Notes on Tactical Use of Qualitative Interviewing

Notes on Tactical Use of Qualitative Interviewing

Joel Lawton with John Hoven

This article was derived from field notes and experiences during three tours to Afghanistan. The purpose of the article is to help tactically deployed soldiers engage in problem solving through conversational interviewing and application of qualitative methods. This paper serves as a guide for MIL-CIV engagements where situational awareness is critical for operational-tactical planning and to enhance intelligence preparation of the battlefield with local knowledge. This guide is intended to be used by tactically deployed military members, those who have regular contact with local nationals in nations or areas where little is known.

Background

My experiences in qualitative-conversational interviewing were gained when deployed to Afghanistan in 2011 as a Human Terrain Analyst for the U.S. Army Human Terrain System (HTS), Training and Doctrine Command (TRADOC), G-2.  Most of what I know was achieved through on-the-job training through daily interaction with Afghans and my specific area of responsibility (AOR). In due course, I developed several techniques using qualitative methods to engage local populations that revealed rich, locally-nuanced, and actionable intelligence. Some tenets may be taught in a formal setting such as how to structure your thoughts, how to build rapport, managing your interrupter, and interviewing etiquette. However, some of the most important aspects to primary research and interviewing are best learned through conducting “real-world” (i.e., not scenario based) engagements. For me this included understanding body language, knowledge of the AOR, general situational awareness, and developing active listening skills. Initially, I learned rather quickly that the interviewer must have an understanding of many socio-cultural nuances and be able to adjust quickly to interviewee’s level of comfort, body language, tone of voice, signs of deception, and willingness to provide information freely. The purpose of this paper is to depict some of my most successful methods used to elicit rich, locally nuanced, and actionable intelligence information using conversational interviewing tactics.  

Methods

All interviews I conducted had several common elements that should be discussed before continuing. Sometimes having only a basic working knowledge of the AOR, the culture, people (e.g., ethnicities, tribes, demographic composition; etc), local and national holidays, and areas of contention was critical to formulate questions for my engagements. My primary sources for the study of an area before going into it were Mission Command System (MCS) databases such as CIDNE, TIGR, and Map-HT. At times, I would find reports that were generated locally on the Forward Operating Base (FOB) where I was located. Derived from the reports, I would seek out the responsible analyst or person for an informal conversation on the area I was to go. I often found that direct and informal engagements with people who have knowledge or have been in the AORs could convey rich and locally nuanced information that is largely absent from structured reports in MCS databases. For instance, if I found that a Human Intelligence Engagement Team (HET) had visited a village I was going to, I would invite them to lunch or meet them in their work area for a conversation on the topic village. I would record my conversation on paper and essentially use it as a baseline to formulate ideas and topic material for questions. This is critical as it gave me some context to develop my first questions to engage my first interviewee.

Second, after reviewing database reports and collaborating with U.S. or Coalition personnel, I physically observed the village or area during the mission itself. This allowed me to develop starting points or opening questions that structured my thoughts and assisted in developing a line of reasoning leading to an initial question set. I would link my prior knowledge to observables in the village that gave me some topics to talk about. As already noted, before a patrol to a village, I would use previously acquired knowledge, intelligence, and even PMESII/ASCOPE (i.e., acquired from databases) variables to formulate ideas about the area. Upon arriving in the village, environmental and ethnographic observables helped me link/relate what I knew about the village to the “ground truth.” For illustration, if I knew that I was going to a specific village and understood that there was a high frequency of SIGACTs in the area before the patrol, I would look for observable indicators that were security or threat oriented. These indicators may include: no children or women on the streets, people ambivalent to speak, or even direct indictors of insurgent activity. Thus, linking prior knowledge to observed indicators, I had a line of questioning. I would then initially exclude topics related to civil liberties, economy, education, development, or other unrelated security oriented questions. Upon interviewing the locals and subsequently found that maybe the security situation has linkage to a lack of economic opportunities, I would then elicit information regarding the newly discovered independent variable. Thus, I would not limit my line of reasoning and questions to one topic. Further, my questions were never codified and limited to a single line of reasoning that tried to prove or disprove a pre-established hypothesis. My questions always progressively elaborated during pre-mission and the mission itself. In other times, I did not have any prior knowledge (i.e., through the above mentioned reporting mechanisms) of an area and I completely relied on observable indictors to structure my initial questions.

