THE ROBUST TEAM: A QUALITY ENGINEERING APPROACH

In general, the traditional approach to evaluating the performance of groups in process has been twofold. The first has been to use a developmental model that provides a summary of the different phases or stages in the life cycle of a group. A popular example of this approach is the forming, storming , norming, performing model of group development. Each phase corresponds to a stage in the group life cycle ” review Volume I, Part II of this series.

The second model has emphasized structural patterns of a group or team. These may be construed in terms of gender, experience, length of service, or positional roles (leader, secretary, or assistant, for example). Using the structural approach, the team can also be analyzed in terms of process; the "peacemaker," the "aggressor," the "blocker," or the "help-seeker," for example, or Resource Investigator, Coordinator, and so on.

Both these models have proven to be useful when trying to describe some aspects of group dynamics, and it may be possible to identify colleagues who fulfill some of these roles or identify teams that have passed through these different stages of development. Unfortunately, such a restricted approach to monitoring team process does not provide any feedback as to whether the team is producing predictable results, nor does it identify problems or opportunities for improvement ” especially breakthrough opportunities. Specifically, no opportunity exists to determine whether team process is "in control" (capable) or whether the group is "out of control" ( chaotic and falling far short of what it could achieve). Some of these issues were addressed in Volumes I and II of this series, and perhaps the reader may want to review them at this time.

A further shortcoming in non-systems approaches to team building concerns team process improvement. As long as the team is operating within "acceptable" parameters, no opportunity or drive to improve or maximize the performance of the team exists. Furthermore, the team usually does not have the ability or training to self-regulate and, through self-regulation, to begin to change and adapt to the continual change taking place in the workplace. These and other considerations suggest that a systems approach to team building may have considerable advantages.

The robust team involves an examination of teams as systems in conjunction with more detailed parallels between a team systems approach and the model put forward by Taguchi as part of his quality engineering methodology (see Volume V of this series). Using this viewpoint, a system is considered as a means by which a user 's intention is transformed into a perceived result. Therefore, if teams are considered in terms of how successfully they transfer energy when they function, it should be evident that there will be parallels between their functioning and the functioning of an engineered system ” as in the P diagram for example. After all, in many ways, a team shares similar features to the manufacturing process of a particular product. Specifications are drawn up (objectives, time scales , etc. are established); the production machinery is put in place (team members are selected); the production process is designed and implemented (teams meet, establish norms, set agendas , and engage in problem solving, decision making, and planning activities, etc.) and the system is regulated by performance criteria (by the individual members ' expectations, assessments, performance appraisals , etc.).

In manufacturing, it is important not to separate the performance of the component from its interaction with other components and its integration into large subsystems of the whole process or product. In teams, it is important not to separate the performance of the individual from his/her relationships to other team members, their interactions, and their membership in sub-teams and the team as a whole ” rather it is of paramount importance to view them as a team system.

TEAM SYSTEMS

Many social psychologists only consider a collection of people to be a group if their activities relate to one another in a systematic fashion. However, it is easier to define a group as a collection of individuals. The word "team," however, as mentioned in Volume I, Part II, is reserved for those groups that constitute a system whose parts interrelate and whose members share a common goal. Some groups can easily be viewed according to this criterion. A soccer club, its manager, and its players constitute a set of parts necessary to the functioning of the whole ” the common aim being to win soccer games . However, when does a newly established team become a good or effective team? To see the answer to this question let us examine the team from a systems approach.

Input

A team has an input or signal. The input is the information, energy, resources, etc., that enter into the system and are transformed through its structures and processes. A broad spectrum of inputs into the system can exist and, depending on the perspective one chooses to take, the boundaries that are drawn around the system can be more or less inclusive of these elements.

A system in which the boundary is closely defined will have only the fixed structures and extant processes within it and will have a wide range of inputs, many of which may enter the system simultaneously . A system that has a very broad boundary might include people, materials, resources, and most information as a part of the system, with the input defined very narrowly as a discrete piece of information or energy.

Signal

The signal as developed in the Taguchi model has a more specific and limited definition. It is an input into the system, but it is limited to the means by which the user conveys to the system a deliberate intention to change (or adjust) the system output. In more general terms, it is the variable to which the system must respond in order to fulfill the user's intent. From this perspective, most of what are traditionally considered inputs into the system, i.e., people, materials, information, and so on, are already part of the system itself, and the signal is the discrete piece of information that determines the amount of energy transformed by the system.

