HCI Glossary

abstraction: A term in a notational system or other information artifact that is defined with reference to the primitive concepts of the system, or with reference to other abstractions that are ultimately defined by primitives. It is frequently used to aggregate many instances, so that all can be manipulated by a single action; thus, a heading style is a typical word-processor abstraction defined in terms of font properties, and all its instances can be altered by altering its definition.
abstraction hierarchy: A hierarchical description of the functional structure of a work domain, in which work-domain purposes are related to underlying physical structures. Also referred to as a structural means-ends hierarchy because links between adjacent levels connect ends (upper nodes) to means (lower nodes).
actions: A lower level of organization in human endeavor, including simple gestures and other motor behavior.
activity: The significant and typically collective endeavors of humans. It is the fundamental concept in activity theory, which conceives of activity as conducted through individual actions and mediated by artifacts. In the cognitive-dimensions framework, an activity is likewise a significant endeavor conducted through individual actions, but six generic types of activity are distinguished (e.g., search, design, etc.); at present, only individual interaction with information artifacts is considered.
ACT-RPM: The latest in a series of computational cognitive architectures developed by John Anderson and colleagues. ACT-RPM contains architectural mechanisms for cognitive, perceptual, and motor performance and learning. It was one of the architectures reviewed in Pew & Mavor (1998). More information about this series of architectures can be found in Anderson (1976, 1983, 1993) and Byrne & Anderson (2001).
adaptationism, adaptationist approach: The thesis that selection pressures have been the most powerful cause of evolution; a useful methodological heuristic is to assume that biological and behavioral structures are the results of adaptation to the environment.
affordance: Perceptual characteristics of an object that make it obvious what the object can do and how it can be manipulated.
affordances: According to the theory of James J. Gibson, these are directly perceived possibilities for action in the environment.
analog representations: Representations that share the structure of the things they represent. For example, a picture is an analog representation. See also propositional representations.
anthropology: The investigation of social structure, social relationships, and individual social action through an emphasis on culture—originally “other cultures” but increasingly the emphasis is on “home” cultures.
apex: A computational architecture using resource scheduling and reactive planning (techniques from artificial intelligence). More information can be found in Freed (1998). It has been used as a GOMS modeling tool (John, 2002; Remington, et al., 2002).
artifacts: A diverse class of human-created systems, applications, tools, and conventions, including language and mathematics, that mediate human activity. Artifacts are the products of prior human activity; they both enable and constrain current human activity, and their use helps to orient the design of future artifacts. An information artifact (as used in the cognitive-dimensions framework) is an artifact designed to store, create, present, or manipulate information, whether noninteractive (e.g., a book or a map) or interactive (e.g., a spreadsheet or a heating controller).
automatic cognitive processes: Processes that are relatively quick and effortless, requiring little attention or monitoring. Well-practiced skills, like walking and driving, or signing one’s name, are examples of automatic cognitive processes. See also controlled cognitive processes.
bounded rationality: The idea, from Herbert Simon (e.g., Simon, 1982), that human agents are rational in that they act so as to meet their goals according to their knowledge; they are only boundedly rational, however, in that they cannot necessarily compute ideal, optimal decisions. Instead, their decisions are bounded by environmental constraints on their performance (such as the need to act quickly), interacting with limits on access to knowledge and limits on the ability to process relevant information.
CCT: See cognitive complexity theory.
claims analysis: An analytic evaluation method involving the identification of scenario features that have significant positive and negative usability consequences.
claims feature: An element of a situation or an interactive system that has positive or negative consequences for people in this or similar situations.
CMN-GOMS: The original version of GOMS created by Card, Moran, and Newell (1980a). The “CMN” was added before “GOMS” when other versions of GOMS began to appear (e.g., CPM-GOMS and NGOMSL), to differentiate the specific representation used by Card, Moran, and Newell from the concepts in GOMS.
cognitive architecture: The fixed structure that realizes a cognitive system. It typically decribes the memories and processors in a cognitive system and how information flows between them. This is in contrast to the knowledge laid on top of an architecture to allow it to perform a task in a particular domain.
cognitive artifact: A manmade or modified tool to support mental activity. Examples include number systems, slide rules, navigational charts, and even language itself. While generally applied to a single individual, within the framework of DCog, a cognitive artifact is also a tool that supports the coordination of information processing between entities within a functional system.
