HUMAN COMPUTER INTERFACING

1 HCI Foundations
1.1 Input–Output channels 1
1.2 Human memory 2
1.3 Individual differences 1
1.4 Text entry devices 1
1.5 Display devices 1
1.6 Devices for virtual reality and 3D interaction

1) INPUT–OUTPUT CHANNELS (HCI)

• Human interaction happens through input (receiving information) and output (sending
information).
• In computers:
o User receives output from computer
o User gives input to computer
• The terms input/output can be confusing because:
o A channel can act as both input and output
o Example: Sight → mainly input, but also output in eye-gaze systems

Human Input and Output

• Input → through senses
• Output → through effectors (body parts)
Five senses:
1. Sight
2. Hearing
3. Touch
4. Taste
5. Smell
• In HCI, important senses:
o Sight
o Hearing
o Touch
• Taste and smell → not commonly used (only special systems)
Effectors used:
• Fingers (main)
• Eyes
• Head
• Voice
• Body

Example (Using a Computer)

• Sight → see screen (menus, icons, windows)
• Hearing → system sounds (beep, voice)
• Touch → feel keyboard and mouse
• Input given by user → typing, mouse movement
1.2.1 VISION
• Vision is the main source of information
• Two stages:
1. Physical reception (eye receives light)
2. Processing & interpretation (brain understands)

Human Eye
• Light → reflected from objects → forms upside-down image on retina
• Retina converts light → electrical signals → brain
Main parts:
• Cornea & Lens → focus light
• Retina → light-sensitive
Photoreceptors:
Rods
• Very sensitive to light
• Work in low light
• Cannot detect fine detail
• Cause temporary blindness in bright light
• Located in edges (peripheral vision)
• About 120 million
Cones
• Less sensitive to light
• Detect color & detail
• 3 types → red, green, blue
• Located in fovea (center)
• About 6 million

Visual Perception
1. Size and Depth
• Object size on retina = visual angle
• Depends on:
o Object size
o Distance
Key concepts:
• Small visual angle → object not visible
• Visual acuity → ability to see fine detail
Size Constancy
• Object appears same size even if distance changes
Depth cues:
• Overlapping objects
• Size & height in view
• Familiarity

2. Brightness
• Brightness = subjective feeling
• Depends on luminance (light amount)
Important points:
• Contrast = difference between object & background
• Low light → rods active
• Normal light → cones active
Effects: Legibility factors:
• Higher luminance → better acuity • Font size: 9–12 points
• But increases flicker • Line length: 58–132 mm

Flicker: • Reading from screen is slower than paper

• < 50 Hz → visible
Contrast:
• Dark text on light background → better
• More noticeable in peripheral vision
readability
3. Color • But more flicker
• Color components:
o Hue (wavelength) 1.2.2 HEARING
o Intensity (brightness) • Hearing provides important environmental
o Saturation (whiteness) information
Facts:
Human Ear
• Humans distinguish:
Three parts:
o ~150 hues
1. Outer Ear
o ~7 million colors (with variation)
• Pinna + auditory canal
• But identify only ~10 colors without training
• Functions:
o Protect ear
Visual Processing Capabilities o Amplify sound
• Brain interprets incomplete images 2. Middle Ear
• Perception remains stable despite • Contains:
movement o Ear drum
• Size, color, brightness appear constant o Ossicles (bones)
Context helps interpretation • Amplifies sound

• Same image → different meaning depending 3. Inner Ear

• Cochlea (fluid-filled)
on context
• Hair cells convert vibrations → nerve
Illusions:
signals
• Lines may appear different lengths (but
equal)
Sound Processing
• Due to wrong interpretation
Sound properties:
Proofreading illusion • Pitch → frequency
• Brain ignores repeated words • Loudness → amplitude
• Timbre → sound type
Reading
Hearing ability:
Stages: • Range: 20 Hz – 15 kHz
1. Visual perception of words Other abilities:
2. Language decoding • Identify sound location
3. Syntax & meaning • Ignore background noise
Eye movement: Cocktail party effect
• Saccades → quick jumps • Can focus on one sound among many
• Fixations → where reading happens (94%
time) In HCI:
• Regressions → backward movement • Mostly used for:
o Alerts
Reading facts:
o Notifications
• Speed ≈ 250 words/min
• Can be used more effectively
• Words recognized by shape, not letter-by-
letter
• Capital letters reduce speed
1.2.3 TOUCH Fitts’ Law
• Provides: 𝐷
𝑀𝑇 = 𝑎 + 𝑏log⁡2 ( +1)
o Feedback 𝑆
• Movement time depends on:
o Warning (hot/cold)
o Distance (D)
• Important for interaction
o Target size (S)

Skin receptors:
1. Thermoreceptors → heat/cold Design implications:
• Large targets → easier
2. Nociceptors → pain
• Short distance → faster
3. Mechanoreceptors → pressure (important)
• Place frequent options closer
• Pie menus → equal distance but use more
Types of mechanoreceptors:
• Rapid → respond to immediate pressure
space
• Slow → respond to continuous pressure

Sensitivity (Acuity) 2) HUMAN MEMORY

• Memory helps us:
• Measured using two-point threshold
o Store information
• Fingers → highest sensitivity
o Retrieve information
• Forearm → lower sensitivity
• Used in:
o Daily activities
Kinesthesis
o Learning language
• Awareness of body position
o Performing actions
• Comes from joints
• It also gives us a sense of identity
Types:
• Rapid → movement
• Slow → movement + position
• Positional → static position

1.2.4 MOVEMENT
• Interaction involves:
1. Stimulus received
2. Brain processes Types of Memory
3. Muscles respond There are three types of memory:
1. Sensory Memory
Time components: 2. Short-Term Memory (STM) / Working
Reaction Time: Memory
• Auditory → 150 ms
3. Long-Term Memory (LTM)
• These memories interact with each other
• Visual → 200 ms
• Pain → 700 ms
• Combined signals → fastest
1.3.1 Sensory Memory
• Acts as a buffer for sensory input
• Practice → reduces time
• Exists for each sense:
• Fatigue → increases time
o Visual → Iconic memory
o Hearing → Echoic memory
Movement Time:
o Touch → Haptic memory
• Depends on:
o Age
o Fitness
Important points:
• Information lasts for a very short time
o Iconic memory → about 0.5 seconds
Accuracy vs Speed
• Continuously overwritten by new input
• Faster response may reduce accuracy
• Skilled users → fast and accurate
 Recency Effect
Examples:
• Last items are remembered better
• Seeing a moving finger in multiple
Important:
positions → iconic memory
• If another task is done → recency effect
• Hearing a question and realizing it later →
disappears
echoic memory
• Shows STM is affected by interference

Attention
Interference
• Transfers information to short-term
• STM is affected when tasks use same channel
memory
Types of channels:
• Helps focus on important stimuli only
• Visual
Reasons:
• Articulatory (verbal)
• Limited capacity of brain
• Prevents overload
Working Memory Model (Baddeley)
Factors affecting attention:
• STM has:
• Interest
o Multiple components
• Need
o Central executive
• Arousal

1.3.2 Short-Term Memory (STM)

• Works as a temporary storage (scratch-
pad)
• Used for:
o Calculations
o Reading
o Holding temporary information

Characteristics:
1. Fast access
• Around 70 ms
2. Short duration
• About 200 ms
3. Limited capacity
• Can remember 7 ± 2 items (digits)
1.3.3 Long-Term Memory (LTM)
• Called Miller’s rule
• Main storage of knowledge

