Chapter 5 – Cognitive

Engineering

HCI: Developing Effective Organizational Information

Systems
Dov Te’eni
Jane Carey
Ping Zhang

Copyright 2006 John Wiley & Sons, Inc

Road Map
Context Foundation Application

4
1 Physical 7 8
Introduction Engineering Evaluation Principles &
Guidelines
3 5 11
Interactive Cognitive Methodology
2 Technologies Engineering
Org & 9 10
Business 6 Organizational Componential
Context Affective Tasks Design
Engineering

12 13
Relationship, Collaboration Social &
& Organization Global Issues
14
Changing Needs of IT
Development & Use

Additional Context

Copyright 2006 John Wiley & Sons, Inc

Learning Objectives
◼ Describe the main elements of human information
processing (HIP) and their interrelationships.
◼ Understand the causes and importance of HCI
complexity.
◼ Explain Norman’s seven-stage model and the design
implications of the execution and evaluation gaps.
◼ Use the Goals, Operators, Methods, and Selection
(GOMS) model to describe a simple task.
◼ Demonstrate how this material can be used to
compare two different HCI designs.

Copyright 2006 John Wiley & Sons, Inc

Introduction
◼ Cognition: is the mental action or process of
acquiring knowledge and understanding through
thought, experience, and the senses.
◼ Cognitive psychology: is a subdiscipline of
psychology exploring internal mental processes. It is
the study of how people perceive, remember, think,
speak, and solve problems.
◼ Cognitive engineering applies knowledge of cognitive
psychology to the design and development of
systems that support the cognitive processes of
users.

Copyright 2006 John Wiley & Sons, Inc

Introduction
◼ A simplified cognitive model of human information
processing serves as the basis for understanding
how cognitive resources such as memory and
attention are utilized in HCI.
◼ Additional models such as Norman’s model of user
activity and the GOMS (Goals, Operator, Methods,
Selection) model further demonstrate how user
interact with computers and the implications for
designers.
◼ Complexity is seen to represent the utilization of
scarce cognitive resources and reducing complexity
is one of the goals of cognitive engineering.

Copyright 2006 John Wiley & Sons, Inc

A Simplified View of Human
Information Processing (HIP)
◼ Figure 5.3 presents a simplified model of Human
Information Processing (HIP) which includes
processors and memories that interact in order to
process information.
◼ There are three types of processors:
◼ Perceptual: senses, detects and accepts input from the
external world and stores parts of the input in the
working memory. Data can be visual or auditory.
◼ Cognitive: interprets, manipulates and makes decisions
about the inputs.
◼ Motor processor: interpreting cognitive decisions into
physical actions.
Copyright 2006 John Wiley & Sons, Inc
A Simplified View of Human
Information Processing (HIP)
◼ Two types of memory:
◼ Working memory
◼ Long-term memory

◼ Attention is needed to facilitate and control this

cognitive model.

Copyright 2006 John Wiley & Sons, Inc

A Simplified View of Human
Information Processing (HIP)
Attention

Verbal
Memory data Processors
Spatial

Working Perception Cognition Motor

Long term
Memory Memory

Figure 5.3 A simplified model of Human Information Processing (HIP),

including memories and processor.

Copyright 2006 John Wiley & Sons, Inc

A Simplified View of Human
Information Processing (HIP)
◼ Performance: the speed and accuracy of the
information-processing task.
◼ HCI developers need to predict the speed
and accuracy of the intended outcome of the
interaction.
◼ Speed and accuracy are dependant on the
constraints posed by the human information
processing system.

Copyright 2006 John Wiley & Sons, Inc

A Simplified View of Human
Information Processing (HIP)
◼ Speed depends on the rate of processing and
transfer of data within the system and
between the human system and the external
world.
◼ Accuracy depends on the capacity to store
and retain information and then the ability to
retrieve it correctly.
◼ Thus speed and accuracy depend on how
well the processors and memory function as
a whole given the constraints of the shared
resources. Copyright 2006 John Wiley & Sons, Inc
A Simplified View of Human
Information Processing (HIP)
◼ Several principles of operations that govern
HIP aim at overcoming the limitations of
processors and memories.
◼ Cognitive processes are characterized as:
◼ Automatic behavior: behavior characterized
by cognitive processes that are fast and
cognitively undemanding.
◼ Controlled behavior: behavior characterized
by cognitive processes that are relatively slow
and cognitively demanding.
Copyright 2006 John Wiley & Sons, Inc
A Simplified View of Human
Information Processing (HIP)
◼ Processing of Images: processing characterized as
spatial, graphic, and holistic.
◼ Processing of verbal information: processing
characterized as sequential, linguistic, and
procedural.