cognitive complexity theory (CCT): A computational cognitive architecture introduced by Kieras and Polson in the 1980s and used as the basis for NGOMSL. It was realized in a production system. More information can be found in Bovair, Kieras, & Polson, 1988, 1990; Kieras & Polson, 1985.
cognitive dimension: A characteristic of the way information is structured and represented—one that is shared by many notations and interaction languages of different types and, by its interaction with the human cognitive architecture, that has a strong influence on how people use the notation and determines what cognitive strategies can be pursued. Any pair of dimensions can be manipulated independently of each other, although typically a third dimension must be allowed to change (pairwise independence). (More exactly, a cognitive dimension is not solely a characteristic of the notation, but a joint characteristic of the notation and the environment in which the notation is used, whether based on paper and pencil or computer or even based on voice and sound.) Examples such as viscosity, premature commitment, and others are defined in the text. Note that in the cognitive-dimensions framework, dimensions are not evaluative per se, but only in relation to a particular type of activity; for example, viscosity is a problem for modification activities but not for transcription activities. See also cognitive-dimensions framework.
cognitive-dimensions framework: This states the pertinent values of cognitive dimensions that are required to support a given type of activity, and thereby provides a means to evaluate an information artifact.
cognitive engineering: A multidisciplinary area of research concerned with the analysis, modeling, design, and evaluation of complex sociotechnical systems. It is sometimes also called cognitive systems engineering.
cognitive ethnography: A qualitative method of data collection used by DCog researchers based around observation. It is “cognitive” because it focuses on computational information transformations within a functional system.
cognitive science: The scientific project dedicated to understanding the processes and representations underlying intelligent action.
cognitive work analysis: An approach to analyzing human work that focuses on how effective courses of action are constrained by both ecological and cognitive considerations.
cognitivism: Maintains a Cartesian dualism, attributing human conduct to the operation of mental predicates. Often associated with a computational theory of mind, it is a predominate paradigm within human-computer interaction (HCI).
common ground: Some piece of knowledge is common ground if all the people conversing know it and they all know that the others know it, too. (See the text for a more formal and comprehensive definition.)
complexity theory: The study of how much time or resources are required to compute things. The complexity of an algorithm or problem is typically measured relative to the size of the problem, n, and expressed in order notation. For example, a time complexity of O(n2) means that the time it takes to perform the calculation increases with the square of the problem size.
computational equivalence: Two representations that, in addition to being informationally equivilent, make the same information equally readily accessible. See also informational equivalence.
computational metaphor (of cognition): Hutchins (1995a, 117) defines a “computation” as referring to “the propagation of representational state across representational media.” The computational metaphor is the position taken that cognition is a form of computation, and that mental state is encoded analogously to computer representations. The term “metaphor.” Within cognitive science, the computational metaphor is also known as the “representational theory of mind,” in which computations are actions on representations. DCog claims that the computational metaphor can be applied to a unit of analysis broader than an individual’s mind (i.e., the functional system).
computer-supported cooperative work (CSCW): The design of systems to support interaction and cooperative working. It emerged in the late 1980s as a result of dissatisfactions with the predominantly cognitivist paradigm employed in human-computer interaction (HCI) and in recognition of the importance of questions regarding organizations, work, and interaction for the design of computer systems.
conceptual framework: A structure describing the concepts in a cognitive system, less specific than a cognitive architecture.
constraints: In the context of cognitive work analysis, factors that limit, but do not prescribe, how effective work activity might be carried out.
controlled cognitive processes: Processes that require monitoring and effort or attention during their execution. See also automatic cognitive processes.
control task analysis: A way of analyzing work that focuses on the control that must be exercised over a work domain and the tasks implied to exercise such control. An analytic phase of cognitive work analysis.
conversation analysis: Originated by Harvey Sacks, this is often coupled with ethnomethodology and is the study of the way in which conversationalists order and accomplish their exchange of speech with each other as a situated and locally organized matter.
cost structure of information: An analysis of the resource and opportunity costs involved in accessing and handling information from a physical or virtual information system.
