CHAPTER FOUR

Decision Analysis Soft Skills

TERRY A. BRESNICK and GREGORY S. PARNELL

Leadership is communicating to people their worth and potential so clearly that they come to see it in themselves.

—Stephen Covey

The mind is like a parachute. It only works when it is open.

—Tony Robbins

4.1 Introduction

In Chapter 1, we identify the decision analysis soft skills, and in Chapter 2, we discuss decision-making challenges and note that the decision analyst must have the soft skills necessary to implement decision analysis as a socio-technical process. Soft skills are the nonquantitative, “personal,” “social,” or “interpersonal” skills that complement the quantitative hard skills of the mathematics of decision making methodologies. Soft skills, such as leadership and facilitation, include the personal skills to prepare for the analysis, such as strategic thinking and researching, and the ability to interact with and manage people, while hard skills involve technical tasks, mathematical modeling, and analysis techniques that we learn and perform individually. Our business and personal success relies on many traits, including honesty, integrity, dependability, friendliness, communications, optimism, etc. However, we focus on the soft skills that are directly related to a decision analysis project and help us interact with decision makers, stakeholders, and experts. For example, good communications is a general soft skill, interviewing an expert to assess a probability distribution is a decision analysis soft skill, and solving a decision tree for the best alternative with the probability information is a decision analysis hard skill. Every experienced decision analyst knows that soft skills are critical, but we have not found a comprehensive list in the decision analysis literature. Based on discussions with colleagues and our experience, we present and describe the following nine decision analysis soft skills: thinking strategically, leading teams, managing projects, researching, interviewing individuals, conducting surveys, facilitating groups, aggregating across experts, and communicating analysis results. We are not attempting here to provide a “cookbook” of steps that has been codified with rules that should be followed for every project, but rather, a collection of diverse ideas, general approaches, and specific techniques that the authors have found extremely useful over the years. The specific soft skills that decision analysts will use in a particular study will depend upon the nature of the decision analysis study, their background and training, their experience level, and their judgment about what is required for success in that study.

The remainder of the chapter is organized as follows. Sections 4.2–4.10 describe the nine decision analysis soft skills listed above. Section 4.11 provides a summary of the chapter.

4.2 Thinking Strategically

Strategic thinking is a personal skill that involves taking a long-term, systematic view of one’s environment, goals, and opportunities. Strategic thinking is recognized as an important attribute for our professional and personal lives (Dixit & Nalebuff, 1993). Liedtka (Liedtka, 1998) identifies five attributes of strategic thinking in practice: systems perspective, intent focused, thinking in time, hypothesis driven, and intelligent opportunism. We add a sixth attribute: speaking truth to power. We will describe each of these attributes using a decision analysis perspective.

• Systems perspective. “A strategic thinker has a mental model of the complete end-to-end system of value creation, his or her role within it, and an understanding of the competencies it contains” (Liedtka, 1998). Strategic thinkers understand the “big picture,” the decision environment, the strategic objectives of the organization, and the challenges that senior leaders face. A systems perspective includes a sound understanding of the strategic planning space of the organization, as well as the decision frame (Chapter 6) of the current study.
• Intent focused. Liedtka describes strategic intent as “the focus that allows individuals within an organization to marshal and leverage their energy, to focus attention, to resist distraction, and to concentrate for as long as it takes to achieve a goal.” A decision analyst should maintain a determined focus on the purpose of the decision analysis study and not be distracted by smaller issues and challenges along the way that might distract from the goal of their study.
• Thinking in time. “Strategy is not driven by future intent alone. It is the gap between today’s reality and intent for the future that is critical.” Strategic thinkers must be able to understand the past, envision the future, and act in the present. Decision analysis must understand the past and know which elements are important to the future. Decision analysts can use scenario planning to creatively think about future opportunities and challenges (Chapter 6). However, they must take actions during the study that position the organization for successful implementation once the decision is taken (Chapter 14).
• Hypothesis driven. Strategic thinking requires a balance of ensuring both creative and critical thinking. The creative thinking is especially important in the design of new alternatives (Chapter 8) and the development of new models (Chapters 9 and 11). Critical thinking requires listening to individuals with different views, collecting data to confirm or reject hypotheses, and using the best available data and sound mathematical techniques to evaluate alternatives.
• Intelligent opportunism. “The dilemma involved in using a well-articulated strategy to channel organizational efforts effectively and efficiently must always be balanced against the risks of losing sight of alternative strategies better suited to a changing environment.” The decision analyst must always be on the “lookout” for significant changes in the environment that could signal major new risks or game-changing opportunities with the potential to make dramatic improvements in value through improved strategies.
• Speaking truth to power. Being a strategic thinker is easy when the decision analyst is presenting exciting new ideas with great potential for value creation. Unfortunately, decision analysts sometimes are the bearers of bad news. Decision analysts must have the integrity and courage to present the unvarnished truth to senior leaders even if there are undesirable potential professional and personal consequences.

