Mastering Information Management Series
Closing the Cognitive Gaps:
How People Process Information
Chun Wei ChooThis article appeared in the
Faculty of Information Studies
University of Toronto
March 22, 1999, Financial Times of London, Mastering Information Management Series. Download the original article here.
Reprinted as a book chapter in Mastering Information Management, edited by Don Marchand, Tom Davenport, and Tim Dickson (FT-Prentice Hall, 2000).
We often see "information" described as a "resource." This tends to imply that information is some "thing" that resides in documents, information systems, or other artifacts. The information is assumed to be constant, unchanging. Its meaning is fixed by its representation in the artifact. A complementary view is to look at information not as an object but as the outcome of people constructing meaning out of messages and cues. Information resides not in artifacts but in the minds of individuals. Individuals actively create the meaning of information through their thoughts, actions, and feelings. Since individuals use information in order to solve a problem or perform a task, the social settings in which the information is encountered determines its value and salience. When we treat information as an object, we are concerned with how to acquire the information that we need, and how to represent the information that we have in order to enable access and processing. When we treat information as subjectively constructed, we are concerned with understanding the social and behavioral processes through which information is enacted and engaged. A fuller understanding of information seeking as social behavior helps us to design better information processes and information systems. This article hopes to provide a first step towards that understanding.
As shown in Figure 1, we divide the information seeking activity into three processes: experiencing of information needs, information seeking, and information use. We examine the cognitive, affective, and situational factors that influence each of these processes.
Figure 1. Human Information Seeking Processes
Since the Second World War, a large number of "information needs and uses" studies attempted to understand how different groups of people experience information needs and how these needs are met (or not met). The information needs and uses of scientists, engineers, physicians, academics, citizens, government officials, public sector employees, special needs groups, managers, and professionals have all been investigated. In recent years, Brenda Dervin of Ohio State University has been active in applying a sensemaking approach to examine how humans perceive information needs as cognitive gaps. In the sensemaking approach, a person is moving through space and time, taking steps through experiences. As long as she is able to make sense of her experiences, movement ahead is possible. From time to time, movement is blocked by the perception of a cognitive gap a situation that the person is unable to make sense of. To bridge this gap, the person seeks information to make new sense and uses the information to help her continue her journey. Dervin and her associates have completed over 40 studies in the past two decades based on the sensemaking approach. Their research suggests that the ways in which people perceive their cognitive gaps and the ways that they want information to help are good predictors of their information seeking behaviors. Better yet, the ways in which people perceive and define their sensemaking gaps can be coded into universal categories that are applicable across different groups of information users. Dervin has identified 8 situation gap categories:
Person sees two or more roads ahead
Person sees one road ahead but it is blocked
Person sees self as having no road
Person sees self as on a road that suddenly disappears
Person sees self as being dragged down a road unwillingly
Person judges how foggy is the road ahead
Person judges how many intersections are on the road
Person judges how many people are also travelling
Cognitive needs are draped in affective responses so that they are as much felt as they are thought about. When sense has run out, the lack of understanding creates a state of uncertainty. Carol Kuhlthau of Rutgers University found that uncertainty causes a number of affective symptoms, including anxiety, apprehension, confusion, frustration, and lack of confidence. These affective states motivate and direct the individuals information seeking and information use experience. Affective responses influence, and are influenced by, the individuals ability to construct meaning, focus information needs, manage moods and expectations, and deepen personal interest in the search. Tom Wilson of Sheffield University suggests that uncertainty and its affective symptoms can constitute a state of stress that the individual has to cope with. For example, research in health information seeking has contrasted "monitors" who prefer high levels of information input to cope with a stressful event and suffer less psychological arousal when they have the information; with "blunters" who prefer less information and suffer greater arousal when they receive a high information input.
At the situational level, information needs arise from the problems, uncertainties, and ambiguities encountered in specific contexts and experiences. Such situations and experiences are composed of a large number of elements that relate not just to subject matter, but also to situational conditions such as goal clarity and consensus, magnitude of risk, amount and locus of control, professional and social norms, time and resource constraints, and so on. As a result, the determination of information needs must not stop at asking "What do you want to know?" but must also address questions like: "Why do you need to know it?" "What does your problem look like?" "What do you know already?" "What do you anticipate finding?" and "How will this help you?" Susan MacMullin and Robert Taylor of Syracuse University suggest that problem situations be analyzed according to eleven "problem dimensions" that amplify information needs and form the criteria by which individuals assess the relevance and value of information (see Sidebar 1: Problem Dimensions).