CPM-GOMS: A version of GOMS developed by John in the 1980s that explicitly joined GOMS to the model human processor. It included perceptual and motor operators as well as cognitive operators. Perceptual, cognitive, and motor operators could run in parallel, subject to resource and information dependencies. More information can be found in Gray, John, & Atwood, 1993; John 1988, 1990; John & Gray, 1992, 1994, 1995.
critical path: The path through subtasks (operators) in a PERT chart that determines the length of the total task. Used in CPM-GOMS.
CSCW: See computer-supported cooperative work.
DCog: The form of distributed cognition developed by Hutchins in the early 1990s. It is distinguished from other uses of the term distributed cognition by its explicitly computational perspective on goal-based activity systems.
descriptive model: A model that describes how a system or person behaves and that provides a framework or context for thinking about or describing a problem or situation. Usually based on data gained through empirical observation, it is often little more than a verbal or graphic articulation of categories or identifiable features in an interface.
design: Encompasses activities and actions directed at producing new artifacts. Design work is collective and multidisciplinary. It often includes professional designers, technologists, and future users of the artifacts.
design rationale: Arguments for why (or why not) a feature or set of features should be incorporated into a design.
direct perception: The theory of James J. Gibson that claims that the visual environment is perceived “directly,” as opposed to being indirectly inferred from sense data.
ecological interface design: An approach to interface design that uses Rasmussen’s abstraction hierarchy and skills-rules-knowledge framework to specify interfaces that support adaptive human operator behavior in the face of events or situations that systems designers did not anticipate.
EPIC: A computational cognitive architecture introduced by Kieras in the late 1990s. EPIC is known for computationally implementing perceptual and motor processes. More information can be found in Kieras & Meyer, 1997; Meyer & Kieras, 1997a, 1997b.
ethnography: Originating in the anthropology of Bronislaw Malinowski, this has come to mean the study of cultural and societal matters from inside their operations. It is associated with fieldwork, which emphasizes the importance of participatory methods for collecting data about social phenomena.
ethnomethodology: Founded by Harold Garfinkel, this is an alternate sociology that eschews the predominantly theoretical caste of sociology in favor of studies of the way in which social order is attested to, constituted, and used within the practical doings of societal members. Ethnomethodological studies of work have focused on the situated practices, methods, and interactions through which members order their work.
Fitts’ Law: A predictive model of human movement developed by Paul Fitts in the 1950s. The model predicts the time for rapid aimed movements, that is, the time to acquire a target of a specified size at a specified distance.
focus: A concept used in focus shift analysis. It deals with the objects of work as the anchoring point of identifying the level of human action. A focus shift draws attention to the shift between one level of action and another. It indicates the dynamics of the situation that is the main point of concern.
formalism: A philosophy or practice that focuses on the manipulation of surface representations or symbols with a disregard for the underlying semantics and meaning. This may be as a strong philosophical stance or it may be because it is believed that, in a certain situation and for a particular end, the symbols capture faithfully the underlying meaning.
formal methods: In computing, this refers to the use of specification and analysis methods based on algebraic or set theoretic mathematical methods. It does not encompass all uses of mathematical analysis in computing, but it is principally applied to those concerned with the formal specification of discrete systems.
formal specification: See formal methods.
formative model: A model that identifies requirements that must be satisfied so that a system can behave in a new, desired way if needed.
functional system: The functional system is the unit of intelligent activity that is analyzed in DCog; it may be composed of a collection of any number of individuals and artifacts. It is bounded by the problem, and it includes all of the entities that compose the problem space and which are involved in problem solving.
goals: The “G” in GOMS, goals are what the user is trying to achieve through interaction with the computer system.
GOMS: GOMS is an analytic technique for making quantitative and qualitative predictions about skilled behavior with a computer system. The acronym stands for goals, operators, methods, and selection rules.
grounding: The process by which common ground is developed.
Guiard’s model of bimanual control: A model that describes the roles of the hands in performing tasks that typically involved two hands, one being dominant or preferred and the other being nondominant or nonpreferred.