4.3 Leading Decision Analysis Teams

Decision analysts must be effective decision analysis team leaders. Leading teams includes identifying the correct decision issues; developing common team goals; motivating individuals to achieve team goals; guiding the team during the study; and helping the client organization to achieve the most value from the study. J. Richard Hackman (Hackman, 2002) describes five conditions necessary for successful teamwork: The team must be a real team, rather than a team in name only; it has compelling direction for its work; it has an enabling structure that facilitates teamwork; it operates within a supportive organizational context; and it has expert teamwork coaching. We describe each of these in the context of a decision analysis.

• Form a team. Assigning the name “team” to a new decision analysis group does not make them act like a real team or perform effectively as a team. The decision analysis team leader must motivate the team to perform as a team. An important first step is establishing team objectives for the study. The team that the decision analysts may have to develop may be the facilitators and experts supporting a project, or it may even be the internal decision board itself (see Chapter 13).
• Provide compelling direction. Groups that have a shared understanding of the importance of their decision analysis project to the organization are more likely to perform effectively. The team leader must develop and communicate this compelling direction.
• Delegate responsibility and authority. The leader must be careful not to try to do everything himself/herself. Key tasks should be delegated to team members along with the authority to make things happen and the responsibility and accountability for results.
• Enable teamwork structure. The leader must design and implement a team structure that facilitates team work. This structure should include an understanding of the responsibilities and authorities of each team member.
• Obtain organizational support. The team leader must insure that the key leaders and stakeholders support the decision analysis project and provide the necessary resources for success. A critical issue is the access to decision makers, stakeholders, and experts. The decision process must be designed to provide access to senior decision makers (see Chapter 5).
• Obtain expert coaching. Taking a college course in decision analysis or performing technical work on a decision study does not prepare one to be a decision analysis team leader. Leadership skills are developing by doing. Studying good leadership traits and examples can be very helpful. Having a coach or mentor can also be very helpful.

4.4 Managing Decision Analysis Projects

There is a difference between leadership and management—“management is doing things right; leadership is doing the right things” (Covey, 1989). In addition to leading teams as described in Section 4.3, leaders must also manage the decision analysis project. Project managers are concerned with effectiveness and efficiency. Managing decision analysis projects includes developing analysis study plans; identifying and scheduling activities; monitoring progress; managing the completion of tasks; and transitioning project activities to individuals and teams with authority and responsibility for decision implementation.

• Develop a project plan. For large projects with many participants, a project plan is essential to identify the responsibilities of all contributors. Depending on the scope and duration of the effort, the plan can range in size from a few pages to a large document.
• Identify and schedule activities. Decision analysis is not a standard set of “cookie-cutter” techniques that are used identically every time. The decision analysis techniques and process must be developed uniquely for each decision analysis project. A common schedule is required to assign and monitor the activities.
• Monitor progress. Every project and every team is different. The project manager needs to monitor progress and, usually, needs to revise activities or reallocate resources to keep on schedule and within cost.
• Manage task completion. The decision analysis study participants are key individuals in the organization. Once their tasks are complete, the project manager needs to reassign team members to other tasks or return them to their “day job” or another project.
• Transition to decision implementers. Many times the decision analysis team does the study up to the decision and then transitions project activities to individuals and teams with authority and responsibility for decision implementation. The greater the participation of the decision implementers in the study, the more likely they are to understand the decision, and effectively execute the decision implementation (see Chapter 14).

4.5 Researching

Researching is an important personal skill. As we discuss the important role of decision analysts in interacting with decision makers, stakeholders, and subject matter experts (SMEs), it is important to note that many times, the analyst must perform research prior to these meetings to better understand the problem domain, identify the key issues, identify objectives, and develop the key questions to ask. The best researchers look for a variety of credible sources from both the problem domain and the decision analysis literature that offer conflicting views on the problem. The amount of research depends on the decision analyst’s prior understanding of the problem domain, knowledge of key terminology, and amount of domain knowledge expected of the decision analyst in the decision process.

Another important role of research is to find decision analysis techniques that are appropriate for the problem being addressed. For example, we may decide to use scenario planning or Monte Carlo simulation and decide that we need to review the best practices in the literature.

4.6 Interviewing Individuals

A frequently used approach for knowledge elicitation is the interview process. Interviews are especially appropriate for senior leaders who do not have the time to attend a longer focus group or the interest to complete a survey. Interviews can be highly effective, but there are many potential pitfalls. The time of high-level decision makers is valuable, so it is essential to keep the interview moving. The interviewer must probe for thorough responses to open-ended questions rather than simple yes/no or few-word answers and should use research and subject matter expertise to ask challenging questions. The questions asked must be clear, and the interviewer must quickly provide clarification as needed. The interviewer must not impose his or her emotions and feelings on the participant. While an occasional war story is okay to stimulate thought, the interviewer must avoid dominating the discussion. The interviewer should be aware of his or her own personal gestures and body language—participants pick up on them easily. Finally, and most importantly, the interview team must keep good notes.