When I knew nothing about the area, I often used the Tactical Combat Assessment Planning Framework (TCAPF) as a gateway tool to start conversation and discover unknown unknowns. I never limited myself to its four standard questions: 1) Has the population of the village changed in the last 12 months?; 2) What is the greatest problem facing the village? 3); Who do you believe can solve this problem? and 4) What should be done first to help the village? In most cases, I operated in an unknown-unknown environment, where U.S. or Coalition Forces knew little to no information about. A little experience and understanding of culture and civil-society allowed me to develop a working hypothesis or discover a known-unknown upon my first interview. My next step was to identify someone for my first interview.

Third, the security environment in Afghanistan and even other nations where the U.S. operates are largely non-permissive, making sampling techniques hard. My primary approach to choosing interviewees was convenience non-probability sampling, as is customary in qualitative research. In other words, I chose the first person that I could find to talk to. However, I did apply some filters to my choice of interviewee. Initially, I tried to identify males, people in non-formal leadership positions, and individuals who resided in the village that I wanted to know something about. I would try to initially exclude those in a formal or informal leadership role, in order to remove as much personal or political bias from responses as possible. My experience led me to believe that key leaders tend to have agendas and will not provide uninfluenced or apolitically motivated information. They thus are trying to influence the interviewer for things like resources, power, social status, or other related benefits. It is also important to note that people without influence, often best identify key leaders. That is because they recognize people that actually have influence, rather than someone telling you they are in a position of authority. I then asked questions to ensure I was talking to a person that actually resided in the village of interest. This allowed me to exclude potential bias and receive information from a local context and not from people who lived outside the village. However, I would not completely exclude their narrative, I would use my conversation with them to elicit information that locals may not want to tell me or had not considered being a resident of the subject village. Lastly, talking typically with male interviewees was critical. Women unfortunately have little influence and talking to them could even have negative repercussions in male dominated societies. Thus, when applying these interviewing tactics, it is important for the interviewer to know the operational environment and the certain cultural nuances that could upset socio-cultural dynamics.

Fourth, when choosing your interviewees, common interviewing best practices apply. This includes receiving informed consent, trying to isolate the interviewee from others overhearing the conversation, and establishing rapport. Ethical interviewers should take necessary means to receive at a minimum a verbal informed consent to be interviewed; which can prevent potential legal issues. When I found my potential interviewee, I would try to isolate them from locals in proximity to prevent my questions from being heard, protect their views, limit groupthink, and reduce fear of retribution for the interviewee talking to me. Also, establishing rapport with the interviewee, I found helped enable a more truthful and free-flow exchange of information. I did this through explaining why I was there, why I was interviewing them, and assuring them that I would protect the confidentiality of the information that they provided. Lastly, I found that a projection of empathy also helped the interviewee feel comfortable with me and the questions I asked.

Fifth, before actually diving into my more important questions, I would establish an interviewee baseline. I usually established a baseline through asking very non-specific demographic questions and observing their non-verbal cues during the building rapport tactic. I would ask some very simple questions and gauge their nonverbal cues such as posture, ease of answering questions, tone of voice, gestures, and other like indicators. These questions were often demographic in nature such as: How old are you? Are you married? How long have you lived in this village? What is your occupation? And so forth. These sorts of questions will likely be answered truthfully. During this line of questioning, I took careful notes of the abovementioned nonverbal indicators, making this my interviewee baseline. I then moved into harder contextual questions.  If the interviewee deviated from this baseline, I knew right away that the there might be something vital to probe further into. To illustrate, I would ask all my demographic questions, record their nonverbal cues, and move into a topic such a local insurgent activity. If the interviewee displayed varying level of discomfort, hesitance to respond, or a combination of like indicators, I knew that the interviewee might know more than they were telling. Also, the use of nonverbal indicators would cue me to change subject or terminate the interview if the interviewee displayed high levels of discomfort.                       