H: A standard operator in a keystroke-level model (KLM), H represents the act of homing the hands between input devices, such as between the keyboard and the mouse. It was empirically determined to take 400 msec by Card, Moran, and Newell (1980b, 1983).
halting problem: Early in computing theory, it was established that it was impossible to produce a program that can reliably tell whether any program will eventually halt (rather than run forever). As well as its intrinsic significance, this is typical of a range of impossibility results that establish the fundamental limitations of computation.
hidden dependencies: Important but invisible links between entities in the cognitive dimensions framework. Frequently the links are visible in one direction but not in the other (cell dependencies in spreadsheets, for example). See the text for more details.
informational equivalence: Two representations that contain the same information. See also computational equivalence.
information artifact: See artifact.
information-foraging theory: A framework that employs an adaptationist methodology to develop quantitative and computational psychological models of user strategies for seeking, gathering, and consuming information.
information processing: Within cognitive science, problem solving is seen as an information-processing activity [See computational metaphor (of cognition)], in which encoded information is acted on and transformed in the resolution of a goal held by a cognitive entity.
information psychophysics: The application of visual psychophysical techniques to perception of information conveyed through a visual representation.
information scent: The relation of environmental cues in the environment (such as bibliographic citations, or World Wide Web link text) to users’ assessments of the value of information sources.
information visualizations: The use of computer-supported, interactive, visual representations of abstract data to amplify cognition.
interaction language: The commands used to instruct an interactive information artifact such as a word processor or a heating controller. These commands are a form of notation, but typically what the user can see as feedback is the effect of the command rather than the command itself. For example, in a word processor’s interaction language, giving the command Delete Word deletes a word; in the same word processor’s macro language, however, giving the same command Delete Word inserts the appropriate symbols into a program. See also notation.
joint action: Both people involved in a joint action intend to do their part and believe that the joint action includes their and the other person’s part. (See the text for a more formal and comprehensive definition.)
juxtaposability: A level of visibility (in the cognitive-dimensions framework) at which any required subset of components can be viewed simultaneously. It is an essential requirement for certain cognitive strategies, such as design by modification of existing material, or for checking consistency of form across instances of similar components. See also visibility.
K: A standard operator in a keystroke-level model (KLM), K represents the act of striking a key on an input device like a keyboard or a mouse. The duration depends on the device and the skill of the user. A variety of quantitative estimates can be found in Card, Moran, and Newell (1980b, 1983).
keystroke-level model (KLM): The simplest GOMS technique. It provides standard keystroke-level operators (D, H, K, M, P, and R) with estimates of duration and rules for placing the M (mental) operators. It abstracts away from the goal hierarchy, multiple methods, and selection rules found in other GOMS techniques. More information can be found in Card, Moran, and Newell (1980b, 1983).
KLM: See keystroke-level model.
latent semantic analysis (LSA): A statistical theory and method of identifying word meaning through analysis of context of word use.
LSA: See latent semantic analysis.
learning as development: A view of learning that emphasizes the triggering and maturation of skills (versus their acquisition through practice). Some versions construe this as biologically prefigured (Piaget), and others construe it as socially mediated (Vygotsky). In this view, people are not understood not only in terms of what they are but in terms of what they are becoming.
M: A standard operator in a keystroke-level model (KLM), M represents the act of mentally preparing to execute a command. M is a “catch-all” operator that may include such unobservable actions as making a decision, remembering a procedure or command, visually searching for information, and so on. Because M encompasses all such actions, it is a powerful approximation to human decision making that makes the KLM a relatively easy modeling method to use. M was empirically determined to average 1350 msec by Card, Moran, and Newell (1980b, 1983).
mediated communication: Human-to-human communication may be mediated by technology—for example, by telephone or video phone, by text chat, or by email.
mental model: A cognitive structure of concepts and procedures that users refer to when selecting relevant goals, choosing and executing appropriate actions, and understanding what happens when they interact with a computer system (or other tool).
methods: The “M” in “GOMS,” methods are well-learned sequences of subgoals and operators that can accomplish a goal. There may be more than one method to accomplish a goal. If so, then selection rules are used to determine which method to use in the current task situation.
MHP: See model human processor.
model: An simplified expression of reality that is helpful for designing, evaluating, or understanding the behavior of a complex artifact such as a computer system.
model human processor (MHP): The information-processing cognitive architecture introduced by Card, Moran, and Newell in 1983. It was never realized in a computational form, but it sufficiently specified its memories, processors, communications, and principles of operation that some quantitative predictions of human behavior could be derived. The MHP was merged with GOMS through CPM-GOMS.