There following are a variety of formats that can be used to frame good questions:

• Fact-finding questions. These are designed to gather verifiable data or facts.
• Feeling-finding questions. Theses are designed to gather subjective information that gets at feelings, opinions, beliefs, and so on.
• “Tell me more” questions. These are designed to follow up or to expand upon what participants are saying.
• Best/least questions. These are designed to test limits of wants and needs; questions often take the form of “what would be best in this circumstance?” or “what is the most critical factor?”
• Third-party questions. Theses are designed to uncover sensitive information in an indirect manner. Often, questions are asked about what other persons would say or do or think about a specific topic.
• “Ruler for a day” questions. These are designed to remove constraints that inhibit or limit responses; they may take the form of “what things would you change if you were in charge?”

Regardless of the questioning technique, it is essential to:

• Obtain the facts. When asking factual questions, ascertain the level of confidence that the participant has in his or her ability to answer.
• Distinguish among facts, feelings, and emotions. With open-ended questions in particular, find out what is known, what is assumed, and what is opinion.
• Frame questions carefully. How an open-ended question is framed and probed can greatly affect the answer. Develop a script and stick to the script unless a potentially interesting issue is raised that has not been thought of before.
• Identify biases from the managers and subject matter experts that could skew judgmental assessments.

Kayser, in Mining Group Gold (Kayser, 2011), indicates that interviewees can be classified by the way they react to interview questions. He describes different types of interviewees:

• Expressers (open words, open emotions). They easily represent their words and feelings and are quick to get things off their chests. The interviewer must pay careful attention to the specific words behind the emotions that the participant chooses.
• Directors (open words, guarded emotions). They tend to use strong positive words and speak quickly and confidently, and have little patience for time wasting. The interviewer must keep things moving; avoid extended side discussions.
• Reactors (guarded words, open emotions). They get easily upset and their emotions are clear. They make heavy use of body language and gestures, their responses are indirect and obscure, and their reasoning is sometimes hard to clarify. The interviewer must be sensitive to emotions and must not let the participant think that his or her emotions are being ignored.
• Stoics (guarded words, guarded emotions). They tend to keep facts and emotions to themselves. The interviewer must be sensitive to body language which may give clues to when to probe further; do not let participants’ discomfort fester and explode.

Since interviews take time, it is important to get the best information possible. The following are best practices from the authors’ personal experiences for each phase of the interview process: planning, scheduling, conducting, documenting, and analyzing interviews.

4.6.1 BEFORE THE INTERVIEW

For interviews with senior leaders and key stakeholder representatives, it is important to prepare a questionnaire to guide the interview discussion. The following are best practices for interview preparation:

• Develop as broad and diverse a list of interviewees as possible. Identify one or more interviewee for each stakeholder group. Review the interview list with the decision maker(s) to ensure that all key stakeholder groups are represented on the list of potential interviewees.
• Begin the questionnaire with a short explanatory statement that describes the reason for the interview, the purpose of the decision analysis, and the stakeholders being interviewed.
• Begin the interview with an “unfreezing question” that encourages the interviewee to think “outside the box” about the future and how that will impact the key elements of the decision.
• Tailor the questionnaire to each interviewee. Make the questions as simple as possible.
• Do not use leading questions that imply the answer is known in advance and that guide the interviewee to agree with the preconceived answer.
• Do not ask a senior leader a detailed question that can be obtained by research.
• End the questionnaire with a closing question, for example, “Is there any other key issue that we have not discussed?”
• Arrange to have an experienced interviewer and a recorder for each interview.
• Decide if the interviews will be for attribution or not for attribution.

4.6.2 SCHEDULE/RESCHEDULE THE INTERVIEW

Interviews with senior leaders require scheduling, and frequently, rescheduling. The more senior the leader, the more likely schedule will be a challenge. The following are best practices for interview scheduling:

• It is usually best to have the interviews scheduled by team members who are representatives of the decision makers and stakeholders since they have better access.
• Depending on the importance of the decision and the difficulty of scheduling, we usually request 30–60 minutes for the interview.
• It is usually best to conduct interviews individually to obtain each interviewee’s ideas on the topic at hand. Additional attendees change the interview dynamics. The senior leader may be reluctant to express ideas in front of a large audience or may defer to staffers to let them participate.
• While some facilitators prefer to provide the interview questionnaire to the interviewees in advance, there is a danger that the recipient will have his or her staff fill it out and we will not get the recipient’s perspective. This can result in the interviewee reading “staff answers.” If we want staff answers, we should consider a focus group or a staff interview.
• The interviews can be done in person or over the phone. In-person interviews are the most effective since interaction is easier; however, sometimes they are not possible and the only practical choice is a phone or video interview.