Chunking
Characteristics:
• Grouping information into meaningful
1. Large capacity
units
• Almost unlimited
• Increases memory capacity
2. Slow access
Example:
• About 0.1 seconds
• Easy to remember grouped numbers
3. Long duration
• Information lasts long
Closure
• Very little decay
• Completing chunks or tasks
• If not completed → leads to errors
Storage Process
• Information moves from STM → LTM through
Pattern abstraction
rehearsal
• Using patterns to remember easily
• Example: phone numbers
Types of Long-Term Memory Meaningfulness
1. Episodic Memory • Meaningful information is easier to
• Stores events and experiences remember
2. Semantic Memory Example:
• Stores: • Objects > abstract words
o Facts • Sentences > words
o Concepts
o Skills Semantic Structuring
• Information linked to existing knowledge →
Semantic Memory Structure easier learning
Semantic Network
• Information stored as connected nodes 2. Forgetting
• Organized in classes and relationships Two main theories:
Key idea: 1. Decay Theory
• Information stored at highest level • Memory fades over time
• Helps in inference 2. Interference Theory
Retroactive interference
Frames • New information replaces old
• Structured data with slots Proactive interference
• Types of slots: • Old information interferes with new
o Fixed
o Default Other factors:
o Variable • Emotional content:
• Represent structured knowledge o Hard to remember short-term
o Easy to remember long-term
Scripts
• Represent common situations Important point:
Elements: • Forgetting may be due to retrieval failure,
• Entry conditions not actual loss
• Results
• Props 3. Retrieval
• Roles Types:
• Scenes Recall
• Tracks • Retrieve without help
Recognition
Procedural Knowledge • Identify when shown
• Knowledge of how to do things • Recognition is easier than recall
• Stored as production rules
Format: Retrieval Cues
• IF condition → THEN action • Help in remembering
Examples:
Long-Term Memory Processes • Categories
1. Storage (Remembering) • Stories
• Done through rehearsal • Patterns
• Visual imagery
Learning Improvements
Total Time Hypothesis Imagery
• More time → better learning • Visualizing helps memory
Distributed Practice Effect • People may add extra details
• Learning is better when spread over time
3) INDIVIDUAL DIFFERENCES (HCI) 4) TEXT ENTRY DEVICES (HCI)
• Earlier, humans were discussed in a general • Text entry is one of the main activities in
way computer use
• It was assumed that all people have similar • Common methods:
abilities and limitations o Keyboard
o Phone keypad
Important Idea o Handwriting recognition
• This assumption is partly true o Speech recognition
• But in reality, all users are different 2.2.1 Alphanumeric Keyboard
• Most common input device
Why Individual Differences Matter • Used for:
• Even though humans share common o Entering text
processes: o Entering commands
o Each person is not the same QWERTY Keyboard
• Designers must consider these differences • Standard keyboard layout
• Named after first six letters: Q W E R T Y
Types of Individual Differences Important points:
1. Long-term differences • Layout is fixed for letters and digits
• Sex • Non-alphanumeric keys vary (UK vs US
• Physical abilities differences)
• Intellectual abilities • Example:
o UK → £
2. Short-term differences o US → $
• Stress
• Fatigue
Why QWERTY is not optimal
• Designed for old typewriters
• Purpose:
3. Changes over time
o Prevent key jamming
• Age
• Common letters placed far apart

Design Considerations Why still used

• While designing systems: • Social reasons:
o Check if any users will be negatively o People already trained
affected o Cost of replacing keyboards
• Poor design may: o Difficult to relearn
o Exclude some users
Alternative Layouts
1. Alphabetic Keyboard
Example
• Letters arranged A–Z
• Visual interfaces:
Points:
o May exclude visually impaired users
• Looks simple
o Unless other senses are used
• Not faster for typing
• No major advantage
General Design Rule
• Systems should:
2. DVORAK Keyboard
• Rearranged keys for efficiency
o Support users who are:
▪ Under pressure
Features:
• 56% typing done by right hand
▪ Ill
• Alternate hands → faster typing
▪ Distracted
• 70% keys on home row
• Should not push users to:
• Less finger movement
o Perceptual limits
o Cognitive limits
Advantages:
• Faster (10–15% improvement)
• Less fatigue
Ergonomic Designs 2.2.4 Handwriting Recognition
• Curved keyboards • User writes → system converts to text
• Split keyboards (left & right hand) Advantages:
Purpose: • Natural and familiar
• Reduce effort • Good for:
• Improve comfort o Notes
• Reduce RSI o Sketching
• Works well in small devices
2.2.2 Chord Keyboards
• Use only 4–5 keys Disadvantages:
• Letters formed by pressing combinations • Not very accurate
• Different handwriting styles
Advantages:
• Needs stroke information
• Very compact
• Slow (~25 words/min)
• Can be used with one hand
• Fast for trained users Important point:
• Online recognition (while writing) is easier
Disadvantages:
• Not familiar
than reading written text
• Not widely used Uses:
• Mobile devices
Uses:
• Pocket organizers
• Courtroom stenographers
• Wearable devices
2.2.5 Speech Recognition
2.2.3 Phone Keypad & T9 • User speaks → system converts to text
• Mobile phones use numeric keys (0–9)

Multi-tap method Advantages:

• Press key multiple times for letters • Natural way of interaction
• Example: Problems:
o 3→d • Errors still present
o 33 → e • Works well only with limited vocabulary
o 333 → f • Difficult with natural language
Problems:
• Slow Other issues:
• Need pause or extra key for same letters • Different accents
• Need training for each user
Modes:
• Noise affects accuracy
• Number mode
• Privacy issues in offices
• Text mode
• Capital letters mode Accuracy:
• ~97% (but still not sufficient)
User behavior:
• Experienced users type fast using thumb Uses:
• Use shorthand • Telephone systems
• Disabled users
T9 Algorithm
• Hands-busy situations
• Uses dictionary
• RSI sufferers
• Press each key once
Example: Three uses of speech:
• 3926753 → “example” 1. Replace keyboard
2. Design new systems based on speech
Advantages:
3. Use where keyboard is not possible
• Faster than multi-tap
• Suggests word options
5) DISPLAY DEVICES (HCI) Jaggies (Aliasing)
• Most computer systems depend on display • Lines and curves look rough or stepped
• Happens because pixels are square
screens to show output.
• Some systems (like cars, hi-fi systems, alarms)
use non-screen outputs, but in computers, Anti-aliasing
screens are essential. • Technique to smooth edges
• Slightly blurs edges

2.4.1 Bitmap Displays – Resolution & Color Brain interprets blur as smooth line
What is a Bitmap Display?
• A screen is made of tiny dots called pixels 2.4.2 Display Technologies
arranged in rows and columns. 1. Cathode Ray Tube (CRT)
• Each pixel shows a color or brightness. How it works:
• Electron beam hits screen
Color Representation • Screen coated with phosphor
• Each pixel stores its color using bits: • Phosphor glows when hit
o 1 bit → only black or white
Scanning process:
o 8 bits → 256 colors
• Beam moves:
o 24/32 bits → almost unlimited colors
o Left → right
More bits = more color options and better image
o Top → bottom
quality
• Repeats ~30 times per second

Colormap Flicker problem:

• A colormap is the set of colors available at one • If refresh rate is low → flicker visible
time. • Reduced by:
• Even if a system supports many colors, it may o Higher refresh rate
show only a limited set at once o Interlacing
o Long-glow phosphor
Resolution Color in CRT:
Resolution has two meanings: • Uses 3 electron beams:
1. Number of pixels (screen size in pixels) o Red
• Example: o Green
o 1024 × 768 → 1024 pixels horizontally, o Blue
768 vertically Combining them produces different colors
• More pixels → sharper and clearer images
Advantages:
2. Pixel density (PPI – pixels per inch) • Cheap

• Shows how closely pixels are packed • Fast response (good for animation)

• Usually between 72–96 PPI • Good color display

Higher density = more detail in small space Disadvantages:

• Bulky and heavy
Resolution Issues • Flicker can cause eye strain
• A screen may support high resolution, but: • Color screens have lower resolution
o Computer may send lower resolution
Effects:
• CRT → stretches image → lower quality
• LCD → shows unused space → black borders
• Can cause:
o Distorted images
o Missing or uneven lines (artifacts)
2. Liquid Crystal Display (LCD) 2.4.3 Large & Situated Displays
How it works:
• Liquid crystals placed between glass layers Large Displays
• Light passes through • Used in:
• Voltage changes crystal alignment o Meetings
This controls whether pixel looks black or o Lectures
white o Advertisements

Important features: Types:

• Flat and thin 1. Plasma Screens
• Lightweight • Large flat displays
• Low power usage • Wide format (16:9)
• Less eye strain
2. Video Walls
Advantages:
• Many screens combined
• Portable
• Show one big image
• Comfortable to use
• No radiation
3. Projectors
• Project image onto screen
Impact:
Types:
• Made laptops and portable devices possible
• RGB projectors
• Changed how people work (e.g., working
• LCD projectors
anywhere)
Problem:
Other Display Technologies
• Presenter’s shadow blocks screen