The experiencing of information needs may lead to purposive information seeking. Purposive information seeking resembles a problem-solving or decision-making process. The individual identifies possible sources, differentiates and chooses a few sources, locates or makes contact with them, and interacts with the sources in order to obtain the desired information. In todays munificient information environment, how does the individual select between sources, and between information from different sources? In an economy where human attention is the scarce resource, how does an individual allocate time and energy when searching for information? An answer suggested by research is that an individual weighs the amount of effort required to use a source against the anticipated usefulness of the information from that source. At the same time, this evaluation of cost and benefit is modulated by the individuals personal interest and motivation, and by the complexity of the task or problem at hand.
At the cognitive level, the individual would select a source that he or she perceives to have the greater probability of providing information that will be relevant, usable, and helpful. Relevance and usability may depend on information attributes such as currency, comprehensivesness, and appropriateness to the specific problem situation. In addition, the individual would be concerned with the accuracy and reliability of the source. Source and information selection would then be guided by the perception of cognitive authority and trustworthiness. Research in information seeking often groups together some or all of these source attributes under the rubric of "perceived source quality" in order to examine its effect on source use.
At the affective level, the individuals degree of personal motivation and interest in the problem or topic would determine the amount of energy that he or she invests in information seeking. Carol Kuhlthau noted that as the information search progresses, initial feelings of uncertainty and anxiety fall as confidence rises. If a clear theme is developed to focus the search, the individual may become more highly motivated, and if the search proceeds well, there is a growing feeling of satisfaction and accomplishment. Drawing on social learning theory, Tom Wilson postulates that since a strong feeling of self-efficacy or personal mastery about using a source leads to greater source use, doubt about ones capacity to use a source properly would conversely lead to that source not being used, even if the source might be perceived to contain relevant information.
At the situational level, the selection and use of sources is influenced by the amount of time and effort that is required to locate or contact the source, and to interact with the source to extract information. At least three different kinds of "effort" or costs may be pertinent: physical effort (e.g. to travel to source); intellectual effort (e.g. to learn a classification system or user interface); and psychological effort (e.g. to deal with an unpleasant source). These source attributes may be bundled together in a category called "perceived source accessibility." The selection of sources then depends on their perceived quality and perceived accessibility. Additionally, the complexity of the task or the uncertainty of the task environment would also influence information seeking. A complex task characterized by numerous interdependent task elements which interact unpredictably may require broader information gathering and processing. Similarly, a complex, volatile external environment may induce greater information scanning. (See Sidebar 2: Environmental Scanning by CEOs.)
Just as there are universal categories of information needs, Brenda Dervin and Robert Taylor propose that there are eight general categories which describe how people use information. Thus, information may be used to: develop a context; understand a particular situation; know what and how to do something; get the facts about something; confirm another item of information; project future events; motivate or sustain personal involvement; and develop relationships, enhance status, reputation or personal fulfillment.
At the cognitive level, the individuals cognitive style and preferences would impact on the processing of information. A number of classifications have been developed to differentiate personality types and cognitive preferences. The Myers-Briggs Type Indicator is a widely used instrument for classifying personality types into 16 categories. Each personality type would display distinctive preferences and modes when gathering and using information (see Sidebar 3: Cognitive Styles). Another common cognitive style variable is "field dependence." Field dependent individuals tend to respond uncritically to environmental cues, whereas field independent individuals orient themselves correctly in spite of environmental cues. Daniel Kahneman (University of British Columbia) and Amos Tversky (Stanford University) discovered that when people use information to make judgments they rely on heuristics to simplify information processing. In certain situations, these simplifications can produce errors or biases. For example, to judge whether an event belongs to a category, people rely on mental stereotypes, but they often ignore other relevant information such as the distribution of the categories in the general population. To judge the frequency or likelihood of an event, people over-rely on recent, vivid, easy-to-recall information. To estimate a quantity they make adjustments from an initial anchor or suggestion. Unfortunately the adjustments are often inadequate.
At the affective level, people avoid using information that will arouse strong, negative emotions in others or in themselves. People use information selectively to avoid embarrassment, conflict or regret; to maintain self-image; and to enhance personal status or reputation. For example, decision makers are known to positively evaluate and continue a course of action even when the available information indicates that withdrawal is necessary to reduce further losses. One psychological factor behind this "escalation of commitment" is the desire to save face. Decision makers persist because they do not want to admit to themselves that they have made an error, much less expose their mistakes to others. In organizations where error-free decision making is valued, managers may attempt to hide their mistakes or postpone their discovery. Another example is the "not-invented-here" syndrome: the tendency of a longstanding group to reject new information from outside the group. Such behavior may be a natural consequence of individuals who over time, increase order and stability in their work environments so as to reduce the amount of stress and uncertainty that they need to face. As a result, the longer the individuals tenure in a group, the stronger their emotional attachment to beliefs and decisions that they helped create, and the more resistant they become towards outside new ideas and information.