4.6.3 DURING THE INTERVIEW

The interview team’s execution of the interview creates an important first impression with the senior leader about the decision team. Some thoughts for conducting interviews:

• The best number of people to conduct the interview is one interviewer and one note taker. An alternative to the note taker is a recorder. Some interviewees may be reluctant to be recorded. If considering using a tape recorder, request permission first.
• Conduct the interview as a conversation with the interviewee. Use the interview questionnaire as a guideline. Take the questions in the order the interviewee wants to discuss them.
• Make the interview interesting to the interviewee. Be flexible, follow-up on an interesting observation even if it was not on the questionnaire. Many times, one interviewee will identify an important issue that no one else has mentioned.
• Ask clear, open-ended, questions that require the interviewee to think and respond. Avoid complex questions that might confuse the interviewee.
• Respect the interviewee’s time. Stay within the interview time limit unless the interviewee wants to extend the interview period. When the interviewee’s body language signals that they have finished the interview (e.g., fold up paper and look at their watch), go quickly to a closing question and end the interview.

4.6.4 AFTER THE INTERVIEW

Documentation of the interview is important to capture the key decision issues since they provide important data for all of the steps that follow. Best practices for documenting the interviews include the following:

• As soon as possible after the interview, type the interview notes. The questions and the answers should be aligned to provide proper context for the answers. The documentation should be consistent with the decision to use the notes with or without attribution.
• The interviewer should review the typed notes and make revisions as required.
• Once the interview notes are complete, the interviews should be provided to the decision team.
• It is good practice to follow-up the session with an individual thank you note (or email)

Freeman Marvin et al., in their “Soft Skills Workshop,” suggest additional best practices for interviewing as shown in the boxed text (Klimack et al., 2011):

4.7 Conducting Surveys

Surveys are a useful technique for collecting information from a large group of individuals in different locations and are most appropriate for junior to mid-level stakeholders. Surveys can be used to gather qualitative and quantitative data on the decision objectives, uncertainties, and alternatives. A great deal of research exists on techniques and best practices for designing effective surveys. Decision analysts can distribute and collect survey data via mail, email, or the Internet. This section provides an overview of survey design and methods for conducting surveys. Surveys require detailed planning to accomplish the team’s goals (The Survey System—Survey design, n.d.). These steps can be followed to plan, execute, and analyze surveys:

• Establish the goals of the survey.
• Determine the sample of stakeholders that will be targeted with the survey.
• Determine how the survey data will be distributed and collected.
• Develop the survey questions.
• Test the survey.
• Distribute the survey and collect data.
• Analyze the survey data.

4.7.1 PREPARING AN EFFECTIVE SURVEY: DETERMINE THE GOALS, SURVEY RESPONDENTS, AND MEANS OF DISTRIBUTING AND COLLECTING SURVEY DATA

The decision team needs to clearly articulate the goals of the survey and the target sample of stakeholders whom they want to answer the survey. Often, surveys for decision problems will be used to collect potential decision objectives and value measures. If the team plans to collect and analyze data from questions with standard answer scales (e.g., “Yes/No” or multiple choice answer scales), it is important to determine the appropriate sample size needed to draw valid statistical conclusions from the survey data. Sample size calculations are described in basic statistics books, and online tools are available to do these calculations (The Survey System—Sample size calculator, n.d.). The team needs to work with the decision maker(s) to determine the appropriate stakeholders to survey.

The method for implementing a survey should be selected before the survey is designed. Popular methods are mail, electronic mail, and web surveys. Table 4.1 provides a listing of some of the advantages and disadvantages of these survey methods.

TABLE 4.1 Advantages and Disadvantages of Surveys

The ability to collect survey responses in a database when using a web survey instrument can be beneficial if subsequent analysis is required. Several online programs now exist to help teams design web surveys, collect responses, and analyze the results. Some popular programs include SurveyMonkey.com (SurveyMonkey), InsitefulSurveys.com (Insiteful Surveys), and the SurveySystem.com (The Survey System).