Copyright 2006 John Wiley & Sons, Inc

A Simplified View of Human
Information Processing (HIP)
◼ Cognitive strategies: plans of mental actions
◼ People tend to select cognitive strategies that
minimize cognitive effort.
◼ Cognitive processing involves:
◼ Heuristics: rules of thumb that depend heavily on the
content and context of the task. Like shortcuts,
heuristics tend to use fewer cognitive resources and
provide us with a natural way to overcome our limited
cognitive resources. They lead to correct answers, but
not always. They are vulnerble to change in context.

Copyright 2006 John Wiley & Sons, Inc

A Simplified View of Human
Information Processing (HIP)
◼ Cognitive processing involves (cont.):
◼ Metaphor: a cognitive process in which an experience
is related to an already familiar concept. For example,
typewriter metaphor can be used to learn about a
computer keyboard.
◼ Mental model: a representation of the conceptual
structure of a device or a system. For example, a
person may build a mental model of the computer
system in which files are stored in rows like books in a
shelf. When you delete a file, you throw it away. This
uses the trash bin metaphor. The mental model is to
pick a file out of the trash bin and put it back on the
shelf.

Copyright 2006 John Wiley & Sons, Inc

The Complexity of HCI
◼ Cognitive engineering is a discipline that applies the
combined knowledge of cognitive psychology and
information technology to the design of artifacts.
◼ Cognitive engineering is based on the assumption
that the user is faced with limited cognitive resources
and, therefore, human computer interaction should
be designed to complement and fit these limitations.

Copyright 2006 John Wiley & Sons, Inc

The Complexity of HCI
◼ Complexity of HCI: the
human resources needed
for interacting with the
computer to accomplish
the task.
◼ The task, as the user sees
it, can be represented as
a set of intentions and
evaluations.
◼ The user engages in
human-computer
interaction to achieve Figure 5.4 HCI as a Bridge between Human and Computer
these interactions and
evaluations. Copyright 2006 John Wiley & Sons, Inc
The Complexity of HCI
◼ When a user is faced with a task and a computer system,
she perceives some level of complexity with this situation.
Each factor in the human-compute-task is, separately and
in combination, a source of complexity and can be viewed
as:
◼ Accomplishing a task with no computer in mind
◼ Operating a computer with no ulterior task in mind
◼ Combining both (i.e. using the computer to accomplish the
task)
◼ Complexity is a function of the amount of cognitive
resources needed by the user to accomplish a given task
with a given computerized system.

Copyright 2006 John Wiley & Sons, Inc

The Complexity of HCI
◼ The more the resources needed the higher the complexity.
◼ The same task can be of different complexity when
supported by different systems.
◼ The same task on the same system may be of different
complexity to different users.

Copyright 2006 John Wiley & Sons, Inc

Gulfs of Execution and Evaluation
◼ Gulf of execution: the gap between the
user’s goal and its computerized
implementation.
◼ Gulf of evaluation: the gap between the
computerized implementation of the user’s
goal and its evaluation by the user.

Copyright 2006 John Wiley & Sons, Inc

 Bridging the gaps using Norman’s
Model
•Establish a goal that needs to be Goals
accomplished.
•Form the intention (or hierarchy of
intentions) that will accomplish the goal. Intentions Evaluation

•Specify the action sequence to

implement the intentions.
Action Interpretation
•Execute the action. specification
•Perceive the state of system resulting
from the action.
•Interpret the system state. Execution Perception
•Evaluate your interpretation against
the expectation based on your
intentions. Physical
system

Norman's seven stage model of

user activity
Copyright 2006 John Wiley & Sons, Inc
 Bridging the gaps using Norman’s
Model
Goals

Task and semantics Intentions Evaluation

Semantics and syntax Action Interpretation

specification

•Example: Summing up three numbers Execution Perception

using a calculator or a spreadsheet.
•Goal: Summing the three numbers
•Intention: enter the three numbers and Physical
receive their total system
•Sequence of action…
Figure 5.6 Norman's seven stage model of
user activity
Copyright 2006 John Wiley & Sons, Inc
Fit and Complexity
◼ The fit between the user’s mental model of
the system and the actual model of the
system affects complexity, too. The greater
the fit, the easier it is to determine how to
translate goals into action. The greater the
misfit, the more difficult and more erroneous
is the process of bridging the execution and
evaluation gulfs.

