Back to Diane Nahl Home Page
Back to Table of Contents of the User-Centered Revolution Article

The New Paradigm Establishes Itself

The user-centered revolution in library and information science was generated by the general availability of search technology for novice end-users who rapidly out-numbered experts and intermediaries. The needs of this new class of searchers have changed the priorities in the information environment. In response, researchers have developed new types of data that more closely represent users' actual experience with systems. This interest led to an evolution in the methodologies applied to user studies over the past 25 years, from survey questionnaires and feedback forms, to interviews, journals, logs, and case histories. The development in methodology represents a shift in focus from gathering anonymous, statistical data on users, to individualizing users through ethnographic and psychological approaches.

With technology came transaction logs, a new online, real-time data gathering method that records every key stroke of every user. Such logs have become a unique source of detailed information on what people do when they search. However, even this level of detail was not sufficient to understand why people make particular command choices and errors. Systematic self-reporting methods were introduced to obtain cognitive and affective data on reasonings, reactions, and decisions. Ethnographic studies using personal journals, protocol analysis, and self-witnessing methods uncovered the overwhelming quality of feelings in the research process. This new knowledge about users is driving the development of new systems and interfaces designed around user-centered principles. However, there was some resistance among designers, administrators, and librarians to the idea of taking care of novices' feelings.

User-Centered vs. System-Centered Approaches

The term "user-centered" arose in the minds of librarians and information specialists who wanted more services for end-users, patrons, clients, students, and the public at large within the information environment. They saw the status quo as "system-centered," that is, insufficiently responsive to user needs and realities. One early observer notes the "growing movement" at the time for user-oriented systems to shift from "source-controlled" to "receiver-controlled"(1). The expression user-centered has become, a maxim, a principle, and represents a new paradigm for defining, measuring, and explaining the behavior of library users, database searchers, and Internet navigators. A search of the literature of the past two decades yields dozens of articles and some books bearing in their title or abstract, one of the following expressions of this new orientation:

* user-centered

* user-friendly

* user-based

* user-oriented

* user-responsive

* client-centered

* human-centered

* people-centered

Two themes emerged in reviewing literature with a conscious user-centered focus: (1) methods of application that seek to alleviate specific user needs, and (2) the integration of theoretical justifications and accounts that researchers have proposed for transforming the information environment through user-centered interventions.

It is worth noting at the outset that the system-centered approach in the design of information facilities routinely advises a testing procedure with actual users within the design process, before features are finalized. However, this technique has proved to be inadequate primarily because user-categories have been based on designers' intuitions rather than on research about how users actually perceive, think, organize, and intend during the information seeking process. As pointed out by Walther and O'Neil in the early days of online users, "Intuition and good judgment on the part of the designer seem insufficient for specifying the features which an interface should possess" (p. 115). Despite this attempt to give users a greater weight in the interface design equation, system-centeredness has remained essentially the same. The reason for this is clarified in the discussion of the advent of the user-centered paradigm and its essential characteristics. In speaking of this basic shift, one commentator points to a distinction between the concept of information as "like a brick" while users are like "empty buckets into which [information] bricks can be thrown" (3)(p. 160). According to Dervin, the invalidity of this metaphor is shown by two outcomes: the "nonuse of information" by both professionals and the general public, and the user-oriented case studies of how people acquire and assimilate information in their daily lives and occupations.

Reviewing the data, Dervin concludes:

The important point here is, when it comes to understanding how and why and with what effect human beings pay attention to, process, and use something that an outside observer calls information, we must start by understanding what the user (or potential user) calls information. Information processing and use are, within the context of relativistic assumptions about information, sense-making activities. The emphasis here is on the word "making" for it denotes that the perceiver of the information is not an empty bucket but is actively making sense" (p. 164-5).

Information as an "observer construct" has had to give way to information as a "user construct" since the "empty bucket has evolved into a thinking, self-controlling human being" and "information changes from brick to clay, moved and shaped in unique ways by each perceiver" (3)(p. 169). Since making sense of the world is part of the human condition, knowledge-building is therefore a "personal information seeking" activity:

Questions asked have dealt with locating self and others (Where am I? Where are they?), with finding possibilities (What can I do?) and evaluating possibilities (Will it work?), and with assessing aloneness (Is anyone listening? Doesn't anyone else agree? Am I the only one?). Perhaps the most important aspect of the findings to date is that they support the premise that looking for information relativistically is a more powerful entry for understanding information needs and use" (3)(p. 170).