4.7.2 EXECUTING A SURVEY INSTRUMENT: DEVELOPING THE SURVEY QUESTIONS, TESTING, AND DISTRIBUTING THE SURVEY

Surveys should be designed to obtain the information to meet the goals of the survey. To maximize response, the survey should be short with clearly worded questions that are not ambiguous from the respondent’s perspective. When constructing questions, care must be taken not to overly constrain the answers. Start the survey with an overview of the purpose of the survey and the goals the decision analyst hopes to achieve from the information provided by the respondents. Here are some general principles that can be followed in developing effective survey questions (Fowler, 1995):

• Ask survey respondents about their first-hand experiences so they can provide informed answers.
• Ask only one question at a time.
• In wording questions, make sure that respondents understand the question. If the question includes terms that could be interpreted differently by respondents, provide a list of definitions to clarify any possible ambiguities. This list of definitions should precede the questions.
• Articulate to respondents the kind of acceptable answers to a question. For objective questions, answer scales can be setup as multiple choice answers from a rating scale or level-of-agreement scale. For open-ended text response questions, the question should be worded so respondents provide information germane to the question. Close the survey with a statement allowing respondents to provide any additional information pertinent to the goals of the survey.
• Format the survey so that it is easy for respondents to read the questions, follow instructions, and provide their answers. For example, answer scales should follow a similar pattern in terms of the order in which they are presented (e.g., the least desirable answer is the first choice ascending to the most desirable answer).
• Orient the respondents to the survey in a consistent way. This can be accomplished with a set of instructions that describe the goals of the survey, the method for completing their responses, and the means for submitting the completed survey.
• Once the survey questions are written, test the survey instrument with a few individuals outside the team. Ask them to complete the survey using the same medium that respondents will use (e.g., by e-mail, mail, or on the web). Ask for input from the test sample regarding the instructions and wording of the questions and answer scales. If a web survey is used, test the method for collecting responses. Use the input from the test sample to improve the survey.
• Once improvements are made, distribute the survey to respondents using the method chosen. Develop a plan for monitoring the response rate and establish when reminders will be sent to respondents who have not completed the survey. The team should also have a standard way to thank respondents for their time and efforts, for example, a thank you note or e-mail.

4.8 Facilitating Groups

Group facilitation is a process in which a person whose selection is acceptable to all members of the group, who is substantively neutral, and who has no substantive decision-making authority, diagnoses and intervenes to help a group improve how it identifies and solves problems and makes decisions, to increase the group’s effectiveness.

A facilitator is a person who helps a group free itself from internal obstacles or difficulties so that it may more efficiently and effectively pursue the achievement of desired outcomes

—Roger Schwarz (Schwarz, 2002)

4.8.1 FACILITATION BASICS

Facilitating groups is a critical decision analysis soft skill. When planning to facilitate a meeting, one size does not fit all. Facilitation can run the gamut from totally noninterventional, “touchy-feely” group processes, to more interventional, structured, and highly quantitative group processes and modeling. Facilitation sessions may be for Groups that get together on a one-time or infrequent basis, have no sense of group identity, and may have little reason to seek or expect consensus. Or it may be for Teams that get together regularly, have a common sense of group identity and purpose, and often seek to achieve consensus.

One unique approach to facilitation is decision conferencing (Phillips, 2007). A decision conference is typically a 2- to 3-day session supported by a facilitation team consisting of a lead facilitator, a model builder, and a scribe (O’Connor, 1984). It is most useful when a decision must be made in a very short period of time, when the stakes are high, when little “hard data” is available, and when most of the subjective information needed to make the decision resides in a small group of experts. The lead facilitator moderates and controls the sessions, elicits information, asks questions, channels responses, and constructs analytical models in response to group input. The model builder operates a computer to build in real time the decision models developed by the group and displays them on a projection screen. The scribe acts as a conference recorder, documenting key decisions and providing a complete audit trail of rationale for the session. The decision conferencing approach to facilitation is described in detail in Appendix C.

The role of the facilitator is challenging and requires multitasking. The facilitator must ask questions carefully using clear vocabulary, clarify and elaborate, seek feedback, and be consistent. He or she must be able to listen to the responses carefully and use listening techniques, such as paraphrasing, repeating word for word, and so on. The facilitator must be cognizant of body language and expressions. The facilitator must be aware of dysfunctional group behaviors, such as late arrivers, silence, chronic objectors to the process, dominators, side conversations, rambling answers, off-the-wall remarks, and so on, and must be prepared to deal with them as they arise. Techniques for dealing with such behavior include (Schwarz, 2002; Kayser, 2011):

• Be friendly, but firm in confronting the behavior.
• Focus on the behavior, not the individual.
• Encourage group members to share responsibility in handling difficult members.
• Recognize acceptable behavior and highlight the impact of disruptive behaviors.
• Privately suggest more functional behaviors with disrupters.
• Avoid sustained one-on-one arguments or exchanges.

Finally, the facilitator must provide feedback to the group in real time.

Facilitator responsibilities include (Kayser, 2011):

• Stimulating the conversation and balancing group participation.
• Understanding (not evaluating) participants’ feelings.
• Refocusing the group based upon desired outcomes.
• Staying neutral on content.
• Staying focused on the situation, not the group behaviors.