The System

The structure of a system comprises aspects of the system that are relatively static or enduring . Process, on the other hand, refers to the behavior of the system. Consequently, process refers to those relatively dynamic or transient aspects of a system that are observable by virtue of change or instability. Traditional models of a system are based upon an input-process-output model. The system acts to transform the energy from the input into the output. This process, once established, is subject to variation due to internal and external factors that produce "error states" or outputs other than the desired output. These outputs can simply be wasted energy or may actually reduce the functional ability of the system itself.

If a particular team has a task to perform, e.g., solving a problem, you can consider the team to be a system that has inputs, output, and a process that allows the team members to transform their energy into the desired outputs. Team process can be defined as any activity (for example, meetings) that utilizes resources (the team) to transform inputs (ideas, skills, and qualities of team members) into outputs (discoveries, solutions to problems, proposals, actions, design ideas, products, etc.).

Often the energy that the team brings to the process is not used to best effect. For example, in a meeting, time may be wasted reiterating points because individuals have not paid attention to what is being discussed or because there is cross talk. This in turn leaves people annoyed and frustrated. These are examples of "error states" or undesirable outputs from the team process.

Output/Response

In traditional systems models, the output is whatever the system transforms, produces, or expresses into the environment as a consequence of the impact its structures and processes have on the input. An output can be anything from a newborn baby to well done barbecued ribs to a presentation to a text return. This is very important to understand because teams, by their nature, are complex and multifunctional. They cannot and should not be configured to produce one kind of response. Most teams will have a whole range of outputs with accompanying measures that will be used to identify how successful they are and how effective they are in transferring energy. The key is to identify appropriate measures that can be used to monitor the team's progress.

The Environment

It is important in attempting to maximize the performance of a team to identify factors that may have an impact on the performance of the team and its ability to maximize the transfer of its input into desired output and over which the team has little or no control. (Remember, the output of the team will be a new design ” however defined ” and it is up to the team to make that design "wanted" in the preset environment. This is not a small feat.) These factors are designated as internal or external to the system. It is these factors that cause energy to be wasted and undesirable output (error states) to occur.

External Variation

In teams, external variation factors may include such things as change in team membership, the environment in which the team is working, changing demands from management, corporate cultural, racial, and gender factors, and so on. In developing a group process, it is important to develop group systems and processes that are robust to these factors. In addition, team goals exert a considerable influence on the behavior of individual members, and goals can vary enormously. They could be output targets that will vary in accordance with the team's task ” problem-solving teams puzzling over the root cause of a problem; design teams considering the optimization of a particular system design to achieve robustness; a marketing team attempting to understand the exact details of customer requirements; or sports teams, each of which will have an entirely different set of performance goals depending upon the sport: soccer, football, tennis, golf, and so on.

Any analysis of working teams should take into account the objectives of the team and the situation in which the team performs because both will have a profound effect on the team functioning.

Internal Variation

Internal variation, on the other hand, relates to factors that are in the team system and its members. People may bring predetermined ideas about the correct design solution. They may have biases about other team members depending on their race, gender, function, grade, and so on. Certain team members may not get along with other team members and will regularly question, challenge, or contradict the others for no apparent reason. The team may not manage its time well and consequently may find itself chronically short of time at the end of meetings.

Team members may not know how to ask open questions that will open up fresh avenues of information. Closed questions will result in familiar dead ends or nonproductive and previously rejected ideas. Team members may not know how to build on the ideas of other team members and, consequently, good ideas may be regularly lost. If the reader needs help in this area, we recommend a review of Volume I, Part II.

The Boundary

At the simplest level, boundaries can be put around almost anything, thereby defining it as a system. In practice, the identification of the boundary is the key to successful system analysis. The classification of factors (signal, control, and variation) that impact on the system is dependent on the way in which the boundary is defined. For example, by setting the boundary of the system fairly wide, to include the team members, environment, resources, information, and so on, leaving only the directive from the champion or the monthly output target outside, more factors would be considered as control factors and fewer as variation. In this case, the directive from the champion would be the signal factor. The team members, environment (or aspect of it), and so on would be control factors.

External variations would then include disruptions to the team process from sources outside the team boundary. Internal variations would include attitudinal, cultural, and intellectual variations among and between team members and variations in environmental conditions (e.g., temperature). By setting a narrower boundary, many of the factors such as environment and resources would be considered external to the system and therefore would become noise factors rather than control factors. These issues are important because they determine the team's strategy for dealing with variations and establishing a means of becoming robust to them.

