Bulletin, October/November 2006

We Don’t Do That Here: Using Cognitive Work Analysis to Learn About Organizational Culture

by Patricia Katopol

Patricia Katopol is a doctoral candidate at the Information School, University of Washington. She can be reached at pfk<at>u.washington.edu.

What is it like to be an administrative assistant in local government? Is it the same as being an administrative assistant in state government? In a software firm? When it comes to the part of the job that involves information, much of the work seems similar. Most administrative assistants find, manipulate, store, retrieve and present information. They do some of this information-related work for themselves, but much of what an administrative or executive assistant does is in support of a manager. When we think about devising new information systems for people who do similar work, what difference does it make where they work? Does the organizational culture matter? 

In a study of the information culture of government support staff personnel and its implication for managerial decision-making, I used cognitive work analysis (CWA) to inform the questions and to provide the analytical framework used to learn about the organizational culture that I believe frames information culture. That is, organizational culture determines behavioral constraints regarding information-related activities (information culture) in this environment, for example creating, retrieving, storing, transferring or exchanging information at work. In this article, I discuss organizations, organizational culture, the constraining effect of institutional forces on the behavior of organizational members and how using CWA as an analytical framework promotes learning about them.

An organization can be defined as an economy or market as well as an adaptive social structure. In an organization, people are joined in shifting alliances and work toward goals that may or may not be shared by others. Further, they may be constrained in attaining those goals by forces internal and external to the organization, forces of which they may not be aware.

Organizations may be rational – a formal system with defined job titles, work rules and hierarchies. It is fairly easy for an outsider to figure out who is doing what, how they’re doing it and with or to whom. For example, Vicente, who reviews the elements of an organization in his work on CWA, appears to describe a typical rational system, that is, one that is purposeful, exhibits organized collective behavior, has a rational criterion for decision making, offers mechanisms for choice and engages in effective planning. 

Alternatively, organizations may be natural, with a focus on people and behavior. People in an organization come to work with their own desires, needs and goals, which they will pursue in addition to, or instead of, the organization’s needs and goals. Or, an organization may be open, a perspective which examines the organization’s inner life as well as its interactions with the environment. Note that the open organization is not without boundaries. Rather, it is engaged simultaneously in two ongoing types of activity – activities that maintain the organization and activities that change the organization. 

Organizations may not fit neatly into one type. Some parts of the same organization may be very open, such as marketing. Other parts, such as research and design, may have to operate in a more or less closed system in order to preserve intellectual property rights.

Organizational Culture
When we think of organizational culture, generally what comes to mind is how the organization feels to those on the inside. What does it feel like to work there? Culture can be briefly described as shared group assumptions. It is a way of looking at the world and solving problems. Culture determines what we find important, how we show it is important and what people consider the correct and incorrect ways to act within the organizational setting. Culture is so important that we teach it to new group members so that they will also do the right things, in the right ways, thereby preserving us (the organization) and our ways. 

When we go into an organization, we get a glimpse of its culture by examining the artifacts – things such as furniture or how people dress; espoused values – what people say are their values; and underlying assumptions – assumptions which determine how the organization sees itself in relation to its environment. 

This integrated view of the organization – one entity whose parts work together to present a united front to the outsider – is only one way of many to examine organizational culture. The existence of formal and informal structures such as departments, divisions, subgroups or multinational units all work against the likelihood that there is one culture in any given organization. If culture formation centers on mission and strategy, goals, means, measurement and correction, it is not surprising that different subgroups in the organization will have different cultures as they pursue their own missions and develop their own strategies and goals in their struggle to gain control over scarce resources in the organization. 

Integration is only one perspective used in analyzing organization culture. There are at least two other perspectives – differentiation and fragmentation – which take into account varying levels of conflict and tension. 

Integration is the perspective used when we talk about a single, cohesive, organizational culture. Assumptions are shared, newcomers are taught the “right” way (the organization’s way) to think about and deal with problems. While it makes sense to look at the points where an organization is in agreement, providing a convenient label for the organization, this search for consistency and consensus has it limits. For one thing, whose consensus? Management may have shared values and assumptions, but what of other subgroups such as support staff or sales? The integration perspective can miss the differences and possible sources of conflict and tensions among people in different groups. 

Thus, in the differentiation perspective we examine where espoused values differ from actions or words. Differences may be due to tensions between hierarchical levels and may be accepted by the organization as a natural part of it. The differences are recognized as part of organizational life – sometimes there’s tension, and sometimes the competing groups or subcultures get along. 

The fragmentation perspective, finally, includes ambiguity, contradictions and tension. Boundaries are unclear, conflict is expected and ambiguity is a norm. 

When we speak to participants about their work, we must listen carefully to what they are also telling us about what it is like to work in that particular organization. What do they say about ritual, what stories and scripts are well known or repeated by organizational members, how do they talk about themselves, their place in the organization and the organization itself? Is jargon used, and if so, how? Is it emotional or task-related? Are there derogatory or complimentary names for departments and people? Is knowledge of jargon used to keep people out? Is humor used as a way to share experiences and bring people together or is it used to put down people and departments? When we go to the worksite, what does the physical environment look like? How do people dress and what do their offices look like? 

Applying the integration perspective can provide an overview of the organization that is useful for gaining an initial understanding. But at the same time as the researcher begins to see the organization as a whole, subcultures and conflicts become visible. Differentiation allows the researcher to see the activities that reduce organizational cohesion. Fragmentation recognizes that organizations are torn by reactions to the environment, struggles to find unity and members with competing goals. I think that it is in the differentiation and fragmentation perspectives that researchers using CWA can find a clearer and more detailed picture of the constraints, goals and priorities of organizational members. 