1. Gas Plasma Displays

Solution: Back Projection
• Used for large screens
• Projector placed behind screen
• Similar to normal displays but bigger

Situated Displays
2. Random Scan (Vector Display)
• Displays placed in specific locations
How it works:
• Meaning depends on location
• Draws lines directly (not pixel grid)
Examples:
• Public displays
Advantages:
• Electronic notice boards
• Smooth lines (no jaggies)
• High resolution
2.4.4 Digital Paper
Disadvantages:
• Expensive What is Digital Paper?
• Poor color quality • Flexible display
• Causes fatigue • Can store image without power

3. Direct View Storage Tube Technologies:

• Used in oscilloscopes 1. Rotating spheres
• Each pixel has a tiny ball
Features:
• Black on one side, white on other
• No flicker
• Rotates to show color
• Stable image

Limitations: 2. Liquid tube method

• Cannot erase part of image • Reflective sphere moves inside liquid
• Low brightness and contrast • Shows black or white
• Poor color support
Key Feature: 2. 3D Mouse
• Image stays even when power is off • Advanced version of normal mouse

Uses: Features:
• Reusable banners • Can:

• Future applications: o Move in 3D space

o Rotate
o Digital books
o Tilt
o Rewritable pages

6) DEVICES FOR VIRTUAL REALITY AND 3D Tracks:

INTERACTION • Position (3D)
• Rotation:
• Virtual Reality (VR) systems allow users
o Pitch
to:
o Yaw
o Move in a 3D environment
o Interact with 3D objects o Roll

• Sometimes normal devices (mouse, Technologies used:

keyboard) are used • Magnetic sensors
• But special 3D devices are also used for: • Ultrasound
o Movement • Mechanical joints
o Interaction
o Viewing Problem:
• User must hold it in air → causes strain
2.5.1 Positioning in 3D Space
• Putting it down may trigger actions
Basic Idea
• In 3D environments, users must:
o Move to a location
3. Dataglove
• Special glove used in VR
o Change direction/orientation
o Manipulate objects
How it works:
This is more complex than 2D interaction
• Optical fibers detect finger movement
Degrees of Freedom • More bending → less light → detected
• Normal mouse → 2 degrees (x, y)
• 3D interaction → 6 degrees: Measures:
o 3 for position (x, y, z) • Finger joint angles
o 3 for rotation: • Hand position
▪ Pitch (up/down tilt) • Wrist rotation
▪ Yaw (left/right turn)
▪ Roll (twisting) Features:
1. Cockpit and Virtual Controls • Very powerful input device
• Based on aircraft/helicopter controls • Captures many movements
Use:
• Games Advantages:
• Simulations • Natural to use
• Highly expressive
Important points:
• Provides detailed input
• Requires high skill and practice
• In PC systems:
Disadvantages:
o Controls are often virtual (on
• Very expensive
screen)
• Hard to use with keyboard
o User uses mouse to control them
This creates two levels of control Applications:
(indirection): • Gesture recognition

1. Mouse → control • Sign language

2. Control → movement • Assisting disabled users

4. Virtual Reality Helmets 2. Shutter Glasses
• Worn on head • Eyes alternately blocked
• Synced with screen
Functions:
1. Display 3D images 3. Polarized Glasses
2. Track head movement • Different light filters for each eye

Use: 4. Red-Green Glasses

• Basic method
• When head moves → system updates view
• Lower quality
Extra use:
Limitations of 3D Displays
• Head direction used for:
• Devices are:
o Navigation
o Bulky
o Object selection
o Heavy
5. Whole-Body Tracking • Viewing angles limited
• Tracks full body movement • Focus problems (eyes vs display mismatch)

Examples: Advanced Idea

• Adjust display based on eye focus
• Leaning on bike simulator
• Still under development
• Standing on trampoline-like device
VR Motion Sickness
Advanced method: • Happens when:
• Cameras track body joints o Eye sees one thing
• White markers placed on body o Body senses another

Purpose: Cause:
• Create immersive experience • Delay between:
• Make user feel inside virtual world o Head movement
o Screen update
2.5.2 3D Displays
Effect:
Desktop VR • Confusion in brain
• Uses normal screen • Nausea (like sea sickness)
• Creates 3D effect using:
Important point:
o Shadows
• Delay > 100 ms → problem increases
o Perspective
o Occlusion
Simulators and VR Caves
Seeing in 3D Simulators
• Human depth perception uses many cues • Example:
o Flight simulators
Stereoscopic Vision
• Each eye sees slightly different image Features:
• Brain combines them → depth perception • User inside physical setup
• Screens show virtual world
Creating Stereo in VR
Step 1: VR Caves
• Generate two images (one for each eye) • Room with screens all around

Features:
Step 2: Show separately • User surrounded by display
Methods: • Can look in all directions

1. VR Goggles Advantage:
• Two small screens • More immersive than headset
• One image per eye
------------------------------------------------- -------------------------------
 2 Designing Interaction
2.1 Models of Interaction, Frameworks and HCI 1
2.2 Interaction Styles 1
2.3 Contexts of Interaction 1
2.4 The Process of Design, Navigation Design 1
2.5 Screen Design and Layout 1
2.6 Iteration and Prototyping

MODELS OF INTERACTION (HCI) Task Analysis

• Study of:
• Interaction involves two main participants:
o Goals
o User
o Tasks
o System
o Intentions
• Both are:
• Helps in system design
o Complex
o Different in thinking and
communication User vs System Language
• User language → Task language
So, the interface acts as a translator between
o Based on human thinking
them
• System language → Core language
o Based on computer operations
Why Models are Needed
• Interaction may fail at many points
Important Point
• Models help to:
• Even in multi-user systems:
o Understand interaction clearly
o Interaction is viewed from one
o Find problems
user’s perspective
o Compare different interaction styles

3.2.2 Norman’s Execution–Evaluation Cycle

3.2.1 Terms of Interaction
• Most important interaction model
Domain
• Area of knowledge or activity Basic Idea
• Examples: • User:
o Graphic design 1. Plans action
o Authoring 2. Executes it
o Factory control 3. Checks result
Concepts
Two Main Phases
• Important elements of a domain
1. Execution Phase
• Example (graphic design):
• User decides and performs action
o Shapes
2. Evaluation Phase
o Canvas
• User checks system response
o Tools

Task 7 Stages of Interaction

• Action performed on concepts 1. Establish goal
Goal 2. Form intention
• Desired result of a task 3. Specify actions
• Example: 4. Execute actions
o Create a red triangle 5. Perceive system state
Intention 6. Interpret system state
• Specific action to achieve goal 7. Evaluate result
Explanation (Simple Flow) Four Steps (Translations)
• User decides what to do → performs action
→ sees result → checks if goal achieved 1. Articulation (User → Input)
• User expresses task using input device
Example: Switching on Light Depends on how easy it is to express actions
• Goal → need more light
• Intention → switch on lamp 2. Performance (Input → System)
• Action → press switch • Input converted into system operations
• Result → light ON or OFF Determines what system states can be
• Evaluation: reached
o If ON → success
o If OFF → new goal 3. Presentation (System → Output)
• System shows results via output
Gulfs in Interaction Depends on how well system information is
1. Gulf of Execution displayed
• Gap between:
o What user wants to do 4. Observation (Output → User)
o What system allows • User interprets output
Smaller gap = easier interaction Depends on how easily user understands
result
2. Gulf of Evaluation
• Gap between: Examples
o System output Poor Mapping
o User understanding • Light switches not matching room layout
Smaller gap = easier to understand results → difficult to use

Limitation of Norman’s Model Good Mapping

• Focuses only on user side • VR datagloves
• Does not fully describe system → natural interaction
communication
Other Points
3.2.3 Interaction Framework (Abowd & Beale) • Some devices limit actions:
o Example: remote control cannot
Main Idea
turn off system
• Extends Norman’s model
• Includes both:
• Output limitations:
o User
o Cannot show all information at once
o System

Four Components • Observation difficulty:

1. User o Hard to interpret:
2. System ▪ Analog clock without
3. Input markings
4. Output ▪ Command line results
Input + Output = Interface
Assessing Interaction
Languages Used • Interaction quality depends on:
• User → Task language o Task being performed
• System → Core language No single system is best for all tasks
• Input → Input language
• Output → Output language
 Covers:
• Input devices
• Output devices
• User’s physical environment
Concerned with:
• Comfort
• Safety
• Efficiency