MOO (multi-user domain object-oriented): A MUD in which the characters, their behavior, and all other services are built and extended using an object-oriented programming language.
motor control: A branch of experimental psychology concerned with the study of human movement.
MUD(multi-user domain): A persistent collaborative environment that is modeled on a geographic space.
NGOMSL: A version of GOMS that is computationally realized in a tool called GLEAN. NGOMSL stands for Natural GOMS Language, and was developed by Kieras in the mid-1980s. More information can be found in Kieras (1997) and Kieras, et al. (1995).
nonfunctional requirements: Qualities of a system under development that are not directly related to its function, such as maintainability or reliability.
normative model: A model that identifies one or a few best ways for a system or person to behave. The model usually offers a criterion or criteria against which to evaluate behavior.
notation: A system of symbols used in specialized fields to represent facts or values (as in a circuit diagram) or to give instructions (as in a programming language), usually subject to rules of combination and ordering (“syntax”). Although the symbols are discrete, there may be an admixture of analog features (relative placement of components in a circuit diagram, or layout of text in a program). This is a wider definition than that of Nelson Goodman’s (1968), for example, but it is more typical of conventional usage. Notations may be persistent (written down) or transient (spoken or otherwise fleeting). See also interaction language.
operator: The “O” in “GOMS,” operators are the actions that the software allows the user to take. Operators can be defined at many different levels of abstraction, but most GOMS models define them at a concrete level, like button presses and menu selections.
P: A standard operator in a keystroke-level model (KLM), P represents the act of pointing, that is, moving a cursor with a mouse. It was empirically determined to average 1100 msec by Card, Moran, and Newell (1980b, 1983), but it can also be calculated with Fitts’ Law.
participatory design: A design movement primarily associated with Scandinavia, in which future users of the artifacts being designed participate in the original design work.
PERT chart: Program Evaluation Review Technique, a methodology developed by the U.S. Navy in the 1950s to manage the Polaris submarine missile program. A similar methodology, the Critical Path Method (CPM), which was developed for project management in the private sector at about the same time, has become synonymous with PERT, so that the technique is known by any variation on the names: PERT, CPM, or PERT/CPM. These methods are used in cognitive modeling to depict the parallel operation of perceptual, cognitive, and motor operators, with resource allocation and information-flow constraints.
PIE model: An example of an abstract formal model—that is, one that is used to analyze a class of systems and usability problems rather than specifying a particular system. The PIE model was developed at York University in the mid-1980s and was one of the first steps in a new stream of formal method work in human-computer interaction (HCI) that began at that time. See: http://www.hcibook.com/alan/topics/formal/
placeholder: Something (a physical artifact, electronic record, or human memory) that explicitly or tacitly maintains the current position within a formal or informal process. See: http://www.hcibook.com/alan/topics/triggers/
practice: When being trained as a carpenter or a nurse, for example, one shares a practice. At the same time, each individual who possesses a practice keeps it up and changes it as well. It is practice that allows us to talk about more than just individual skills, knowledge, and judgment, and not just about a “generic” human being. Practice is shaped historically, which is of particular relevance for design and use of computer applications.
preattentive processing: The rapid detection of visual features theorized to occur in parallel before the operation of selective attention.
predictive model: A model that allows metrics of human performance to be determined analytically without undertaking time-consuming and resource-intensive experiments.
Problem Behavior Graph: A graphical depiction of search through a problem space.
problem solving: Searching through a problem space from a known start state to a desired end state, or one of a set of desired end states, applying operators of uncertain outcome to move from state to state. Problem solving typically refers to a path through the problem space that includes explorations of deadend paths and backing up to prior states. (see skilled behavior for a contrasting type of behavior)
problem space: A mental representation of a problem, including the start state, the goal state, and the operators or moves that allow transitions between states. According to Newell and Simon’s (1972) theory of problem solving, humans solve problems by constructing and searching a problem space.
procedural knowledge: Also called “how to do it” knowledge. The knowledge of which operators to perform to move from a known start state to a desired state in a problem space.