Copyright 2006 John Wiley & Sons, Inc

User Activity with Multiple Intentions
◼ Users in organizational settings usually interact with
computers to achieve complex goals.
◼ Norman’s model of user activity can be useful for
describing the user activity involved in more complex
tasks by modelling multiple intentions to accomplish a
single goal, (e.g. checking a new sales forecast and
summing the corresponding revenues).
◼ Each of these two intentions describes how the
seven stages are organized to bridge the gulfs
between the computer and the user.

Copyright 2006 John Wiley & Sons, Inc

The GOMS Model for Describing HCI

◼ GOMS: Goals, Operators, Methods, and Selection

rules are the elements of a model that describes
purposeful HCI.
◼ Goals specify what the user wants and intends to
achieve.
◼ Operators are the building blocks for describing
human-computer interaction at the concrete level.
◼ Methods are programs built with operators that are
designed to accomplish goals.
◼ Selection rules predict which method will be used. For
example, “If the mouse is working, select ‘point to an
item on screen’, if not select ‘choose OPEN option in
file menu’”.

Copyright 2006 John Wiley & Sons, Inc

 Start

Using GOMS Choose Top Level The goal can be broken down into
Goals more detailed subgoals

Draft Operators and

Methods for Each Goal

Select Methods

Check for Goal Attainment

Repeat
until
Satisfied Further
Yes Refinement
Needed?

No

Stop

Figure 5.9 A flow chart for building GOMS

Copyright 2006 John Wiley & Sons, Inc
Using GOMS for text editing
◼ Operators
◼ Mental primitives for flow of control:
◼ Accomplish the goal of <goal description>
◼ Report goal accomplished
◼ Decision: if<operator> then<operator> else<operator>
◼ Goto step <number>
◼ Memory stage and retrieval
◼ Recall that <working memory object>
◼ Retain that <working memory object>
◼ Forget that <working memory object>
◼ Retrieve LTM that <long-term-memory object>
◼ Primitive external operators
◼ Move mouse
◼ Press key <key name>
◼ Type in <string of characters>
◼ Move-cursor to <target coordinates>

Copyright 2006 John Wiley & Sons, Inc

Errors
◼ Errors: deviations from intentional behavior that is
either skill, rule or knowledge based.
◼ Classification of errors based on behavior type:
◼ Skill based behavior: automatic behavior that is
predefined and requires minimal cognitive resources.
◼ Rule based behavior: controlled behavior that relies
on predefined rules of behavior that are contingent on
particular situation encountered.
◼ Knowledge based behavior: highly controlled
behavior that requires assessment and generation of
new rules of behavior, and is demanding of cognitive
resources.
Copyright 2006 John Wiley & Sons, Inc
Causes of Errors

Copyright 2006 John Wiley & Sons, Inc

Summary
◼ Cognitive engineering focuses on developing
systems that support the cognitive processes of
users.
◼ The simplified model of HIP demonstrates how
cognitive resources such as memory and attention
are utilized.
◼ Norman’s seven-stage model of user activity is a
framework for applying the simplified model to HCI.
◼ Complexity is seen to represent the utilization of
scarce cognitive resources.

Copyright 2006 John Wiley & Sons, Inc

Summary
◼ A good fit between the user’s mental model of the
system and the actual mental model is one way of
reducing complexity.
◼ The GOMS model is also introduced (GOMS stands
for Goals, Operators, Methods, and Selection rules)
to explain in more detail how cognitive resources are
used in specific interactions.
◼ Much like Norman’s user activity model, the GOMS
model is based on levels of interaction that bridge the
gap between the abstract (psychological) task and
the concrete (physical) system.
◼ Thus, the notions of fit, complexity, and levels of
interaction are central to understanding cognitive
engineering in HCI.
Copyright 2006 John Wiley & Sons, Inc