2. Dialog Design & Interface Styles

• Related to input side of framework

Focus:
• How user gives commands
• How interaction is structured

Covers:
FRAMEWORKS AND HCI • Articulation
• Frameworks are used to: • Performance
o Explain interaction details Mostly associated with system side, even though
o Discuss other related aspects of it involves user input
HCI
They help us understand different areas 3. Presentation & Screen Design
connected to interaction • Related to output side of framework

ACM SIGCHI Framework Focus:

• A framework is proposed by ACM SIGCHI • How system shows information
Curriculum Development Group Includes:
• It is similar to the interaction framework • Layout
Used to organize different HCI-related areas • Display of data

4. Social & Organizational Context

• Surrounds the entire framework

Includes:
• Work environment
• Social factors
Affects:
• How interaction happens
• How systems are used

Important Point
• All these areas:
o Affect system design
Relation to Interaction Framework o Affect user performance
• Different HCI areas are mapped onto parts
of the interaction framework
1. Ergonomics
• Focuses on:
o User side of interaction
INTERACTION STYLES (HCI) Advantages:
• Interaction = dialog between user and • Easy to use
computer • Uses recognition (not memory)
• The interface style affects how this dialog
Disadvantages:
happens • Needs good grouping and naming
Different styles change: • Hierarchical menus can be confusing
• Ease of use
• Speed Types:
• Learning • Text-based menus
Common Interaction Styles • Graphical menus
1. Command line interface
2. Menus 3.5.3 Natural Language
3. Natural language • Interaction using normal human language
4. Question/answer & query dialog
5. Form-fills & spreadsheets Advantages:
• Easy and natural for users
6. WIMP
7. Point-and-click
8. 3D interfaces Problems:
1. Ambiguity in structure
3.5.1 Command Line Interface (CLI) • Example:
• User types commands to interact with o Sentence meaning unclear
system
2. Ambiguity in words
Features:
• Same word → different meanings
• Uses:
o Commands
3. Context dependency
o Abbreviations
• Requires background knowledge
o Function keys
Advantages: 3.5.4 Question/Answer & Query Dialog
• Direct access to system
Question/Answer
• Very powerful
• System asks questions
• Flexible (can use parameters)
• User responds step by step
• Good for repetitive tasks
Advantages:
Disadvantages:
• Easy to learn
• Hard to learn
• Good for beginners
• Requires memory (no visual help)
• Commands may be confusing Disadvantages:
• Limited flexibility
3.5.2 Menus
• Less powerful
• List of options shown on screen
Features: Query Language
• User selects option using: • Used to search databases

o Mouse Features:
o Keyboard • Requires:
Advantages: o Specific syntax
• Easy to use o Knowledge of database
• Uses recognition (not memory)

Disadvantages: Problems:
• Needs good grouping and naming • Complex when multiple conditions used

• Hierarchical menus can be confusing • No direct confirmation

3.5.5 Form-Fills & Spreadsheets Features:
• Simple
Form-Fills
• Often used in:
• Screen looks like a form
o Web
• User fills fields
o Multimedia
Advantages:
Important point:
• Familiar
• Not limited to mouse
• Easy to use
• Also used in touchscreens
Features:
3.5.8 Three-Dimensional Interfaces
• Allows:
o Navigation Types:
o Editing 1. 3D appearance
o Blank fields • Buttons look raised

Use: Helps identify clickable items

• Data entry 2. 3D workspace
• Objects shown with:
Spreadsheets
o Perspective
• Grid of cells o Depth
Features: • Objects far away appear smaller
• Cells contain: 3. Full VR systems
o Values • User moves in virtual world
o Formulas
Advantages:
Advantages: • Natural interaction
• Flexible • Uses real-world experience
• Updates automatically
• User can experiment with values
Important Idea:
• 3D interfaces make users feel like:
o Interacting, not just viewing
Input and output are combined
THE CONTEXT OF THE INTERACTION (HCI)
3.5.6 WIMP Interface
Meaning: • Interaction between user and system does
• Windows not happen alone
• Icons • It occurs within a context (environment +
• Menus surroundings)
• Pointers This context affects:
• User behavior
Examples:
• Performance
• Microsoft Windows
• Motivation
• MacOS

Features: Types of Context

• Most common interface 1. Physical Context
• Graphical and interactive • Environment around the user
• Example:
3.5.7 Point-and-Click Interface o Workplace conditions
• User clicks on objects
2. Social & Organizational Context
• Users work in:
Examples:
o Groups
• Clicking links
o Organizations
• Clicking buttons
This has a strong influence on interaction
Influence of Social Environment Possible Outcomes
1. System is rejected
Presence of Other People 2. Users become unmotivated
Effects on performance: 3. Users adapt system in their own way
For known tasks: Important Point
• Competition → improves performance
Users must be involved in design process
• Desire to impress others → improves
performance Positive Effects of Technology
• New technology can:
For new tasks: o Motivate users
• Fear of failure → reduces performance o Improve work
If it is:
Importance of Privacy • Well designed
• Users need privacy to: • Matches user needs
o Learn • Integrated with work
o Experiment Importance of Feedback
• Without privacy → hesitation and mistakes
Why feedback is needed
• Keeps user informed
Motivation of Users
• Prevents confusion

Sources of Motivation Without feedback

• Fear • User may feel:
• Allegiance o Bored
• Ambition o Uncertain
• Self-satisfaction o Frustrated

Good system design

Job Satisfaction
• Every action should have:
• Depends on:
o Clear result
o How well user performs tasks
o Ease of system use Handling delays
• Feedback should show:
If system is: o System is working
• Difficult Prevents frustration
• Frustrating
Then: HCI IN DESIGN PROCESS
• Job satisfaction decreases In many cases, HCI experts are involved too late
• Performance decreases • Systems are:
o Designed and built first
Mismatch Between System and Work o Checked for usability later
This causes problems
Problem
• Systems often chosen by managers
 Best companies:
• May not match actual user needs
• Consider usability from the beginning
 Usability should be:
Reasons • Designed in, not just tested later
• Managers focus on:
o Results MAIN PHASES OF DESIGN PROCESS
• Not on: There are four main phases + iteration loop
o Actual work process
1. Requirements (What is needed)
Purpose:
• Understand what users want and need
 4. Iteration & Prototyping
Important idea:
• Designs are not perfect at first

Iteration:
• Repeated improvement

Prototyping:
• Build early versions of system
Steps: Purpose:
• Study current situation: • Test with real users
o How people work • Get feedback
o What tools they use
Evaluation:
Techniques: • Identify:
• Interviews o Problems
• Observation o Improvements
• Videotaping
5. Implementation & Deployment
• Studying documents
Purpose:
Important method: • Build final system
• Ethnography
Includes:
o Observing users in real environment
• Coding
2. Analysis • Hardware creation
Purpose: • Documentation
• Organize collected data
Result:
Goal: • System ready for users
• Identify:
Important Practical Points
o Key issues , User tasks
Time vs Quality Trade-off
Techniques:
• Limited time available
Task Models
• Cannot make system perfect
• Show how users perform tasks
Must balance:
Scenarios • Quality
• Stories describing interaction • Cost
Used to: • Time
• Represent current system
Reality of Design
• Imagine future system
• Not possible to fix all problems
3. Design • Important question:
Purpose: Which problems are worth fixing?
• Decide how system will work
Key Insight
Includes: • Finding problems → easy
• Rules • Fixing problems → possible
• Guidelines • Choosing which to fix → difficult
• Design principles
Activities: Final Thought
• A “perfect” system may mean:
• Plan navigation
o Too much time spent
• Design screen layout
o Poor planning
Support from theory: Better to have:
• Cognitive models
• Acceptable system
• Organizational issues
• Delivered on time
• Communication understanding
NAVIGATION DESIGN (HCI) Four Important Questions for Each Screen
• Design is not just about the system, but the 1. Knowing where you are
whole socio-technical environment • System should show current position
• However, at this stage we focus on the • Example:
system itself o Breadcrumbs