At the situational level, the norms and rules of the social group, profession, or organization can influence information processing and use. Irving Janis of Yale University has observed how highly cohesive groups are susceptible to "groupthink." This happens when group members seek concurrence to such an extent that they compromise the processing and use of information, choosing to ignore or undervalue information that threatens group beliefs and solidarity. Donald Schon of the MIT describes how each profession develops its own language, values, overarching theories, and role definitions. Members adopt these as frames of reference through which information is processed to describe reality, explain phenomena, and reaffirm professional identity. Edgar Schein of the Sloan School of Management defines organizational culture as a pattern of shared assumptions developed by the organization as it learns to cope with problems of external adaptation and internal integration. Because the assumptions have worked well enough, they are considered valid and are therefore taught to new members as the correct way to perceive, think and feel in relation to those problems. As a result, organizational culture develops a shared framework for people in organizations to collectively make sense of information. An important part of organizational culture is organizational politics. In a contest for influence and power, information may be used as a resource to protect vested interests or to justify preferred courses of action.
Human Information Seeking: An Integrated Model
The three processes of information needs, information seeking, and information use may be integrated into a general model of how humans seek information. As shown in Figure 2, the individual experiences information needs as she perceives gaps in her state of knowledge or her ability to make sense. The perception of information needs is shaped by cognitive, affective, and situational factors. The individual may choose to suppress this information need by for example, avoiding the problem situation, so that no information seeking ensues. Alternatively, the individual may decide to bridge this gap of knowledge or understanding through purposive information seeking. During information seeking, the selection and use of sources and information depends on perceived source accessibilty, perceived source quality, task complexity, and personal interest. Information may also be received "incidentally" as a result of the individuals habitual scanning of the media or conversations with others, even though these activities were not directed at addressing specific information needs. The outcome of information seeking is a set of noticed, selected information that is a very small proportion of the total information that is received. How this information is then processed and put to use depends on the cognitive style and preferences of the individual, the emotional responses that accompany information processing, and the social and cultural context surrounding information use. The final outcome of information use is a change in the individuals state of knowledge, allowing the individual to make sense or take action. This in turn generates new experiences and new information needs, so that the cycle of information seeking is continuous.
Figure 2. Human Information Seeking: An Integrated Model
Implications for Practice
Our discussion here suggests some approaches to improve information management.
(1) Design information systems not just to answer queries, but to provide useful and useable information that would help individuals solve work-related problems and deal with specific requirements of problem situations.
System designers need to move beyond analyzing flows of data to understanding how the meaning of information is constructed, and how the context of organizational work is negotiated. For example, users should be able to query systems not just with account numbers or keywords, but also with life situations ("I am buying a house - what government information sources should I consult"), task descriptions ("I am writing a project plan on P get me the information that would help me"), and sensemaking questions ("What are the assumptions guiding our interpretation?").
(2) Increase awareness about the nature of human information seeking and processing, including an understanding of cognitive styles and limitations, and the ways that habitual routines and emotional defenses can block learning.
Cognitive diversity invigorates an organization, so the intention is not to pigeonhole people according to their presumed strengths and weaknesses. Instead, the goal is to ensure a lively mix of styles, skills and sensitivities in a group so as to heighten vigilance in information processing.
(3) Educate everyone to manage information quality and information quantity.
System designers need to understand how people assign value and salience to information. Users need to know how to evaluate the quality of sources and information, and how to balance between source quality and source accessibility. In restricting information quantity to prevent overload, users should also guard against premature closure: divergent information gathering (that consults many sources) is necessary to prepare the ground for convergent action-taking.
(4) Develop an organizational culture that values and encourages information sharing.
Some of the best information sources are ones colleagues in the same organization. Ironically, as organizations become more information-intensive, the less likely it is that members would share their information freely. Organizations now have to work at creating and sustaining affective systems and cultural climates which promote the sharing of information and knowledge.
Sidebar 1: Problem Dimensions
Information is typically needed in order to solve a problem. Users therefore value information that is not just on-topic, but that also helps them deal with the specific requirements of the problem situation at hand. Susan MacMullin and Robert Taylor of Syracuse University have identified 11 problem dimensions that amplify information needs beyond simply subject matter to include these situational contingencies. These problem dimensions also form the criteria that users apply to judge the value and utility of information.
