RESEARCH ISSUESSHAWNEE VICKERY, Feature Editor, Eli Broad Graduate School of Management, Michigan State University Operations Psychologyby Douglas M. Stewart, Eli Broad Graduate School of Management, Michigan State University As researchers, we are constantly seeking opportunities to make new and innovative contributions to our fields. In operations management (OM), one such opportunity is provided by a powerful research tool that is unfamiliar to many operations management scholars and has gained little attention in our literature. This tool has the capability of opening up new avenues of study, reinvigorating stagnant bodies of research, and improving the relevance and accessibility of our work in the eyes of practitioners. I am not referring to some forgotten algorithm or mathematical technique, but rather to the use of psychology. We have generally given insufficient attention to the role of the psychology of the participants in service and production systems, and grand opportunities exist for those who are willing to tackle this shortcoming. Our colleagues in other fields of business research routinely use psychology to successfully address research problems relevant to their areas of study. Organizational behaviorists, for instance, have drawn on psychology to help address questions that interest them, such as what motivates employees; how to keep their goals aligned with those of the organization; and what are the character traits of leaders that cause others to follow them? Even the very quantitatively oriented finance researchers have made major advances in understanding how financial markets operate through the introduction of psychology into their work. For example, Tverskey and Khanneman's prospect theory has provided insight into why and when people buy and sell, which was simply not captured by purely numerical analysis. Obviously, psychology does not have a role in every OM research topic. However, there are many more opportunities than one would initially expect. After all, even the Flexible Manufacturing Systems (FMSs) require human technicians for troubleshooting and reconfiguration. The question remains why, when there are so many opportunities, have we paid so little attention to psychology in our research? Perhaps the answer lies in the history and nature of operations management inquiry. Central to operations management research is the study of processes, and the specific form studied largely influences how the related research is conducted. Historically, our attention has focused almost exclusively on those manufacturing processes where the full application of Taylor's principles had reduced the people in the process to not much more than organic machines. With all human aspects of the system minimized, it was then a reasonable assumption to disregard psychology, and focus exclusively on the mechanical aspects of the process. Such assumptions are no longer valid in much of today's research for two reasons. First, we are facing diminishing returns from our analysis effort for many established research topics. With decreasingly incremental gains from each new algorithm or variable included in our analysis, the relative importance of the human aspect of those "organic machines" increases. At some point the heretofore-ignored psychological variables will outweigh our non-psychological refinements. The second, and perhaps more important reason, is that the nature of the processes we are studying has changed. In many of the processes studied today, people are acting less like machines and more as information processors and decision makers who guide, customize and improve upon the process. Some of this is due to our interest in new processes that are very people-intensive, such as those found in services, and our interest information systems (IS), which are intimately tied to decision making. But, it is also due to changes in the manufacturing environment that have elevated the shop floor worker beyond the realm of pure physical manipulation, and have required him to engage in greater mental processing. This is not to say that no scholars in our field have ever addressed psychology in their research. In fact, there is a substantial history of the use of psychology, particularly in industrial engineering, Total Quality Management (TQM), and service operations. Take, for instance, workstation design. Industrial engineers have long been able to rely on purely quantitative methods to facilitate minimal and natural movement when designing workstations for manual tasks. But when the workstation will be used for monitoring or controlling a production task, such as in a refinery or paper mill, psychological issues became very important. Individual gauges on the control panel are designed to be easily monitored while the process is within normal parameters, yet also provide obvious warning cues as process parameters drift away from their target levels. Moreover, the gauges and controls must correspond in an intuitive manner during corrective actions (e.g., moving a lever upward does not cause the gauge to move downward). The overall displays are designed to mimic the operators' mental model of the system, as well as its physical reality, so that the information provided does not require excessive interpretation prior to its use in decision making. And, should alarm warnings sound, they are designed to be easily distinguished one from another, and mutable once they have sounded to avoid distracting from the ensuing problem solving activities (Vicente & Rasmussen, 1988). Many psychological issues have also surfaced in TQM research. Often these issues have concerned the effectiveness of quality circles, such as the optimum circle size, and various methods to facilitate brainstorming and problem solving. Other psychological issues have centered on providing motivation for better quality through specifying who is responsible for rework, and informing employees about how and where their jobs fit in the grander scheme of the process. Because of their human-intensive nature, services have been another area where psychology has often been seen in operations management related research. Researchers have looked at psychology in terms of the roles and role conflict between the participants in the service encounter (Shamir, 1980), measures of the service orientation of employees (Hogan, Hogan, & Busch, 1984), and the impact of the physical surroundings and dress on the participants in the service encounter (Bitner, 1992). Despite this history, there remains a general reticence to using psychology to address operations management problems. Perhaps this is due to a limited familiarity with the subject within our research community, or perhaps it arises from a feeling that the use of psychology does not constitute good OM research. Regardless of the cause, this reticence