Levels of Interaction 2. Knowing what you can do

When using a system (e.g., word processor), • Clear actions:
interaction happens at multiple levels: o Buttons
o Links
1. Widgets
Avoid confusion:
• Buttons, menus, icons
• Decoration vs clickable items
• Help user understand:
o What action to perform 3. Knowing where you are going
• User should know:
2. Screens / Windows
o What will happen next
• User must:
o Find information
Avoid:
• Trial and error
o Understand grouping of elements

3. Navigation within Application 4. Knowing where you’ve been

• Helps avoid confusion
• User should:
• Example:
o Know what will happen after an
o Browser history
action
o Back button
o Understand current position

4. Environment Important Issues

• Interaction with: Icons
o Files • Not always clear
o Networks • Should include:
o Other applications o Labels
o Tooltips
Starting Point for Navigation Design
Before designing structure, ask: Modes
• Who will use the system? • Same action → different meaning in
• How do they think? different contexts
• What will they do? Problem:
This helps design user-centered structure • Can confuse users

Types of Structure Solution:

• Make modes clearly visible
1. Local Structure
• Focus on single screen/page Feedback
• System must confirm:
2. Global Structure
o What user has done
• Focus on:
o Entire system Helps:
o Movement between screens • Avoid errors
• Build confidence
5.6.1 Local Structure
Lost in Hyperspace
Goal-Seeking Behavior • User feels:
• Users:
o Disoriented
o Have partial knowledge
• Happens when:
o Move step by step toward goal
o No clear navigation
Not always shortest path, but should feel:
• “I am getting closer to my goal”
5.6.2 Global Structure – Hierarchy Representation
• Organizing system into levels
Network Diagram
• Shows:
Types of grouping: o Screens
• Function-based o Transitions
• User-based
• Module-based
Can show:
• What leads to what
Advantages: • Sequence of actions
• Easier to understand • Multiple paths
• Logical structure
5.6.4 Wider Context
Problems:
• Different users think differently
System is not isolated
• Works with:
• Wrong grouping → confusion
o Other applications
o Devices
Important Rule
Structure should match user thinking, not Implications
designer thinking 1. Style Issues
Hierarchy Depth • Follow platform standards
• Deep hierarchy → difficult navigation • Example:
o Standard icons
Better approach:
• Broad categories 2. Functional Issues
• Show multiple levels • Support:
o File handling
Menu Breadth
o Data formats
• Many options can be okay if:
o Copy-paste
o Well organized
• Not limited strictly to 7±2 3. Navigation Issues
• Support:
5.6.3 Global Structure – Dialog o Links between systems
o Opening files in correct apps
• Flow of interaction between user and system
Web-specific Issues
Important Idea
• Users may:
• Interaction is not always hierarchical
o Enter from any page
• Includes:
• External links:
o Sequences
o Not under control
o Steps
o Can change anytime
o Actions

Example
• Moving between:
o Items
o Orders
o Records

Dialog Characteristics
• Has:
o Flow
o Branches
o Loops
 3. Decoration
• Use:
o Boxes
o Lines
o Colors
o Fonts
Purpose:
• Show grouping
• Improve clarity

SCREEN DESIGN AND LAYOUT (HCI) 4. Alignment

• A screen must: Text
o Display information clearly • Left aligned
o Allow user interaction
Numbers
• It is complex because it involves:
• Right aligned or decimal aligned
o Psychological principles
Shows magnitude clearly
o Graphical design
Names
Basic Design Principles
• Better in:
1. Ask o Columns
What is the user doing? o Or reversed (surname first)

2. Think Multiple Columns Problem

What information is needed? • Large gaps → hard to read
What comparisons will the user make? Solutions
In what order is information needed? • Dotted lines (leaders)
3. Design • Background shading
• Adjust alignment
Form follows function
Layout should depend on user tasks 5. White Space (Very Important)
• Space between elements helps:
5.7.1 Tools for Layout
o Separation
1. Grouping and Structure o Grouping
• Related items should be physically grouped o Highlighting
• Helps users:
Uses of Space
o Understand relationships
1. Separate sections
o Navigate easily
2. Create structure
Example Structure 3. Highlight important items
• Administrative info
Design Tip
• Billing details
• Delivery details
Blur your view → check layout balance
• Order details
5.7.2 User Action and Control
Logical grouping → easy understanding
1. Entering Information
2. Order of Groups and Items
• Forms & dialog boxes:
• Screen should follow natural user order
o Show info
Example:
o Allow input
• Fill billing → delivery → order items

Important Points Design Rules

• Match user thinking • Align fields properly
• Set tab order correctly • Avoid jagged layouts
• Sometimes force order (e.g., payment first)
Better Solutions Key Principle:
• Right-align labels Match presentation to purpose
• Place labels above fields
Example:
2. Logical Layout • Alphabetical list → find file name
• Follow: • Date order → find recent files
o Left → Right
2. Multiple Representations
o Top → Bottom
• Interactive systems allow:
Matches reading behavior o Sorting
3. Knowing What to Do o Changing views

Problem: Helps different user needs

• Hard to identify: 3. Aesthetics vs Utility
o Active elements
Important Idea:
o Their function
Good-looking ≠ usable
Solution:
Advantages of Aesthetics:
• Use:
• Improves satisfaction
o Standard buttons
• Increases productivity
o Clear labels
o Consistent design Problems:
• Too much design → confusion
Example Problem:
• Example:
Button “Bold”
o Fancy background → unreadable
• Is it current state OR action?
text
4. Affordances
Trade-off:
Definition: • Beauty vs usability
Objects suggest how to use them
Examples: 4. Color Usage
• Button → push Problems:
• Handle → pull • Overuse → distraction
• Color blindness issues
In Interfaces:
• Icons → clickable Rule:
• Buttons → press Use color carefully
Important Point: Support, not replace information
Affordances depend on user experience 5. 3D Effects
• Experts → double click
Advantages:
• Beginners → don’t know
• Good for:
o Objects
5.7.3 Appropriate Appearance
o Graphs

1. Presenting Information Problems:

• Text becomes hard to read
Depends on: • 3D charts confusing
• Type of data:
o Text 6. Localization / Internationalization
o Numbers Definition:
o Tables
Adapting software for different languages &
• Technology
cultures
• Purpose
Techniques: Evaluation of Prototypes
• Use resource identifiers
1. Formative Evaluation
• Separate text from code
• Done during design
Challenges: • Purpose:

1. Language Direction Improve the system

• Left → Right 2. Summative Evaluation
• Right → Left • Done at the end

2. Cultural Differences • Purpose:

• Symbols differ Check if system is good enough

Example: Methods of Evaluation
• Tick ✔ vs Cross ✖
1. Expert Evaluation
• Meaning varies by culture
• Experts check using guidelines
Example:
ITERATION AND PROTOTYPING (HCI)
• “Knowing where you are”
• Human situations are complex • “Knowing what you can do”
• Designers are not perfect
2. User Evaluation
So, first design is usually not perfect • Real users test the system:
Therefore, iteration is needed o In controlled environment
What is Iteration? o In real-life environment
• Iteration means: Iteration Cycle
Repeating design → evaluation → Steps:
improvement 1. Design
• It starts early and continues until: 2. Prototype
Design becomes acceptable 3. Evaluate
What is Prototyping? 4. Find problems
• Prototype = early version of design 5. Redesign
Repeat until:
Types of Prototypes: No more problems can be fixed economically
1. Paper prototypes
o Sketches Important Concept: Hill-Climbing Approach
o Storyboards
2. Software prototypes What is Hill Climbing?
o Tools like: • Improve step by step
▪ Shockwave • Always move towards “better” design
▪ Visual Basic
3. Physical mockups Example Idea
o Device models • Start somewhere
• Keep improving
Why Prototyping is Used • Reach the best nearby solution
• To:
o Test ideas Problem: Local Maximum
o Find problems
Explanation:
o Improve design
• You may reach:
Best solution in that area
BUT
Not the best overall solution
 3 Cognitive Models
3.1 Goal and Task hierarchy 1
3.2 GOMS, Linguistic Models 1
3.3 Physical and device Models 1
3.4 Cognitive architectures 1
3.5 Communication and collaboration models 1
3.6 Face to face communication, Text Based communication 1
3.7 Group working 1
3.8 Task Analysis, Task decomposition. 2

GOAL AND TASK HIERARCHIES

­ In Human–Computer Interaction, many models assume that users achieve their goals by breaking
them down into smaller and manageable parts.
­ This process follows a divide-and-conquer approach, where a complex task is divided into subgoals,
and each subgoal is further divided until the task becomes simple enough to perform.
­ This results in a hierarchical structure of goals and subgoals, which helps in understanding how
users interact with systems.