Production system: A programming language made up of conditional statements. If all conditions are satisfied, then action is taken.
profile: In cognitive-dimensions analysis, a statement of what is required (in cognitive-dimensions terms) to support a given generic activity. The profile of an activity states for each cognitive dimension whether that dimension is material and, if so, what value it should have (e.g., for modification, viscosity must be low, visibility must be high, etc.).
progressive evaluation: In the cognitive-dimensions framework, the ability to review or test a partially complete structure, to check on progress to date it is important for novices, and becomes less so as one gains experience. Some systems allow only a complete design to be reviewed; others allow a review at any stage.
propagation of representations: A component of information-processing activity in which sequences of transformations from one representational form (or medium) to another result in information-processing activity. Hutchins claims that “representational states are propagated from one medium to another by bringing the states of the media into coordination with one another” (1995a, 117).
propositional representations: Representations that have a more-or-less arbitrary structure; for example, a word or a sentence in a natural language. See also analog representations.
R: A standard operator in a keystroke-level model (KLM), R represents the time a user spends waiting for the computer to respond. Only the time actually spent waiting—beyond any mental preparation time (M)—is included in R. R must be estimated for each computer system modeled.
representation: This can be seen as encoded information, either as a symbolic abstraction of a thing (classical cognitive science), or as a distributed set of nodes (PDP) that, together, have meaning. According to the representational theory of mind, human brains operate on symbolic representations, or codes. DCog extends this to show how transformations to representations need not be entirely symbolic, but may be enacted through manipulations on physical media that have a representational status (e.g., a navigational chart or a drawing).
representational state: This is defined by Hutchins (1995a, 117) as “a configuration of the elements in a medium that can be interpreted as a representation.” Problem solving occurs by successive rerepresentations of the problem (i.e., a representation of the problem) through a series of intermediate representational states into a solution (i.e., a representation of the solution). DCog researchers attempt to make the representational state of a functional system explicit and document how changes to its representational state result in goal-directed problem-solving activity.
scenario: A narrative or story that describes the activities of one or more persons, including information about goals, expectations, actions, and reactions.
schedule chart: The graphic depiction of a PERT or CPM chart.
secondary notation: In the cognitive-dimensions framework, extra information in means other than formal syntax. Examples include layout in programs and circuit diagrams and penciled annotations on music. See the text for more details.
selection rules: The “S” in GOMS, selection rules are the rules people use to choose between multiple methods that accomplish the same goal. They typically depend on characteristics of the task or user’s personal preferences or knowledge.
set theory: One of the foundations of nearly all mathematics and formalism in computing is the manipulation of sets (unordered collections of things) and functions.
situated action: One can plan to descend a set of rapids in a canoe, and one can plan and replan as one goes along, but one cannot (successfully) plan the descent and then merely execute the actions. Suchman, in her 1986 book “Plans and situated actions”, analyzed action as necessarily improvisational. Actions are undertaken in response to the constantly changing physical and social environment. An important consequence of this is to make plans and planning a resource for action, and not simply as the determinant of action.
situation model: A mental model derived from a text that represents what the text is about.
skilled behavior: Movement through a problem space by applying a known operator at every state, from a known start state to a known end state. (see problem solving for a contrasting type of behavior).
skills-rules-knowledge framework: A taxonomy of three qualitatively distinct levels of cognitive control. An expert human operator may exhibit any or all levels of cognitive control, depending upon the situation. This is not to be confused with models that describe stages of acquiring expertise or skill.
soar: A computational cognitive architecture developed by Allen Newel, John Laird, and Paul Rosenbloom in the early 1980s. It was used as the exemplar in Newell’s 1990 book Unified Theories of Cognition. It has undergone continual development from its inception in many universities and has also been commercialized for complex modeling in military simulations and intelligent agents for video games. Soar was one of the architectures reviewed in Pew & Mavor (1998). More information about this series of architectures can be found in Newell (1990).
socially distributed cognition: This is the theoretical position that goal-directed group activity can be understood in computational terms.
social-organizational analysis: A way of analyzing work that focuses on how work is organized and shared across people and supporting tools. It is an analytic phase of cognitive work analysis.
sociology: The investigation of social structure, social relationships, and individual social action.