can trap us within unproductive research paradigms. In essence, we become the man from the old adage who, having only a hammer, treats every problem as a nail. We must be willing to take up another tool when our existing ones are not right for the problems at hand. As a case in point, consider the issue of service quality. Researchers have struggled for years with the issue of how to improve service quality. Some have focused on designing the service package to more closely meet the customer's needs, others on developing a service culture that motivates and empowers employees to exceed customers' expectations. Still others have focused on how to best recover from service failures through the use of tools such as unconditional service guarantees. These approaches have used either existing research tools, or direct observation of what successful companies have done, and they have met with varying levels of success. At the process level of analysis, our efforts have been primarily concerned with the application of statistical process control (SPC) tools to reduce variation in the service environment. While this is consistent with our OM toolkit, services have few process variables that can be measured sufficiently well to allow the concise application of such statistical techniques. (In fact, usually the analysis has consisted of measuring processing or response times. A special issue of Quality Progress on customer service (November 1993) contained six articles on using SPC in services, all of which used time as the measurement variable.) Moreover, the most common forms of service failure often appear to have little to do with variation. Our myopic focus on the research tools at hand has kept us from stepping back from our work and directly asking, "Why, do services fail to deliver as intended?" If we are willing to do this, two reasons surface. The first reason focuses on an improper allocation of resources, which leads to unacceptable waits for customers, harried employees, and an emphasis on tasks that are not proportionate with their importance to the customer. These are design issues currently being addressed by existing research streams. The second reason why services fail is ignored in the existing literature-human error. Once this is apparent, the logical course of action is to investigate how and why these errors arise and how they can be prevented. To answer these questions we must obviously seek the aid of psychology. As it turns out, the psychology literature has devoted substantial attention to the cognitive mechanisms underlying human error. (For a good overview of this literature, see Reason, J., Human Error, New York: Cambridge University Press, 1990.) The application of this research to the service environment can yield interesting and useful results. For instance, when viewed in terms of the cognitive mechanisms involved, I have found that services are prone to a limited range of errors, such as omissions and double-capture slips (see sidebar on page 13) by the provider. These errors tend to take the form of strong-but-wrong responses, where the error is identifiable as the correct response for another situation. I have also seen, as expected, that customers and providers both make errors that lead to service failures. What is interesting here is that the cognitive mechanisms underlying customer errors differ from provider errors (Stewart, 1997). For example, although past research has demonstrated that retail customers make routine skill-based errors (Reason & Lukas, 1984), these errors do not appear to have an effect on service quality. The reason is that customers don't appear to attribute blame for these errors to the service firm. Instead, those customer errors that affect service quality are usually the result of a limited subset of rule- based cognitive error mechanisms such as selectivity and matching bias (Stewart, 1997). The implication is that, as far as the customer is concerned, we shouldn't spend much effort on preventing simple action slips, but rather, greater attention should be given to the elicitation of appropriate scripts and decision frameworks. For instance, many of us have experienced the feeling of walking into a restaurant and being unsure of what actions were expected of us. We asked ourselves, "Should we wait to be seated, should we find a table, or should we look for a place to place our order?" Our ensuing behavior was dictated by which restaurant script we chose to follow-be it the formal sit-down script, the cafè script, or the self-service script. If we chose poorly, the result was an error on our part, such as waiting an inordinate amount of time for a host or hostess who did not actually exist. The task facing the service designer in this situation is to reduce script ambiguity through highlighting the supporting and removing the conflicting cues for script choice, and then ensuring that the service process steps are consistent with the desired script. In this example, psychology has not only helped us describe the service quality problem in an innovative manner, but also has provided an established theoretical base that we can build on to understand why service errors are made, and hence find guidance about how they might be addressed. Undoubtedly, other opportunities for the use of psychology in operations management abound. For example, the growing research interest in better integrating IS and OM should draw heavily on psychology. This would allow us to better address questions such as: what information should be provided continuously; what should be provided on demand; what should the system prompt for; what form should the information be conveyed in (audible, written, graphic); and what level of abstraction should be present? Underlying each of these questions is the psychology concerning the ability of human operators to process information. The reader is encouraged to examine his or her research and consider how psychology might be of benefit. Careful reflection should yield many potential opportunities, some perhaps in well-established research areas where numerical analysis may have overlooked positive and negative psychological ramifications, others in completely new areas of study. Although it may entail some extra effort to become familiar with the relevant literature, the introduction of psychology into our work can provide new theoretical justifications for our methods. And, if the results do not actually increase the applicability and relevance of our work, the inclusion at least demonstrates to our audience that we have considered the possibility. At the very least, a little psychology would help make our work more accessible to the practitioner. As our students demonstrate to us over and over, most managers usually do have a better intuitive grasp of psychology than math anyway. References
|