Concept of Goal Hierarchy

A goal hierarchy represents the way in which a user organizes tasks mentally. Instead of solving a problem
all at once, the user divides it into smaller parts and solves each part step by step.
For example, if a user wants to produce a report on sales of HCI textbooks, this main goal is divided into
several subgoals such as:
• Gathering the required data
• Creating tables and histograms
• Writing descriptive content
Each of these subgoals can be further divided. For instance, gathering data may involve:
• Finding the names of textbooks
• Searching a database
• Filtering and organizing results

This process continues until the tasks become simple and routine. At the end, we get a structured
hierarchy of goals, where higher-level goals depend on the completion of lower-level subgoals.

Issues in Goal Hierarchies

While creating goal hierarchies, several important issues arise.

1. Granularity (Where to Start and Stop)

­ One major issue is deciding how detailed the task breakdown should be.
­ Tasks can be decomposed to a very low level, such as individual hand or eye movements, or they
can remain at a higher level.
­ Similarly, the starting point of analysis can vary from small tasks to very large goals like “staying
alive.”
­ This concept is called granularity, and it depends on the purpose of the design.
­ Different design problems require different levels of detail.
2. Unit Task
­ Most models focus on a level called the unit task.
­ A unit task is a task that does not require problem-solving by the user because it is already learned
and routine.
­ For example, pressing a key or selecting a menu option is a unit task.
­ Although it is simple for the user, identifying and designing such tasks can be complex for the
designer.

3. Multiple Methods
­ Users often have more than one way to achieve the same goal.
­ Therefore, a goal hierarchy must include different possible methods and a way to choose between
them. This makes the system flexible and closer to real user behavior.

4. Error Handling
­ Another important issue is that users are not perfect and may make mistakes.
­ Goal hierarchies usually describe how an ideal user performs a task, but they do not effectively
predict or handle errors.
­ This is a limitation of many hierarchical models.

GOMS MODEL
The GOMS model, developed by Card, Moran, and Newell, is one of the most important models used to
describe user interaction. GOMS stands for:
• Goals
• Operators
• Methods
• Selection

Components of GOMS
1. Goals
­ Goals represent what the user wants to achieve.
­ They act as a reference point in memory, allowing the user to evaluate progress and return to the
task if errors occur.

2. Operators
­ Operators are the smallest units of action that a user performs.
­ These include both physical actions, such as pressing a key, and mental actions, such as reading or
thinking.

3. Methods
­ Methods describe the different ways in which a goal can be achieved.
­ For example, a user may close a window using a menu option or by pressing a keyboard shortcut.
­ Both are valid methods for achieving the same goal.

4. Selection
­ Selection rules determine which method is chosen.
­ These rules depend on factors such as user preferences, experience, and system state.
­ For instance, a user may prefer using the mouse in general but switch to keyboard shortcuts in
specific situations.
Characteristics of GOMS
The GOMS model focuses mainly on expert users performing routine tasks. It provides a structured way
to analyze user actions and can predict performance, such as:
• The time taken to complete tasks
• The memory required for tasks
However, GOMS does not deal well with learning processes or complex problem-solving tasks.

COGNITIVE COMPLEXITY THEORY (CCT)

­ Cognitive Complexity Theory (CCT), developed by Kieras and Polson, extends the GOMS model by
adding more detail and predictive power.
­ It provides a more realistic description of user behavior.
­ CCT uses production rules to represent user behavior. These rules follow the format:
IF condition THEN action
The condition checks the contents of working memory, and if it is true, the action is performed.

Working Mechanism
­ In CCT, multiple rules are active at the same time.
­ At any moment, any rule whose condition is satisfied can execute.
­ This makes the system flexible and capable of representing complex behavior.

Example
Consider a task where a user needs to insert a missing space in a text. The process involves:
• Identifying the mistake
• Moving the cursor to the correct position
• Inserting the space
• Exiting the insertion mode
Each of these steps is represented by production rules. The system continuously checks conditions and
executes actions accordingly.

Key Features of CCT

1. Proceduralization
­ In expert users, repeated actions are stored as a single chunk.
­ For example, inserting a space becomes a single action rather than multiple steps.

2. Novice vs Expert Behavior

­ CCT distinguishes between novice and expert users.
­ Experts perform tasks quickly using chunks, while novices perform each step separately and check
results frequently.

3. Error Representation
­ CCT can represent errors in the system, such as performing actions in the wrong mode.
­ However, it cannot predict when errors will occur.

PROBLEMS AND EXTENSIONS OF GOAL HIERARCHIES

1. Artificial Hierarchies
­ Sometimes, goal hierarchies are influenced by system design rather than actual user behavior.
­ This can lead to unnatural task structures.
2. Closure Problem (Important Example)
A well-known example of the closure problem is seen in early ATM systems.
The user’s main goal is to get money, and once this goal is achieved, the user may forget remaining steps
such as collecting the card.
Example Sequence
• Insert card
• Enter PIN
• Enter amount
• Collect money
• Collect card

Problem
Users often take the money and leave the card behind because they feel their main goal is completed.
Solution
Banks changed the system so that the card is returned before dispensing money. This ensures that all
subgoals are completed before the main goal is achieved.
General Rule
A higher-level goal should not be considered complete until all its subgoals are completed. However, it is
difficult to predict exactly when a user feels that a goal has been achieved.

LINGUISTIC MODELS
­ In Human–Computer Interaction (HCI), the interaction between the user and the computer is often
viewed as a form of communication similar to a language.
­ Because of this, several modeling techniques have been developed based on linguistic concepts.
­ These models aim to describe how users interact with systems using structured rules, similar to
grammar in natural languages.
­ Linguistic models are not only used to describe the structure of interaction but also to analyze user
behavior and measure the cognitive difficulty involved in using an interface.
­ Many dialog design notations are based on these linguistic principles, and one of the most
common methods used is Backus–Naur Form (BNF).

1. BACKUS–NAUR FORM (BNF)

­ Backus–Naur Form (BNF) is a formal notation used to describe the syntax of a language.
­ In HCI, it is used to represent the structure of user interaction with a system, focusing only on the
sequence of actions and not on their meaning.

Key Characteristics
BNF describes interaction at a purely syntactic level, which means:
• It focuses only on how actions are structured
• It ignores the meaning (semantics) behind those actions
• It represents interaction as a set of rules similar to programming languages

Basic Structure of BNF

BNF consists of rules of the form:
name ::= expression
• The symbol ::= means “is defined as”
• The left side contains a non-terminal (abstract concept)
• The right side contains terminals and non-terminals
Types of Elements
1. Non-terminals
Non-terminals represent higher-level actions or concepts.
They are written in lowercase.
Example:
• draw-line
• choose-points

2. Terminals
Terminals represent actual user actions.
They are written in uppercase.
Example:
• CLICK-MOUSE
• DOUBLE-CLICK-MOUSE

Operators in BNF
1. Sequence (+)
This indicates that actions occur in order.
Example:
select-line + choose-points

2. Choice (|)
This indicates alternatives.
Example:
choose-one | choose-one + choose-points

Example Explanation (Line Drawing Task)

A graphics system allows users to draw a polyline. The steps include:
• Selecting the line tool
• Choosing points
• Indicating the last point
The BNF rules describe how these actions are structured step by step.

Recursive Rules
BNF allows recursion, meaning a rule can refer to itself.
Example:
position-mouse ::= MOVE-MOUSE + position-mouse
This means the user can move the mouse any number of times.