socio-organizational Church-Turing hypothesis: The recognition that organizations perform, among other things an information-processing role, and the supposition that this means we are likely to see similar structural elements and processes in the physical and social aspects of the organizations as we do in electronic computers. See: http://www.hcibook.com/alan/topics/ecology/
spreading activation: This is a computational process that determines activation values over a set of interassociated cognitive structures. The spread of activation from one cognitive structure to another is determined by weighting values on associations among chunks. Activation values indicate degree of relevance to ongoing cognitive processes.
state: In computing, this usually refers to the inner memory of a computer at a particular point of time, but more generally it is that in the present that encapsulates all that of the past that can effect things in the future.
state transition network (STN): A representation of dynamic systems including states (usually as labeled circles or boxes) and arcs labeled by actions that form transitions between the states.
status-event analysis: A collection of informal and formal analysis techniques that focus on the differences and relationships among discrete event phenomena (e.g., a key press, a project deadline) and more continuous status phenomena (e.g., the current mouse position, the weather, the current screen display) See: http://www.hcibook.com/alan/topics/status/
strategies analysis: A way of analyzing work that focuses on different ways that a particular control task might be carried out. An analytic phase of cognitive work analysis.
STN: See state transition network.
subgoal: A goal that must be accomplished in service of a higher-level goal. Goals are often expressed as a hierarchy of goals and several levels of subgoals. The lowest level of subgoals are accomplished through operators, which are decomposed no further.
symbolic interactionism: Originated in the social psychology of George Herbert Mead and the sociological methodology of Herbert Blumer, this places emphasis on the individual in explanations of the transactions of people and society.
syndetic modeling framework: A framework that looks at the interactions among different kinds of models—in particular, models of system behavior and models of human activity. See Barnard, May, Duke, and Duce (2000).
task analysis: Any process that identifies and examines tasks that must be performed, either by users, other agents, or the system itself, to accomplish the goals of an interactive system.
task-artifact cycle: All human artifacts both enable and constrain human activity. Human use of current artifacts thus suggests possibilities and requirements for the design of future artifacts, which in turn will ultimately both enable and constrain human activity.
tradeoff: An issue (often in design) that is understood to have competing arguments, usually positive and negative impacts of an option.
transfer of learning: A learning and memory phenomenon in which what is learned in one situation facilitates understanding and behavior in a similar situation.
use/users: Many human activities incorporate computer applications. In human-computer interaction (HCI), the terms use and users are applied to the common properties of computer applications in work activity, and to the people who use computer applications as part of their daily practice. These terms are somewhat unfortunate, as the people rarely construe their own activity as computer use per se or see themselves primarily as users of computer equipment.
viscosity: In the cognitive-dimensions framework, this is “resistance to change.” A viscous system is one that requires many individual actions to achieve what is conceptualized as a single change. It is the opposite of fluidity. See the text for more details.
visibility: In the cognitive-dimensions framework, this is the ability to view components easily whenever necessary. See also juxtaposability.
visual working memory: A limited-capacity visual store that is distinct from verbal working memory. It is a core component of modern cognitive theory.
work-domain analysis: A way of analyzing work that focuses on identifying the functional structure of the work domain with which a human operator will interact (rather than identifying tasks to be performed in the work domain). An analytic phase of cognitive work analysis.
worker-competencies analysis: A way of analyzing work that focuses on the cognitive competencies required of or evident in workers, given their training, expertise, and the way information is represented to them. An analytic phase of cognitive work analysis.
yoked-state space hypothesis: The claim that users of representational devices (such as computer systems) need to mentally represent the device itself, the domain to which the device refers, and the way in which the device represents the domain. See Payne, Squibb, and Howes (1993).
Z notation: A particular notation for formal specification developed principally at Programming Research Group Oxford University. It is based on set theory, with mechanisms to allow large specifications to be separated into parts and structured. See Spivy (1988).
zero-parameter models: Models that can make a priori predictions of quantitative performance of users on an interactive system. The are called zero-parameters because no parameters need to be set from data collected on the system in question; all numeric parameters can be set through a task analysis and pre-existing data from prior research.
zone of proximal development: The inventory of capabilities people can currently demonstrate with assistance (human and material support) and therefore may, in the future, be able to achieve by themselves.