People are constrained from doing as they wish when acting in the organization. From just being polite to people you don’t like to not telling the boss what you really think about his idea, there are many constraints on actions. However, often we do not consciously realize that they are there or why they exist. Institutionalism provides one explanation for these constraints. It provides a reason why people do not act rationally, why organizations continue to act in ways that, from the outside, appear to be inefficient and illogical and why the actions (including technology implementations) that succeed in one place, fail in another. 

Institutionalism grew out of the attempts of disciplines such as sociology, political science and economics to explain why these actions take place. As organizations adopt practices that increase their likelihood of survival, the organization’s actions become less rational and less efficient. As efficiency becomes less likely due to institutionalism, the organization takes steps to decouple its activities from its structure. This decoupling hides the internal workings of the organization from evaluators who may notice inefficiency. So we find that schools produce students, not learning, and local government doesn’t govern, but produces rules and regulations. 

Institutionalism affects the actions of organizations because it affects how the people within them act. For example, professionals such as accountants and nurses are bound by professional codes of conduct that determine what actions are legitimate behaviors for them. In turn, these constraints on legitimate action affect the actions of the organization. Even support staff have legitimate behavior and know that acting outside that legitimacy would call them into question. In this manner institutionalism shapes the organizational culture as it provides mechanisms and rationales that reinforce the perceived right ways to act in a given situation, including information-related behaviors. If we fail to recognize that forces such as institutionalism and organizational culture shape the actions of organization members and that they will not, indeed cannot, act the way we want them to simply because we develop a better system, we will not understand why our attempts to improve systems and services so often fail. 

Organizational Culture and CWA 
CWA is a multifaceted approach to analyzing human information behavior that informs data collection and analysis. For example, actors in a work environment have information behaviors that are affected by hierarchy, conflict and institutional pressures that force them to act in ways that are seemingly illogical or are not directly responsive to the task at hand because they address some other need that must be filled. Asserting that people will seek to fill an information need without considering all of the other personal, contextual and task-based constraints facing them does not adequately address the complexity in which many work related information behaviors take place. 

CWA examines the actions of actors (a person acting in a context, not simply a user of a system) who are affected by constraints (factors outside of the actor’s control and which shape his behavior). These constraints occur in the following dimensions: work environment, work domain, organization, activity and actor characteristics. All the dimensions are important to understanding the actor’s environment and task in order to design better systems and services for the actor’s use. It is, however, analysis of the work domain (the environment in which the work is done) and the organization that can provide the most insight into organization culture.

To have data for the work domain and organization analysis, I asked questions that were framed to elicit information about the department’s role within the larger organization, departmental responsibilities and how work is assigned both to the actor and the department. I asked about work and personal relationships among staff at the same and different hierarchical levels and whether ideas and suggestions proffered by support staff were encouraged and respected. Participants tended to go into detail when responding to these open-ended questions and told me personal stories of what it was like to work there, who or what department really had control of things, the extent to which they were free to do their jobs as they saw fit and how they saw their roles in the organization. 

In my study of government support staff, I added the element of historical background to the CWA-informed interview protocol. Participants were asked how work was done previously and how those procedures compare to those currently used. I believe this factor to be an important addition to the CWA framework, especially where non-technical systems such as social networks or the work process itself, rather than a technology, may be important to the design of new systems. While CWA rightly focuses on the current work process and environment, it is beneficial to know what came before and whether actors found these processes and procedures to be useful and effective. For example, in government organizations where managers and their preferences for certain work routines can come and go with the change of political administrations, there may be an organizational memory of previous practices that worked well and that might be revived if applicable to the current situation. Until recently CWA was used primarily to inform the design of complex systems such as power plant controls. In such cases, it is of little use to include old or obsolete technologies in the design of new systems. However, as we can now use CWA to inform systems and services that are not technological in nature, I think it makes sense to ask participants about previous work routines and systems. In addition, asking these questions offers insights into the organizational culture. For example, was an effective process changed at the whim of a new manager or was it a necessary response to changes in the work? 

Finally, by using CWA, the researcher has the freedom to apply the theories that are most useful to the study. This study uses institutionalism to explain pressures that force actors to behave in certain ways; forces that shape the organizational culture and in effect, pre-determine the way actors will go about their information-related activities. Actors’ information behaviors are affected by their roles at work, connections with other people in the workplace and their professional training. Information behavior at work is rarely an isolated activity whose methods and boundaries are not framed by organizational culture. 

Using CWA forces us to look beyond the work and examine the culture in which work takes place. I suggest that using CWA provides a more complete examination of the work because we can now see its particulars. That is, we see the work through the actors’ eyes. We learn what it means to do the work, in that place, with those particular co-workers and managers, and all constrained by institutionalism and organizational culture. Only when we have an appreciation of the culture in which people work can we say we understand the work well enough to suggest changes in the system. 

For Further Reading
Fidel, R., Pejtersen, A.M., Cleal, B., & Bruce, H. (2004). A multidimensional approach to the study of human-information interaction: A case study of collaborative information retrieval. Journal of the American Society for Information Science, 55(11), 939-953.

Martin, J. (1992). Cultures in organizations: Three perspectives. New York: Oxford University Press.

Powell, W. W., & DiMaggio, P. J. (1991). The new institutionalism in organizational analysis. Chicago: University of Chicago Press.

Scott, W. R. (1998). Organizations: Rational, natural and open systems. Upper Saddle River, NJ: Prentice-Hall, Inc. 

Vicente, K. (1999). CWA. Mahwah NJ: Lawrence Erlbaum Associates.