Measures of Complexity in BNF

The complexity of an interface can be measured using:
1. Number of rules
o More rules indicate a more complex interface
2. Number of operators (+ and |)
o More operators indicate more complex structure

Limitations of BNF
• It does not consider meaning (semantics)
• It ignores user perception of system responses
• It focuses mainly on input actions rather than feedback
2. TASK–ACTION GRAMMAR (TAG)
Task–Action Grammar (TAG) was developed to overcome the limitations of BNF. It provides a more
cognitive approach by considering:
• Consistency in commands
• User knowledge
• Relationships between actions

Purpose of TAG
TAG aims to better represent how users actually think and learn interfaces by:
• Highlighting patterns and regularities
• Reducing learning effort
• Representing real-world knowledge

Handling Consistency
One of the main advantages of TAG is that it can represent consistent command structures.
Example: UNIX commands
• cp (copy)
• mv (move)
• ln (link)
All follow a similar structure, and TAG captures this consistency using parameters.

Parameterized Rules
TAG uses parameters to represent similarities.
Example:
file-op[Op] := command[Op] + filename + filename
This shows that different operations follow the same pattern.

World Knowledge
TAG includes the concept of world knowledge, meaning:
• Users already know certain things
• They do not need to learn them again
Example:
Commands like FORWARD, BACKWARD, LEFT, RIGHT are easier to learn because they match real-world
understanding.

Known-item Concept
TAG uses a special form called known-item to represent knowledge already known by the user.
Example:
known-item[Type=word, Direction]
These rules are:
• Included in the description
• Not counted in complexity
This makes TAG more realistic than BNF.
Congruence
Congruence refers to how well commands are logically related.
Example:
• NEXT ↔ PREVIOUS
• UP ↔ DOWN
Consistent pairs are easier to learn than mixed or unrelated commands.
TAG represents this using feature sets like:
F(‘next’)
Advantages of TAG
• Captures consistency in commands
• Includes user knowledge
• Reflects real learning behavior
• Provides better prediction of usability

Limitations of TAG
• Depends on designer’s assumptions about user knowledge
• Requires careful judgment
• May vary depending on user background and language

Aspect BNF TAG

Focus Syntax only Cognitive + syntax
User knowledge Not considered Included
Consistency Not handled well Clearly represented
Complexity measure Based on rules More realistic
Learning prediction Weak Strong

PHYSICAL AND DEVICE MODELS

­ Physical and device models in HCI focus on how users physically interact with systems using input
devices.
­ Unlike cognitive models that study thinking and problem-solving, these models deal with motor
actions, device behavior, and performance measurement.
Two important models are:
• Keystroke-Level Model (KLM)
• Three-State Model

12.5.1 KEYSTROKE-LEVEL MODEL (KLM)

­ The Keystroke-Level Model (KLM) is a model used to predict how long it takes a user to perform a
task using a computer system.
­ Unlike complex cognitive models, KLM focuses on simple, routine tasks that involve physical
actions.
It is mainly applied to unit tasks, which are small actions that usually take less than 20 seconds, such as:
• Searching and replacing text
• Changing font style
• Editing a word
It does not deal with complex tasks like designing diagrams, because those must first be broken into
smaller subtasks.

Two Phases of Task Execution

KLM divides a task into two phases:
1. Acquisition Phase
In this phase, the user:
• Understands the task
• Decides how to perform it
• Forms a mental plan
KLM does not analyze this phase.
2. Execution Phase
In this phase, the user:
• Performs physical actions
• Uses keyboard, mouse, etc.
KLM focuses only on this phase and assumes the user is already an expert.
Operators in KLM
KLM breaks user actions into seven types of operators:
1. K – Keystroking
This includes:
• Pressing keys
• Typing characters
• Using modifier keys (Shift, Ctrl)

2. B – Button Press
This refers to:
• Clicking a mouse button

3. P – Pointing
This involves:
• Moving the cursor to a target
• Depends on distance and size of the target

4. H – Homing
This refers to:
• Moving hands between devices
• Example: keyboard → mouse

5. D – Drawing
This includes:
• Drawing shapes or lines using a mouse

6. M – Mental Preparation
This represents:
• Thinking before an action
• Small pauses while recalling steps

7. R – System Response
This includes:
• Time taken by the system to respond
It is ignored if the user does not need to wait.
Example Explanation
Suppose a user corrects a typing mistake:
Steps include:
• Move hand to mouse
• Point and click
• Return to keyboard
• Delete character
• Type correct character
Each of these is mapped to operators like:
• H (move hand)
• P (point)
• B (click)
• K (type)
• M (mental preparation)
Time Calculation
Total execution time is calculated using:
T = TK + TB + TP + TH + TD + TM + TR
Each operator has a standard time value based on:
• User skill
• Device type
• Task complexity

Mental Operator (M) – Important Concept

The mental operator is based on chunking:
• If the user performs a familiar action (like typing a word), only one mental step is needed
• If the action is unfamiliar, more mental steps are required
This depends on user expertise.
Advantages of KLM
• Provides quantitative predictions of task time
• Helps compare different systems or methods
• Useful for designing faster interfaces
• Does not require building the system first

Limitations of KLM
• Only works for simple tasks
• Assumes expert users
• Does not consider errors or learning
• Accuracy depends on assumptions

12.5.2 THREE-STATE MODEL

­ The Three-State Model explains how different input devices behave based on states of interaction.
­ Even if devices seem similar (mouse, touchscreen, light pen), they behave differently due to
sensory and motor characteristics.

States in the Model

The model defines three states:

State 0 – No Tracking
• Device is not in contact with the system
• System cannot detect position
Example:
• Finger not touching touchscreen

State 1 – Tracking
• Device is active
• System tracks movement
Example:
• Moving mouse without clicking

State 2 – Dragging / Action

• A button is pressed
• Movement causes action
Example:
• Dragging an icon with mouse
Examples of Devices
Mouse (State 1–2 Device)
• State 1 → Move cursor
• State 2 → Drag objects

Light Pen (State 0–1–2 Device)

• State 0 → Not touching screen
• State 1 → Touching screen (tracking)
• State 2 → Button pressed

Touchscreen (State 0–1 Device)

• State 0 → No touch
• State 1 → Touch detected

Device Limitations
If a device lacks required states:
• Additional keys can simulate missing states
• Example: using keyboard with touchscreen

User Effort and Accuracy

The model highlights that:
• Different states require different effort
• State 2 (button pressed) causes tension
• Accuracy and speed vary across states

Relation to Fitts' Law

Fitts’ Law calculates time taken to point at a target:
Time = a + b log₂(D/S + 1)
Where:
• D = distance
• S = size of target

Key Insight
The constants (a, b) depend on:
• Device type
• User skill
• Device state (0, 1, or 2)

Example Observation
• Touchscreens are less accurate initially (state 0 → 1)
• Mouse dragging (state 2) reduces accuracy due to tension

Advantages of Three-State Model

• Simple and easy to understand
• Helps compare input devices
• Useful for interface design

Limitations
• Does not cover complex interactions
• Focuses only on device behavior
• Needs combination with other models
COGNITIVE ARCHITECTURES
Cognitive architectures are models that describe how humans process information, think, and perform
tasks while interacting with systems.
In earlier models like GOMS, CCT, and KLM, there were certain assumptions about how users think and
act:
• GOMS assumes users solve problems using goals and subgoals (divide-and-conquer)
• CCT assumes the use of long-term and working memory with production rules
• KLM is based on the Model Human Processor, which explains human motor and mental actions

Another common assumption is that human understanding operates at different levels such as:
• Semantic (meaning)
• Syntactic (structure)
• Lexical (words)

12.6.1 PROBLEM SPACE MODEL

The Problem Space Model is based on the idea that human problem solving can be seen as searching
through a space of possible states to reach a goal.
It comes from research in artificial intelligence (AI) and cognitive psychology by Newell and Simon.