Looking at information seeking relatively requires one to recognize that situation-specific factors are more important to users than general considerations. For example, users' interests and involvement ("issue oriented") are better predictors of information-seeking activity than users' education or experience. When individuals see their situation as giving them options for decision or movement, they are more likely to engage in information seeking than when they see themselves under total constraint (3)(p. 171-2). Another reason that information is "relativistic" is that "the effect of the information received is the use created for it by the user." In other words, the situation-specific, momentary information need determines what information will have impact at that time. Research in this new "relativistic information framework" can help systems designers make the shift to this user-centered orientation.

Replacing the Centrality of the System

In the early days of the new paradigm there was not yet a full appreciation of the need to remove the centrality of the system. User-centered consciousness was not fully awakened until one saw the necessity of designing systems through understanding users. The following is an example of an early attempt to promote "an empirical approach to user-centered design:"

The problems which accrue to the human user by virtue of being made a component of the system -- of the user actually being placed on-line to the computer -- were largely ignored until very recently. A user protest led to calls for a shift of emphasis from the elegance of algorithms and computational efficiency to the discovery of ways to make the on-line user interface more acceptable to the user. (2)(p.114).

As it turned out, more than "a shift of emphasis" was required. For instance, Walther and O'Neil define the flexibility of an interface as "the factors inherent in the interface program which would make it easier or more convenient for the user to express his commands, given that the commands had to follow a specific syntax" (2)(p.116). However, this orientation remains part of the system-centered paradigm since it passively accepts the priority or fixity of the system, though it recognizes that acceptability and convenience to users is to be taken into account. Another example is in the earlier attempts to use empirical or objective measures of user satisfaction. Researchers often turned to existing "attitude measures" such as the semantic differential (4) which provides subjects with a series of bi-polar scales upon which to rate their attitude toward aspects of the system:

The results indicate that the users of the flexible version [of the online editor], irrespective of any other factors, rated their version as more tolerant, more flexible, more like a person, more friendly, and more pleasant than the inflexible version" (2)(p.117).

Future development of the user-centered paradigm will view such attempts as insufficient. Attitude scales have three inherent weaknesses relating to the object being rated, the timing of the rating, and its content. These problems need to be remedied by utilizing different techniques that avoid them.

First, the object to be rated by the user needs to be the user's experience, not the system. Thus, not, "On a scale of 1 to 7, how easy is this editor to use?" but rather, "On a scale of 1 to 7, how much confusion did you experience with this editor?" Second, the act of rating needs to be in close proximity to where the activity occurs, not at the end of a session, or even later, but concurrently, during the activity itself. Thus, more than a single probe is necessary, since a user's experience may vary greatly from one minute to the next, depending on various difficulties. Third, the content of the rating category must originate from the user's mental structure rather than the experimenter's. Otherwise, measures such as attitude scales, survey questions, census data, and personality tests represent system-centered accommodations to user protests of user-unfriendly systems.

More recent statements of user-centered system evaluation and design continue to promote the centrality of users (5) but there is a shift in emphasis away from treating users as a method of determining baseline measures for design standards, toward "application-based evaluation procedures." This represents an orientation toward operational effectiveness rather than design efficiency. In this view, system redesign can be motivated by how inconvenient it is to learn rather than how effectively it can perform under ideal conditions.

Criteria for User-Friendliness

Wallace has proposed a framework for evaluating the user-friendliness of a system (6). The list includes the following types of measures:

* users must be given options for preferences

* online system operation and assistance should be available at beginner and more advanced levels

* system behavior has to be transparent to users

* warning messages must be benign or light-hearted

* system design must take into account the physical and psychological needs of users

* system use should require no special skills

* the general language of novices and their routine level of communicative skills should not be surpassed in instructions or explanations

* the system should behave in a uniform way to allow users to anticipate functions

* a variety of types of problem-solving should be available for particular operations

* learning by doing should be the preferred mode of instruction and presentation

One project that implemented these and additional user-friendly features is known as ELSA, "An Intelligent Electronic Library Search Assistant" (7). Four design concepts guided the development of ELSA:

* using labels that are transparent to users (e.g., buttons with the names "Run Search," "New Search," and "Show Full List"

* providing assistance with controlled vocabulary (e.g., a synonyms list)

* actively helping users explore topic areas (e.g., providing suggestions for narrower topics and screening out more distantly related items)

* accommodating different styles of interaction (e.g., an option to type in search items or select them with the mouse from a list)

Another historical example is the user-centered redesign of the MELVYL system in response to empirically discovering the answer to the question, "What do users really want?" According to Farley (8), the process of collecting and maintaining a comment database optional to users, has uncovered bugs in the system as well as numerous user-unfriendly features that could form the basis for redesign. Today, with Internet networking of libraries, MELVYL has learned to "compete in attractiveness" to users who are more sophisticated and computer literate, though still novice searchers.