CONTROLLING A TEAM PROCESS: CONFORMANCE IN TEAMS

A tale in Hellenic mythology describes the behavior of Procrustes ” an innkeeper by the Corinthean peninsula. Procrustes took his clientele, people of definite natural shape and size , and either stretched or truncated their limbs so that they might fit the mattresses he provided. There are many echoes here of the original approach to quality, "We know what you want, we will design it, you will buy it and you will like it." Or the now famous quality euphemism, "We are not sure of what is really quality, but we sure know it, if we ever see it."

Fortunately, this philosophy is being transformed into a "customer-driven approach" and the pursuit of Total Quality Excellence through DFSS. It is not entirely unreasonable, therefore, when it comes to monitoring groups or teams, to identify an alternative to the current emphasis on fitting the behavior of team members into behavioral roles through a "Procrustean" method, that is, by squeezing identity and function into personality models like those of Belbin, Myers-Briggs, Bion, and so on, through normalization and pressure to conform. Remember, one of the diversity issues is the fact that everyone is different and we are all much better because of that difference.

This is particularly the case when old norms are not questioned and challenged regularly or when personality models are used to avoid genuine personal contact or in place of a genuine understanding of the uniqueness of others.

STRATEGIES FOR DEALING WITH VARIATION

There are four basic strategies for dealing with variation and its effect on the performance of a system: ignore the variation, attempt to control or eliminate the variation, compensate for the variation, or minimize the effect of the variation by making the system robust to it. Adopting the first of these strategies would mean accepting that teams will never function efficiently , but hoping that they will "do the best they can under the circumstances." As with an engineering system, this strategy would result in a lot of unhappy customers.

Generally, with engineering systems, you are encouraged to adopt strategy four first, reverting only to strategies two and three as a last resort because they are difficult and expensive to implement. While strategy four should also be chosen in the case of the team system wherever possible, you have greater flexibility in many cases to consider the other two options.

Controlling or Eliminating Variation

Procrustes' behavior is an example of controlling inner variation. While this approach to variation might be considered extreme, you may have some scope for selecting team members with the right characteristics for effective teamwork as well as the necessary technical expertise.

External variations are perhaps a little easier to deal with. For example, you could ensure that meetings are held away from the shop floor to reduce distractions due to noise (in the audible sense!) or hold them at an off-site location to minimize interruption.

Compensating for Variation

The principal means of compensating for variation is by providing some feedback on its effect on system output. The link between structure and process ”the way in which structure determines process, and for your purposes perhaps more importantly, the way that process determines structure ” is found in the concept of feedback loops. Feedback loops are so named because they are circular interrelationships that feed information from output back to input.

Information is transmitted within the system and is used to maintain stability, to bring about structural changes, and to facilitate interaction with other systems. Even the simplest model of the effective team includes this concept of feedback loops. By employing information feedback loops, systems may behave in ways that can be described as "goal seeking" or "purposive."

Negative feedback allows a system to maintain stability as in the case of the most commonly quoted example, a thermostat. A thermostat is controlled by negative feedback so that when the temperature increases above a certain level the heating is switched off, but when the temperature decreases sufficiently the heating is switched on. The process of maintaining stability is called "homeostasis." The capacity for such control is engineered into some mechanical systems and occurs naturally in all biological and social systems. Threats to the stability of the system will be countered in a powerful attempt to maintain homeostasis.

System Feedback

One alternative approach is to monitor those aspects of team behavior that are observable (i.e., gather "the voice of the process"). Descriptive Feedback offers a non-judgmental method of monitoring what happens in working groups. It allows team members to notice when team process is in control and meeting or exceeding predetermined expectations or drifting out of control and reducing potential. Descriptive Feedback provides three basic functions:

1. It makes explicit what is happening during team process.

2. It describes those characteristics of team process behavior, relationships, and feelings that may degrade or go out of control and inhibit the potential of the team.

3. It determines what, if anything, needs to be changed in order to facilitate continuous improvement in team process.

Feedback over time enables a team to establish performance-based control limits. By using these data, specific characteristics or variables relating to team process can be plotted over time. This will identify patterns that emerge and that can be used to identify and capture the degree of variability of the team. Some patterns are related to "in control" conditions, others to "out of control" conditions, just as the patterns of points on a control chart can be used to establish whether a manufacturing process is in control or out of control.

Based on feedback that describes what people notice and how they feel, the team is able to regulate its process and identify opportunities for improvement.