On occasion, conflict occurs and must be managed by the facilitator. Some friction can be good if it stimulates critical thinking, enhances creativity, or minimizes stagnation. However, if the conflict becomes disruptive, the facilitator can manage it by clarifying opposing points of view, defining areas of agreement and disagreement, using effective problem-solving techniques, and in some cases, trying to resolve differences. Specific techniques include (Kayser, 2011):

• Collaboration. Looking for “win–win” solutions; extensive effort is typically spent on generating innovative solutions, on probing deeply into the problem, and on brainstorming.
• Avoidance. Often results in a “no winner, no loser,” no solution, situation; participants can choose to withdraw, give up, or just leave the process.
• Accommodation. Often results in someone “winning” and someone “losing”; typically requires someone to give in, yielding to others’ positions or smoothing over concerns with additional concessions.
• Compromise. The goal is “no winner, no loser”; tradeoffs are the norm as is looking for middle ground or the lowest common denominator. Solutions reached are rarely the best solution.
• Domination. Someone “wins” someone “loses.” This is often achieved by overwhelming those with opposing positions, by unyielding defense of one’s own position until others become weary or convinced, or by pulling rank. Frequently, arguments get heated and emotional.

As experienced facilitators, we strive for collaboration.

4.8.2 GROUP PROCESSES

Many of the inputs used to develop analytical models are elicited using well-documented group processes. Various methods are used to take advantage of the positive aspects of group dynamics and to minimize the negative effects as described later in the chapter.

4.8.2.1 Stages of Group Development. 

All collaborative groups evolve over time as members learn to trust and work with each other, whether the group meets for a one-day decision conference or a year-long project. Bruce Tuckman identified four stages of group development—a model that many facilitators still use today to understand and guide group process (Tuckman, 1965).

• Forming. When a new group forms, its members are consciously or unconsciously searching for trust, meaning, and a sense of belonging with the group. Facilitators must be aware of these needs and insure that each session begins with an activity or exercise that allows everyone to participate and get engaged in the work at hand.
• Storming. Inevitably, the problem-solving process begins to heat up under the pressures of conflicting perspectives and information. People jockey for influence. Patient and impatient people clash. Trust is tested, and confusion around goals and roles begins to surface. If there are deadlines and constraints, this stage can be tense. Facilitation can help a group move through this stage efficiently by turning the focus of the group members’ anger and frustration on the emerging model of the problem and away from each other.
• Norming. As members get to know each other better and get a grip on the problem, they begin to reconcile differences and come to agreement on decision-making processes, resources, timing, and standards. Norms are the formal and informal rules that make up the operating system of productive work. Facilitators can suggest at this stage a wide range of group processes, decision aiding tools, and analytical techniques that will help the group move forward.
• Performing. The final stage of team development involves the coming together of all the experience and skills of the group members to get results for each other and the organization. If the decision conference or working session is to be repeated for another aspect of the problem, the group members will continue to build upon their newly acquired collaboration and problem solving skills.

4.8.2.2 Planning. 

Meeting management is all about being organized, and collaboration needs careful planning as well. As a best practice, we highly recommend conducting a dry run with a small group of colleagues to insure that terms are understandable and questions are interpreted as intended.

Kayser suggests that every facilitator use a cornerstone document on which all collaborative sessions rest—The PDORA document. PDORA stands for Purpose, Desired Outcomes, Roles, and Agenda (Kayser, 2011). For the facilitator, the PDORA document will be an effective tool in keeping sessions on course and preventing breakdowns. In a similar vein, Marvin uses the mnemonic GRASP to help facilitators plan and prepare for a collaborative working session:

• Goals and outcomes
• Room and logistics
• Agenda and time available
• Support team, tools, and techniques
• Participants and observers.

4.8.2.3 Pulsing. 

Decision making and problem solving require two types of thinking: divergent thinking and convergent thinking. Groups that do not spend enough time on divergent thinking risk solving the wrong problem or falling into groupthink. Groups that spend all of their time on divergent thinking never come to conclusion and become nothing more than a debating society (Marvin, 2006).

The ideal group process spends an adequate amount of time in divergent thinking—gathering information, uncovering goals and objectives, and generating creative alternatives—and then switches to convergent thinking for evaluating, prioritizing, and reaching consensus. This creates a repeatable natural cycle or “pulse” by the group as it works through pieces of the problem. The facilitator helps maintain group rhythms of divergent and convergent thinking.

There is ample evidence that groups often fail to explore adequately the “problem space” or the “solution space.” When a group convenes to make a decision or solve a problem, it is useful to follow the “tunnel—funnel” approach diagramed in Figure 4.1. This process encourages divergent thinking as a prelude to convergent thinking.

FIGURE 4.1 Divergent and convergent thinking.