Problems lie on a continuum between ...
1 Design ...
Options, alternatives, ranges ...
Small, detailed sets of data
2 Well-structured ...
Hard, quantitative data ...
Probablistic data on how to proceed
3 Simple ...
Path to goal state ...
Ways to reduce problem to simpler tasks
4 Specific goals ...
Goal operationalization and measurement ...
Preferences and directions
5 Initial state understood ...
Clarify unclear aspects of initial state ...
Initial state not understood
Soft, qualitative data to define initial state
6 Assumptions agreed upon ...
Information to help define problems ...
Assumptions not agreed upon
Views of the world, definition of terms
7 Assumptions explicit ...
Range of options, frames to analyze problems ...
Assumptions not explicit
Information to make assumptions explicit
8 Familiar pattern ...
Procedural and historical information ...
Substantive and future-oriented information
9 Magnitude of risk not great ...
Cost-effective search ...
Magnitude of risk great
Best available information: accurate, complete
10 Susceptible to empirical analysis ...
Objective, aggregated data ...
Not susceptible to empirical analysis
Experts opinions, forecasts, scenarios
11 Internal imposition ...
Clarification of internal goals, objectives ...
Information about external environment
Sidebar 2: Environment Scanning by CEOs
Most studies of how people use information sources found that the perceived accessibility of a source was a major, sometimes the most important, determinant of source use. For example, scientists, technologists, and managers are often sensitive to perceived source accessibility, so that a library or information center on the next floor or even a few offices away may be infrequently visited, even though the users recognize that the library contains more complete and current information than their close-at-hand sources.
However, a recent, countrywide study of how CEOs in the Canadian telecommunications industry scanned their business environments for information about trends and developments observed a different pattern. For these CEOs, the perceived quality of a source (in terms of its reliablity and relevance), was a more important predictor of source use than perceived source accessibility. The study (by Chun Wei Choo of the University of Toronto) observed that CEOs invested time and effort in contacting and interacting with less accessible sources such as customers, competitors, and business associates. These personal sources were rated highly by the CEOs for their ability to provide accurate and usable information. The study suggested that the switch of emphasis from source accessibility to source quality was because the CEOs were trying to make sense of a complex and ambiguous business environment, and because they were personally motivated and interested in learning about external trends and developments.
Sidebar 3: Cognitive Styles
One of the most widely used personality assessment instrument in the world is the Myers-Briggs Type Indicator (MBTI) classification, which is developed from the work of Carl Jung. MBTI analyzes personality types based on four pairs of traits:
Introversion versus Extraversion:
Introverts draw mental energy from themselves whereas extroverts draw energy from others.
Sensing versus "Intuiting":
Sensing types rely on information perceived through their five senses. Intuitive types rely more on patterns, relationships, and hunches.
Thinking versus Feeling:
Thinking types use information to make logical decisions based on objective criteria. Feeling types depend on personal values to decide between right and wrong.
Judging versus Perceiving:
Judging types move quickly to closure by making use of the available information. Perceiving types keep their options open by taking their time to gather sufficient information.
These four pairs of attributes are combined to create a matrix of 16 personality types. Each personality type is expected to display distinctive styles and preferences when processing and using information, as indicated above.
Sidebar 4: Further Reading
Choo, Chun Wei. 1998. The Knowing Organization: How Organizations Use Information to Construct Meaning, Create Knowledge, and Make Decisions. New York: Oxford University Press.
Dervin, Brenda. 1992. From the Mind's Eye of the 'User': The Sense-Making Qualitative-Quantitative Methodology. In Qualitative Research in Information Management, edited by J. D. Glazier and R. R. Powell. Englewood, CO: Libraries Unlimited.
Kuhlthau, Carol. 1993. Seeking Meaning: A Process Approach to Library and Information Services. Norwood, NJ: Ablex Publishing Corporation.
Taylor, Robert S. 1991. Information Use Environments. In Progress in Communication Science, edited by B. Dervin and M. J. Voigt. Norwood, NJ: Ablex Publishing Corporation.
Wilson, T. D. 1997. Information Behaviour: An Interdisciplinary Perspective. Information Processing & Management 33 (4):551-572.
Chun Wei Choo is associate professor at the Faculty of Information Studies, University of Toronto, Canada. His research interests include information management, knowledge management, and organizational learning. He welcomes correspondence on this subject; firstname.lastname@example.org; web http://choo.fis.utoronto.ca/.