Minimizing the Effect of Variation

The Parameter Design approach used in quality engineering ” see Volume V of this series ” is concerned with minimizing the effect of variation factors by making the system robust. This involves identifying control factors ” in this case, aspects of the team process that are within the control of the team and that can be used to reduce the impact of variation factors without eliminating or controlling the variation factors themselves . An example of a "control factor" functioning in this way is the use of Warm-Up and its consideration of "place" (layout, heating, lighting, ventilation ) so that best use is made of the facility provided and distractions are minimized, even though the place itself and many of its features cannot be changed.

The key to a successful team lies not only in identifying those parameters that are critical for the efficient transformation of inputs to the team process into outputs but also in doing this with minimal loss of energy in error states and maximum robustness to variation factors in the environment. Different types of teams with different outputs required of them would have different parameters established for their most efficient performance. Many of the structures, processes and skills that could be used as control factors in a team process have been identified in Volume I, Part II of this series.

Through this process of observation, it is possible to establish control limits in a wider area of team performance. A number of the factors that have an impact on team performance can be observed and regulated through feedback, and "tolerance" for them can be established depending upon the makeup and objective of the team.

These factors include warming up and down, place, task, maintenance, process management, team roles, agenda management, communication skills, speaking guidelines, meeting management, exploratory thinking guidelines, experimental thinking guidelines, change management, action planning, and team parameters.

The traditional approach to engineering waits until the end of the design process to address the optimization of a system's performance ” in other words, after parameter values are selected and tolerances determined, often at the extremes of conditions and often without considering interactions among different components or subsystems. When the components and subsystems are integrated, and if performance does not meet the target value or the customer's requirements, parameter values are altered . Consequently, though the system may be adjusted to operate within tolerance, this process does not guarantee that the system is producing its ideal performance.

Similarly, traditional approaches to building teams have selected team members according to a number of factors: predetermined skills and knowledge, established roles for team members, and implemented structured norms. They also have waited until the end of the process of team design in order to optimize performance. If the team does not perform within the accepted values of these parameters, then it is adjusted: team members are changed, roles are redefined, norms are more strictly enforced. This, however, is against performance criteria that do not necessarily optimize the team's performance nor add to the motivation or job satisfaction of the team members.

The shift suggested in Parameter Design in engineering (and that may be applied to teams as well) is to move from establishing parameter values to identifying those parameters that are most important for the function of the process and then determine through experimental design the correct values for those parameters. The key is to establish the values that use the energy of the system most efficiently and that are resistant to uncontrollable impact from other factors internal or external to the system itself.

MONITORING TEAM PERFORMANCE

One way of monitoring team performance has already been suggested, namely the use of Descriptive Feedback. Gathering "the voice of the process" enables the team to evaluate its performance and to continuously improve its efficiency and hence its effectiveness, before completing the task. Preliminary work in using process-control charting from Statistical Process Control suggests that there is opportunity for application to group process. This provides a second means of monitoring and continuously improving the team's performance. Critical "control factors," identified using the Parameter Design approach, could be measured and monitored in this way. Based upon further refinement, it may be possible to establish control limits, targets, and tolerances for these factors.

System Interrelationships

A systems model of processes differs from traditional models in many ways, one of which is the notion of circular causality . In the non-systems view, every event has its cause or causes in preceding events and its effects on subsequent events: the scientist seeks the cause or effect. Using the linear method of causality, ultimate causes are sought by tracing back through proximate causes. However, many phenomena do not "fit" the linear model: the relationships between them ” and the relationships between the attributes or characteristics of the elements ” do not conform to this linear approach to causality.

In engineering systems, a direct cause and effect relationship often exists between the component of the system and the transformation of the input into an output. A steering wheel channels the input of the vehicle operator directly into the output of the system. That is, turning the steering wheel to the right or left actually turns the wheels of the vehicle to the right or the left. However, it is equally clear that error states or phenomena are nowhere near as simple or linear in the causal relationship. Feedback loops and circular causality create very complex interactions. Similarly, the choice of lubricants may not affect the performance of the system until months or years later, when early deterioration of a transmission would result in difficulty shifting gears.

Similarly, in teams, some cause and effect relationships are clearly related in time and others are not. Interventions by a timekeeper will affect the ability of the team to stick to its agenda. But other factors have more circular relationships. In a global problem-solving team, changing seating arrangements from the long-tabled boardroom style to a circular arrangement will result in more universal eye contact among team members, which may increase the team's communication. This leads to enhanced exchange of information, which may lead to a clearer identification of the problem which will, in turn, lead to a more targeted search for relevant data, which will finally lead to a root-cause identification for the problem. Changing the seating arrangement may enhance finding a root cause more quickly than might have been the case in boardroom seating, and the cause and effect chain may be quite intricate .