4.8.2.4 Pacing. 

The facilitator helps manage the group task and time available, spurring on the group when progress is too slow, reining back when the group is jumping to conclusions, and guiding the group toward its goals. This role is critical to success because groups usually try to accomplish a lot in a limited amount of time. Pacing is all about channeling the energy of the group.

There are three situations when the facilitator can use good group process techniques to channel group energy. These techniques are called Slow, Flow, and Go (Marvin, 2006).

4.8.3 FOCUS GROUPS

Facilitation is also required for focus groups. The term “focus group” is attributed to psychologist Ernest Dichter (Focus Groups, 2011). Focus groups are typically small groups of 6–12 people, and are primarily used to solicit qualitative opinions and attitudes about how satisfied users or customers of a service or product have been. They are also used to gather data on new or proposed products, services, or ideas. They can also be useful for identifying decision objectives, eliciting evaluation criteria, or identifying key uncertainties. They are typically shorter and more focused in scope than decision conferences or other facilitated working sessions. While interviews typically generate a one-way flow of information, focus groups create information through a discussion and interaction between the group members. As a general rule, focus groups should comprise between 6 and 12 individuals. Too few may lead to too narrow a perspective; too many will lead to some individuals not being able to provide meaningful input. As with interviews, the focus group facilitation team needs to devote time to prepare for, plan the execution of, and analyze data from focus groups.

4.8.3.1 Preparing for the Focus Group Session. 

Developing the goals and objectives of the focus group session is critical to success. A few best practices for preparing for a focus group session:

• Develop a clear statement of the purpose of the focus group and what it hopes to achieve from the session.
• Develop a profile of the type of participant who should be part of the session and communicate that to the project client. Select a participant pool with the project client.
• Select and prepare a moderator who can facilitate a discussion without imposing his or her own biases on the group.
• Schedule a time and location during which this group can provide 60–90 minutes of uninterrupted discussion.
• Develop a set of questions that are open-ended and which will generate discussion. Do not use “Yes/No” questions that will yield little discussion. The most important information may come out of discussion resulting from a question posed to the group.
• Conduct a “dry run” with colleagues to insure questions convey what is intended.

4.8.3.2 Executing the Focus Group Session. 

The most important components of executing the session are the moderator and the recording plan. Here are some thoughts for the execution of a focus group session (McNamara, n.d.):

• The moderator should review the session goals and objectives, provide an agenda and the plan for recording the session.
• Ask questions and allow participants a few minutes to discuss their ideas. The moderator should ensure equal participation to prevent a few individuals from dominating the group.
• A good technology solution for facilitating focus groups is the Group Systems software (Group Systems Homepage, n.d.). This technology facilitates groups in brainstorming activities and generating ideas. It helps mitigate the impacts from individuals who tend to dominate discussions because participants type their ideas on a computer in response to questions generated by the moderator. It also helps record the data.
• Do a video and audio recording of the session if possible. If not, use multiple note takers.
• On closing, tell participants that they will receive a record of the decision objectives from the session.
• Follow up the session with an individual thank-you note for each participant.

4.9 Aggregating across Experts

When the decision analyst deals with multiple experts, it is often necessary to aggregate across experts to come up with a single position or assessment. Some of the more common techniques include:

• Working toward consensus through discussion (i.e., all can live with it)—the decision analyst/facilitator assists the group of experts in conducting a structured conversation designed to enable open exchange of information with the goal of informed consensus. Opportunities are made available to explore different perspectives, faulty assumptions, and hidden agendas. The ultimate goal is to reach consensus on a topic or assessment. The decision analyst must also recognize circumstances when consensus is just not achievable. There may be some differing perspectives that are parochial or political in nature, there may be just no middle ground positions, and achieving consensus may not be the goal of all participants. When this happens, it is useful to forego aggregating across all participants and instead, reporting a minority position that can be used as part of a sensitivity analysis to determine if it impacts the decision.
• Voting across participants, majority rules. This can be a quick way to determine how close or how far apart participants are. Whether it is voting on scores or weights or probabilities, each participant gets a vote, results are openly displayed (often using spreadsheets), opportunities are provided to understand and discuss differences, and often, revotes are allowed. Typically, a “majority rules” approach is used to reconcile the differences.
• Mathematical aggregation (such as averaging, using copulas to develop joint distributions, etc.). Mathematical approaches can be quick, relatively easy, and relatively common in aggregating across experts, but unfortunately, they are often applied inappropriately. The most common approach is averaging across voters. While this does represent the mathematical mean of the voters, much information can be lost, particularly when votes are highly bimodal. In such cases, the average does not represent what the group is saying at all. It is far better to look at the data to see if outliers are skewing results, and then try to explore reasons for outliers. If after discussion, outliers remain that skew results, it is better to initially eliminate the outliers, recalculate averages, and do sensitivity analysis on whether or not eliminating the outliers has an effect on the decision. The specific averaging approach should be a function of the types of scales being used.
• Combined approaches. Two or more of the above approaches can be used together to eliminate the shortfalls of a single approach. For example, the group can use the voting approach for a first pass to stimulate discussion, but then the facilitator, together with the group, can examine the data visually and determine which statistic (mean, mode, and median) best represents the holistic center of mass of the group.