Sixth, after I established my baseline, I progressed into the interview itself. I would advance into my interview using one of two methods: Asking specific questions that are linked to some known-unknown identified by prior reporting or working from unknown-unknown situations (i.e., where I knew nothing about the village or the area). When working from a known-unknown situation, I would ask questions related to the context, and deviate as necessary with newly discovered information. When I knew nothing about the AOR or village (unknown-unknown), I used TCAPF questions to open the conversation and asked follow-on questions for each standardized TCAPF question. For example, I would ask the first TCAPF question: “Has the population of the village changed in the last 12 months? This may be followed-up by the questions, “Why do you think the population has increased/decreased;”  “How has this increase/decrease impacted the economy;” “Why do you think this about the economy,” and “Who in your khan [family/extended tribe] could I talk to about this that may have more information?”     

In both interviewing tactics, I would use direct and indirect questions and the use of hypotheticals in any interview to elicit information. I found that sometimes not asking a direct question such as, “how often do the insurgents visit your village?” may provide more accurate and unrestrained information. I would ask battery of semi-related questions to arrive at this answer. Questions may include: “When you travel between your village and the next, have you ever been stopped on the road at a checkpoint that was not ran by the government,” “When was the last time this happened,” “Do you have to pay any money to get to where you want to go,” “If you wanted to walk by foot at night between your village and the next, could you, if not why not,” “Does your neighbor feel that it is safe here, if not why,” and “Do you have police in your village and how often do they patrol in front of your home?” A combination of similar questions succeeded by probing and follow-on questions can provide detailed security-related answers and even other locally-nuanced information. Deferring from asking very direct questions at times helped the interviewee remain comfortable; without them feeling as they may receive retribution for answering very specific questions.

My experience also found that many sub-cultures and groups of people that are very socially connected prefer not to answer direct questions that may be related to how they perceive something individually. It is often better to ask questions that are oriented to communally held perceptions, not individually. In Afghanistan for example, I found it hard for individual Afghans to give me a definitive answer based on how they felt, but were willing to provide communally held beliefs. The interviewer needs to have the mindset they are speaking to the community, not the individual; which also makes it critical that the interviewer identifies people that reside in the community of interest (detailed in the third point above). 

Throughout the interview, I would observe the interviewee’s body language and nonverbal indicators to compare to their baseline (detailed in the fifth point above). As I asked more detail-oriented questions and observed deviation from their baseline, I would scope my questions to that topic, but may divert to indirect questions in order to keep the interviewee comfortable. If I observed extreme levels of discomfort, hesitation, or the general disinterest, I may digress or terminate the interview and move onto another person.    

Seventh, after the conclusion of your interview and moving onto someone else, close the interview. Closing the interview included being empathetic to any views expressed by the interviewee and thanking them for their time. I would also try to collect any contact information and the Military Grid Reference Systems (MGRS) location of where the interview took place. The MGRS location may provide supporting data for entry into MCS databases for future link analysis (if used).

The final tactic I employed was to rate the reliability of the information I received from the interview. This was often a subjective rating, but provided usefulness when trying to validate and coordinate future collection efforts. A “Social-Cultural Informant and Information Reliability Matrix” was developed to rate the reliability of the interviewee and the information content. Even though the matrix is used in a subjective context, information could be “confirmed” or rated “possibly true” with subsequent interviews providing supporting or validating evidence (see Appendix A for “Socio-Cultural Informant and Information Reliability”).     

 I will outline some of the techniques I employed to garner time-sensitive and actionable intelligence through the employment of tactical qualitative interviewing. Below are the tenets of qualitative primary collection and analysis described as above:

Know the Operational Environment and Socio-cultural Nuances

In order for a person engaged in intelligence collections through interviewing, they must be able to gain the trust of an interviewee to be able to elicit valuable information. They must:

• Give consideration to cultural dynamics and ethic fissures.
• Avoid personal bias and defer from letting personal judgment begin forming preconceived notions. When conducting research to find unknown-unknowns, this becomes critical. Bias and personal judgment may lead you to start forming a faulty hypothesis.

Furthermore, before conducting research, the collector should make the effort to study current events that may have impacted the area of interest.

• Review SIGACTS.
• Review all known facets of PMESII-PT variables affecting the area.
• Know what has been promised in the past to the area and what was delivered. [This may be an area of contention].

Have a Starting Point – For the Unknown-Unknowns and Known-Unknowns

The collector should have an idea of their purpose on the mission and what can be gained through tactical questioning.

