ended response options), so that you have some easy-to-score data. The open-ended questions will then provide you with the rich filler to elaborate upon such responses. If you collect demographic variables, be sure to ask these questions at the end of the interview. For example, we asked people their formal job title, industry, total years of work experience, time with the firm, time in current position, and years of education completed. By asking these questions at the end of the interview, you will avoid enticing participants to guess your hypotheses, or to think about how the answers to such questions "are supposed to" affect their responses. 
Be sure to give the interviewee feedback as appropriate (JUDD et al., 1991). You could, for example, praise an insightful response by saying something like "Uh-huh, this is the kind of information we want" or "Thanks, we appreciate your frankness." If the participant provides an inadequate response to a question, try saying "You mentioned two things ... is there anything else you'd like to add?" When he/she responds with a vague I-don't-know response, then say "Let me read the question again; perhaps I didn't make it too clear" or "There's no wrong or right answer, I am just seeking your opinion." Or you may provide the alternative of revisiting the question at the end of the interview. 
And when the interviewee does not want to give a close-ended response to a close-ended question, you can say something like "Well, in general, what would you say? or "Taking everything into consideration, what do you think?" or "On the basis of the way things look to you now, what do you think?" We found all of these types of responses very useful, and we had to be mentally prepared to respond with them in the appropriate situations. In this way, you will maximize the potential of each interview. 
In terms of the length of the interview, each may vary to some degree from one person to the next. Some people are able and willing to expound more than others. We found that if a potential participant asked how long the interview would take, and the interviewer responded with an answer above 15-20 minutes, most persons either bowed out of participating or said they could not spend that much time. We discovered this is in our pilot study, and it affected our ultimate interview length and the number of questions we planned to ask. Know the length of the interview you want to conduct, determine the length of the interview your participants are willing to give, and ask questions that will fit in that specific timeframe. Again, this is where your thorough pre-testing will pay off. 
2.3 Post-interview phase
There are lessons we learned in the post phone interview phase, as well. Being familiar with these ahead of time will further enhance the quality of your study. Each is summarized in Table 1 and discussed in turn. 
At the end of the data collection phase, it is important to prepare your data for analysis. Immediately after each interview, re-read your interview notes from the phone conversation to make sure they are complete and accurate. Do not procrastinate or put off this task, as enticing as it might be. If you do, then the integrity of your study's data and ultimate findings may be jeopardized. 
Prior to conducting the study, you should have determined the data analysis method that will be used. So once all of your interviews have been conducted, it is time to move into data analysis. First, realize that you are likely to be overwhelmed by the incredible amount of rich, descriptive data in your possession. However, methods exist for making sense of this type of data. While not the focus of the current paper, there are various methods that can be and have been used in qualitative studies (e.g., content analysis, thematic analysis, case analysis, unobtrusive measures, etc.). 
We used thematic analysis (PATTON, 1990). That is, we constantly compared interviewee responses, with the goal of organizing the data into systematic categories of analysis by seeking recurring themes. This was a very time consuming process. We also kept a research project journal during the data collection phase as a third source of data in an attempt to identify themes in the data that we were seeing over the course of time. 
In the end, however, when you plan to start "making sense" of all your phone interview data, do not underestimate the time it will take you to generate themes in the responses and to categorize participants responses into aggregate-level themes. We found that doubling any original time estimates for completion of our data analysis steps was necessary. You will also need to be good at seeing the forest (and not all the trees) or you will not progress sufficiently through the mounds of rich data that have been collected. 
Conducting qualitative field research is dynamic but also affected by many factors that must be proactively addressed by the researcher. What is needed in research methods coursework, at all levels, is a more realistic portrayal of the ambiguity and complexity surrounding field research and suggestions for successfully navigating through the process. This paper represents an initial attempt to address that need. The suggestions discussed in this paper should educate the novice researcher about the reality of using phone interviewing as a data collection technique. 
Campbell, Donald T. & Stanley, Julian C. (1963). Experimental and quasi-experimental designs for research. Boston, MA: Houghton Mifflin Co.
Cook, Thomas D. & Campbell, Donald T. (1979). Quasi-experimentation: Design & analysis issues for field settings. Boston, MA: Houghton Mifflin Co.
Judd, Charles M.; Smith, Eliot R. & Kidder, Louise H. (1991). Research methods in social relations (6th edition). Fort Worth, TX: Holt, Rinehart, and Winston, Inc.
Lincoln, Yvonna S. & Guba, Egon G. (1985). Naturalistic inquiry. Newbury Park: Sage.
Mitchell, Victoria L. & Zmud, Robert W. (1999). The effects of coupling IT and work process strategies in redesign projects. Organization Science, 10, 424-438.
Patton, Michael Q. (1990). Qualitative evaluation and research methods (2nd edition). Newbury Park: Sage.
Schmitt, Neal W. & Klimoski, Richard J. (1991). Research methods in human resources management. Cincinnati, OH: South-Western Publishing Co.
Williams, Mary L. (1993). Measuring business starts, success and survival: Some database considerations. Journal of Business Venturing, 8, 295-300.
Lisa A. BURKE is an assistant professor at Louisiana State University in Shreveport. Her primary research interests include management training, development, and education.
Lisa A. Burke
One University Place
LSUS, Shreveport, LA 71115
Monica K. MILLER is an associate at Kraft Foods in Columbus, Ohio. She was a former student of Dr. Burke's at the University of Dayton, where they conducted research on intuitive decision making.
Monica K. Miller
Burke, Lisa A. & Miller, Monica K. (2001). Phone Interviewing as a Means of Data Collection: Lessons Learned and Practical Recommendations [30 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 2(2), Art. 7, http://nbn-resolving.de/urn:nbn:de:0114-fqs010271.