Regardless of the technique used for aggregation, it is important to decide up front what the rule for achieving consensus will be; the group must agree on how the decision will be made.

4.10 Communicating Analysis Insights

The communication of analysis insights to decision makers and key stakeholders is such a critical decision analysis soft skill that we have devoted Chapter 13 to the topic.

4.11 Summary

In this chapter, we review the soft skills necessary for interacting with decision makers, stakeholders, and experts. The decision analyst should be a strategic thinker, a team leader, and a project manager. We describe the soft skill attributes for each of these three roles. Researching is the starting point for understanding the problem domain and understanding the key components of the decision. A frequently used approach for knowledge elicitation is interviewing. Interviews can be highly effective in capturing the preferences of senior decision makers and their staff, but can carry with it many pitfalls. In addition to research and interviews, surveys, when done properly, can provide a wealth of information by collecting information from large groups of individuals in different locations. In the chapter, we present best practices for all stages of the interview process and for preparing and conducting surveys.

In many cases, it is necessary to go beyond the techniques described above to bring together in one location a wide range of people, including decision makers, stakeholders, and subject matter experts. Facilitation of such groups is a critical soft skill. Whether in formal decision conferences or in less formal workshops, the role of the facilitator is challenging and requires an extensive set of skills. Facilitators must enable open exchange of information, encourage consensus building, deal with conflict, and deliver collaborative results. When we need to bring together small groups to gather information on well-focused topics, we can use focus groups to encourage discussion and interaction among group members. Familiarity with group processes, techniques for aggregating across experts, decision-making barriers, and cognitive biases found in Chapter 2 provides essential knowledge for the successful facilitator. The final soft skill is communicating analysis insights to decision makers and stakeholders, which we describe in Chapter 13.

KEY TERMS

REFERENCES

Covey, S. (1989). The 7 Habits of Highly Effective People. New York: Simon & Schuster, Inc.

Dixit, A.K. & Nalebuff, B. (1993). Thinking Strategically: The Competitive Edge in Business, Politics, and Everyday Life. New York: Norton.

Focus groups (2011). Retrieved from Wikipedia.

Fowler, F. (1995). Improving Survey Questions—Design and Evaluation. Applied Social Reserach Methods Series. Thousand Oaks, CA: Sage Publications.

Group Systems Homepage (n.d.). Collaboration for leaders. http://www.groupsystems.com/, accessed October 17, 2011.

Hackman, J.R. (2002). Leading Teams: Setting the Stage for Great Performances. Boston: Harvard Business School Press.

Kayser, T. (2011). Mining Group Gold, 3rd ed. New York: McGraw-Hill.

Klimack, W., Marvin, F., Wicker, P., Buckshaw, D., Kloeber, J., Leonardi, D., et al. (2011). Soft Skills Workshop: Real World Skills for Decision Analysts and OR/MS Professionals. Vienna, VA.

Liedtka, J.M. (1998). Linking strategic thinking with strategic planning. Strategy and Leadership, 26(4), 30–35.

Marvin, F. (2006). The Four Faces of Facilitation: A Guide for Facilitation of Problem-Solving and Decision-Making Groups.

McNamara, C. (n.d.). Basics of conducting focus groups. http://managementhelp.org/businessresearch/focus-groups.htm, accessed October 17, 2011.

Moore, C.M. (1987). Group Techniques for Idea Building. Thousand Oaks, CA: Sage Publications.

O’Connor, M. (1984). Methodology for corporate crisis decision making. In S. Andriole (ed.), Corporate Crisis Management. New York: Petrocelli Books.

Osborne, A. (1953). Applied Imagination. New York: Charles Scribner & Sons, Inc.

Phillips, L. (2007). Decision conferencing. In R. Miles, D. Von Winterfelt, & W.E. Edwards (eds.), Advances in Decision Analysis, pp. 375–398. New York: Cambridge University Press.

Schwarz, R. (2002). The Skilled Facilitator. San Francisco: Jossey-Bass.

The Survey System—Sample size calculator (n.d.). http://www.surveysystem.com, accessed October 17, 2011.

The Survey System—Survey design (n.d.). http://www.surveysystem.com/sdesign.htm, accessed October 17, 2011.

Tuckman, B. (1965). Developmental sequence in small groups. Psychological Bulletin, 63(6), 384–399.