The collector should be developing mental starting points or opening questions that may reveal intelligence information or data. Thus, they should be thinking of what requirements can they help answer (e.g., PIRs, IR, RFI; etc).

For instance, my starting point was through the use of a predefined question set called TCAPF (or, Tactical Conflict Assessment and Planning Framework) used in assessing development variables in the District Stability Framework (DSF). The questions listed in TCAPF were:

• Has the population of the village changed in the last 12 months?
• What is the greatest problem facing the village?
• Who do you believe can solve this problem?
• What should be done first to help the village?

Once you have a starting point to engage individuals with, you can follow-up or dig deeper into issues to gain valuable intelligence. Essentially, a starting point gives you a reference point to work from without asking: “So, tell me about the Taliban in your village?”

Try to Identify People of Little Influence, Urgency, or Power to Initially Question

When in your area of interest, you must first identify an individual or person to talk to. In the business, public sector, and academic circles, people with power, legitimacy, or influence are by default the “movers and shakers.” They can get things done…However, agendas, bias, or the need to influence may prevent the collector from gaining substantive information.

• This actually goes against various stakeholder identification theories such as the “Stakeholder Salience” model used in various forms of project management.
• This approach helps prevent interviewee levels of bias.
• Helps prevent political agendas being captured by the collectors as truths.

Try to Isolate Your Interviewee From a Group of People

• This tactic allows the interviewee to speak freely without fearing scrutiny or contempt from their peers.
• Helps the interviewee have higher levels of comfort, knowing that someone else from their community is not listening.
• Helps prevent groupthink.
• Prevents others or the group in proximity from hearing your logic/questions and forming answers based on group dynamics; rather than their personal thoughts.
• Take notice to your surroundings and the people around you. Once you have conducted enough interviews, it will become obvious who is from there and who is not. 

Establish Rapport With Your Interviewee While Establishing a Baseline:

The collector should have high emotional intelligence or be able to gauge body language, tone of voice; etc. The collector should begin the interview with asking questions that are easy and that build rapport with the interviewee. The collector should appear sincere. A good starting point is demographic questions. Other questions may include:

• Ask, what do you do for a living…?
• How long have you lived here? Example reply: Wow that is a long time, you must really like it here and know how things truly “work.”
• How old are you…?
• How is your family…?

As you are asking these questions, take notice to the interviewee’s tone of voice, body language, quickness to respond, eyes, hands, etc. The interviewee’s body language during the rapport build of your interview becomes your baseline.

The Interview

As you progress into your interview, start asking some of your predefined questions.

• As you ask them, elaborate on the questions to drill-in on details.
• Such as through the use of TCAPF questions, one of the questions: Who do you believe can solve this problem? I would then:
  • Wait for a response…and follow-up with questions such as:
  • Really why him? Do you trust him; etc?
• While asking follow-up questions, I would observe their body language and signs or hesitation or deception. I may then follow-up with the questions: Why? Who else do you trust? Who else can do the same things as this person?

Throughout the interview, I may inject hypotheticals such as: “We know this village might have a hard time with the Taliban, and we appreciate your partnership with us. When and if they come through here, do they make you go to their representative rather than XYZ person you just said (identified from TCAPF question)?

• When asking direct and sensitive questions such as this one, take notice to any signs of discomfort in the interviewee. Also, there may be signs of deception which may suggest actionable intelligence information, even if a direct or truthful answer is not given.
• Drill-in more on these questions and know when to back-off. The use of hypothetical’s can be a useful tool to elicit information and give meaning to unknown-unknowns.

Close the Interview

• Re-establish rapport to not wear-out your welcome.
  • Thank them.
  • Wish them prosperity; etc.
• Document your field notes and move on to other people or persons identified in your first interview, such as informal leadership. It is also important to ensure you document your impressions and observed body language. 

Rate the Reliability of the Information You Collected

• After the conducting your interviews, use an assessment tool that can map or match your collected information to probability or likeliness of being confirmed.
• Furthermore, you can increase the likeliness or probability of truth or validation as individual interviewees provide similar narratives or answers to like questions.
  • This can lead to actionable intelligence!
  • Preponderance of evidence can allow you to quantify your qualitative results.
  • My initial analysis and collection techniques were qualitatively structured.