The development of intelligent search assistants such as ELSA would not have occurred without the climate created by the user-centered revolution that gave it scientific legitimacy and made research grants available for user studies. The first phase of the transformation concluded when user-friendliness became a legitimate scientific research issue. An important thrust came from the human-computer interaction field populated by engineers, programmers, and applied psychologists.

Toward Human-Centered Design

In a whimsical mood, one of the architects of the user-centered revolution in the area of systems design and technology, describes a still common attitude regarding the introduction of new technology:

One of the things that stands out when talking to long-term users of poorly designed systems is that these people take great pride in their skills. They had to go through great difficulties to master the system, and they are rightfully proud of having done so. ... Rather than ease the situation for those who follow, it becomes a sort of initiation rite. The hardy survivors of the experience claim to share a common bond and look with disdain upon those who have not been through the same rites. They share horror stories with one another" (9) p.176.

Norman reveals that he "got attacked by hundreds of professional programmers across the country" when he wrote an article criticizing the UNIX system as user-unfriendly: "If I didn't approve of UNIX, they told me, I had no business using it. Besides, who was I anyway to criticize computer software? In other words, you weren't allowed to criticize unless you were a professional. Being a mere user of the stuff didn't qualify" (9)(p.177). However, a new paradigm was in the ascendancy, and is achieving full victory. The user-centered revolution not only gives users more clout in the design equation but works to make users the central point of focus and interest for planners, backers, designers, engineers, and trainers of information systems.

Understandability and usability are two principles that apply equally to design and documentation issues. According to proponents of "cognitive engineering," (10) understandibility of instructions depends on providing a good conceptual model of the task or system. Good models are those that allow users to predict the effects of their actions. Users will invent false models when the appearance of some system features (the visible setup) does not match the functions desired (10)(p.16). The designer's model is formalized by the design process itself. The user's model evolves through interaction with the system. The system image consists of the visible structure or appearance of the interface (controls, keys, instructions, labels, visualization aids or interface metaphor, etc.). All communication between designer and user takes place through the system image. If the system image does not make the design model clear, users end up with a wrong mental model as evidenced by symptoms such as inability to predict, difficulty in using, having a shallow understanding, and inability to solve problems. One of the key factors in helping users develop an accurate conceptual model is to provide immediate feedback for every action or choice made.

User Errors

By studying the psychology of everyday actions, cognitive engineers have discovered many reasons why people make errors. First, one should note that people blame themselves for errors: "It is as if they take perverse pride in thinking of themselves as mechanically [or technologically] incompetent" (10)(p.35). Second, experience has shown that good system interfaces can be designed that make errors easy to detect and easy to reverse or, if not, have minimal consequences. Good designers heed the maxim: If an error is possible, someone will make it. One common source of errors is inadequate mapping, where users develop misconceptions in their mental models of the system. The existence of conflictual habits is another source of error (e.g., frequently hitting the wrong key for a particular operation). Users are reassured when they are reminded that making errors is routine in everyday automated activities such as keyboarding. Natural situations allow for routine correction of errors, such as retyping or abandoning a sentence midway and beginning a new sentence. This explains why error recovery procedures have become essential in designing computer systems in the user-centered era.

Norman (10)(p.107) has identified several types of user errors. Capture errors occur when a frequently performed activity suddenly takes charge instead of the intended one (e.g., hitting the wrong Function key; hitting Return instead of an F key; waiting while staring at the screen without having entered the command; etc.). Description errors result from performing a correct action but on the wrong object, which occurs when we are distracted, bored or fatigued. Data-driven errors occur when the sight of some object or information intrudes into an action sequence (e.g., typing the wrong information on a line; choosing an unintended menu item). Associated activation errors occur when an association of ideas or thoughts trigger a non-relevant action (e.g., typing an incorrect subject heading phrase). Loss of activation errors occur when the goal of an action is momentarily forgotten or when part of a sequence is temporarily inhibited. Mode errors occur when users forget what mode they are in when there are several (e.g., typing a command appropriate to another mode). Inherent in creating a user-friendly information environment is the requirement to discover what errors users routinely make, to expect them, and to provide accessible remedies for them:

For the first time in history, we are truly free to make machines that fit human needs, independent of mechanical constraints. New information technologies can enhance the power of human thinking, for machine plus person can do more than either alone, but only if the technology complements human abilities. We need more information-processing tools that complement our thinking, reasoning, and memory skills as comfortably as the calculator enhances our arithmetic skills. We can transform the hard technology of computers and information processing into soft technology suitable for people if we start with the needs of the human users, not with the requirements of the technology" (11)(p. 52-3).

At this stage, according to one commentator, designers of information systems are now committed to a "user-centered design philosophy" and have begun implementing interfaces that take into account "end-user information seeking needs" (12)

.Back to Diane Nahl Home Page
Back to Table of Contents of the User-Centered Revolution Article