Basic Concept
A problem space consists of:
• A set of states (situations)
• A set of operations (actions that change states)
The goal is to:
Move from an initial state to a desired goal state

Rational Behavior and Knowledge-Level Systems

In AI, intelligent behavior is described as rational behavior, meaning:
• Actions are directed toward achieving a goal
• The system behaves logically based on knowledge
Such systems are called knowledge-level systems, which include:
• An agent (user/system)
• An environment
• Knowledge about goals and actions

Difference from Traditional Machine Models

In normal computation:
• Machines do not understand goals
• Programmers define the problem and solution
But in the problem space model:
The system represents goal-oriented behavior
Four Main Activities in Problem Space

1. Goal Formulation
• The system identifies what needs to be achieved
• Creates the initial state and goal state

2. Operation Selection
• Chooses the best action to move toward the goal
• Based on current state

3. Operation Application
• Executes the selected action
• Changes the state

4. Goal Completion
• Checks if the goal is reached
• Stops when desired state is achieved

Working Process (Cycle)

The process works as a continuous cycle:
1. Observe environment
2. Formulate goal
3. Select operation
4. Apply operation
5. Check if goal achieved

Recursive Nature (Important Concept)

A key strength of the model is recursion:
• If required knowledge is missing
• A new problem space is created
This creates a stack of problem spaces, where:
• New problems are added
• Completed ones are removed

Application – Soar Architecture

The problem space model is used in systems like Soar cognitive architecture.
Example application:
• Programmable User Models (PUMs)
• Used to simulate user behavior
• Helps detect design problems before implementation

Advantages
• Represents goal-directed behavior
• Models problem solving clearly
• Can predict errors and missing knowledge

Limitations
• Complex to implement
• Requires detailed knowledge representation
• Not always practical for simple tasks
12.6.2 INTERACTING COGNITIVE SUBSYSTEMS (ICS)
The Interacting Cognitive Subsystems (ICS), proposed by Barnard, is a different type of cognitive model.
Unlike other models, ICS:
Does not focus on sequences of actions
Focuses on overall mental processing and interaction

Main Idea
ICS views humans as an information-processing system made up of multiple interacting subsystems.
It provides a holistic (complete) view of:
• Perception
• Thinking
• Action

Integration of Two Approaches

ICS combines two psychological traditions:

1. Information Processing Approach

• Based on memory and mental processing
• Focuses on how information is handled

2. Computational Approach
• Based on AI and language processing
• Focuses on representation and structure

Structure of ICS
ICS consists of nine subsystems:
1. Peripheral Subsystems (5)
• Directly interact with the physical world
• Example: visual system (what we see)

2. Central Subsystems (4)

• Handle internal mental processes
• Example: propositional system

Subsystem Components
Each subsystem includes:
• Inputs and outputs
• Memory storage
• Processing functions
• Transformation mechanisms

Example: Propositional Subsystem

This subsystem handles knowledge such as:
• Meaning of words
• Relationships between concepts
Example:
Knowing that a word:
• Has four syllables
• Starts with “P”
• Refers to a place
Proceduralization (Very Important Concept)
ICS explains how users become experts through proceduralization:

Novice Users
• Think about each step
• Slow and error-prone

Expert Users
• Perform actions automatically
• Use stored procedures
• Faster and more accurate

Example
At an ATM:
• Novice → thinks step by step
• Expert → performs actions quickly without thinking

Role in Design
ICS helps designers:
• Create interfaces that support automatic behavior
• Reduce cognitive effort
• Encourage familiar patterns

Advantages
• Provides a complete view of cognition
• Explains learning and expertise development
• Useful for design guidance

Limitations
• Complex and difficult to apply
• Not focused on detailed task sequences
• Hard to measure quantitatively

COMMUNICATION AND COLLABORATION MODELS

­ Communication and collaboration are very important aspects of Human–Computer Interaction
(HCI), especially when more than one user is involved in a system.
­ A famous idea expressed by John Donne states that no person exists independently; everyone is
connected to others.
­ This idea clearly explains why communication is essential in both real-world and computer-based
environments.
­ In HCI, systems are not only used by individuals but also within groups and social contexts.
­ Even systems designed for single users are influenced by the environment and people around the
user.

1. COMPUTER-SUPPORTED COOPERATIVE WORK (CSCW)

The field of Computer-Supported Cooperative Work focuses on:

• Systems that support group activities and collaboration
• Understanding how computers affect cooperative work
CSCW includes:
• Groupware systems such as email and conferencing systems
• Study of social interaction and teamwork using computers
It covers both:
• Specific tools (like messaging systems)
• General effects of technology on group work

2. IMPORTANCE OF HUMAN COMMUNICATION

Effective communication is the foundation of collaboration. Many systems are designed to support
communication between users, especially when they are physically distant.

FACE-TO-FACE COMMUNICATION
Face-to-face communication is considered the most basic form of communication in terms of technology,
but at the same time, it is the most advanced and sophisticated form when we consider how humans
actually communicate.
• It is not limited to just speaking and hearing
• It involves multiple communication channels working together
• It allows high productivity and natural interaction
This makes face-to-face communication the most effective and complete communication method.

1. MULTIPLE COMMUNICATION CHANNELS

Face-to-face communication includes several channels:
• Speech and hearing for exchanging words
• Body language to express emotions and intentions
• Eye gaze (eye contact) to show attention and understanding
These channels work together to create rich and meaningful communication, which is difficult to
achieve in computer-mediated systems.

2. TRANSFER EFFECTS AND PERSONAL SPACE

Transfer Effects
When people use computer-mediated communication (like video calls or audio calls):
• They carry their expectations and social rules from face-to-face communication
• These rules are usually unconscious, so users may not realize when they are broken
As a result:
• Users may feel uneasy or uncomfortable
• They may think others are rude or unfriendly, even if the issue is actually due to the
communication medium

Personal Space
In face-to-face communication:
• People naturally maintain a comfortable distance between each other
• This distance depends on:
o Noise level
o Situation
o Cultural background
Important aspects of personal space:
• People move closer if they cannot hear properly
• People feel uncomfortable if someone stands too close in front of them
• Distance is more acceptable when someone stands at the side or behind
Cultural Differences
Personal space varies across cultures:
• North Americans → stand closer
• Britons → maintain more distance
• Southern Europeans and Arabs → stand even closer
This can lead to misunderstandings in cross-cultural communication.

Problems in Video Communication

In video conferencing:
• Camera angle and zoom may distort personal space
• One person may appear too close, while the other appears far away
This creates discomfort and confusion in interaction.

3. EYE CONTACT AND GAZE

Eye contact plays an important role in communication.

Functions of Eye Contact

• Shows whether the listener is paying attention
• Helps express:
o Interest
o Confusion
o Boredom
• Creates a sense of engagement and social presence
Regular eye contact helps maintain smooth communication.

Social Meaning of Eye Contact

• People who avoid eye contact may seem untrustworthy or uninterested
• Eye contact also reflects authority and power relationships

Problems in Computer Communication

• In audio communication → eye contact is completely missing
• In video communication:
o Camera placement may distort eye direction
o People may appear to be looking down instead of at each other
Even though facial expressions are visible, true eye contact is difficult to achieve.

Role in Conversation Focus

Eye gaze also helps in identifying objects or topics:
• Example: when someone says “this object,” the listener can understand by seeing where they are
looking
This function is often lost in computer-mediated systems.

4. GESTURES AND BODY LANGUAGE

Use of Gestures
People use hand movements and body posture to:
• Point to objects
• Indicate direction
• Support spoken words
This is called deictic reference, where meaning depends on gestures.
Problems in Computer Systems Solutions in Groupware
• Video systems may not clearly show gestures Some systems try to solve this problem:
• Important information may be lost • Group pointer → a shared cursor to
5. BACK CHANNELS, CONFIRMATION AND INTERRUPTION point at objects
• Shared work surfaces → show hand
What are Back Channels?
movements along with the screen
Back channels are:
• Small signals from the listener to the speaker
These help simulate real-world
• Examples:
pointing behavior.
o Nods
o Facial expressions
o Small sounds like “uh-huh”

Functions of Back Channels

• Show that the listener is understanding
• Provide feedback during conversation
• Help the speaker adjust their message
They allow conversation to flow smoothly.
Problems Without Back Channels
• The speaker may feel unsure if the listener understands
• The listener may appear uninterested
• Communication becomes uncomfortable

In Different Communication Media

• Video → some back channels available
• Audio → only verbal signals
• Text → no back channels

6. TURN-TAKING
Turn-taking is the process of:
• Switching roles between speaker and listener

How Turn-Taking Works

• Speakers pause briefly to allow others to speak
• Listeners may signal intention to speak using:
o Words like “um”
o Body movements

Role of Gaps in Speech

• Very short pauses indicate an opportunity to speak
• Long pauses may create confusion or discomfort

Maintaining or Claiming a Turn

• Sounds like “um” or “er” signal:
o The speaker wants to continue
o Or the listener wants to interrupt

Problems in Communication Systems

• In half-duplex systems (like walkie-talkies):
o Only one person can speak at a time
o Back channels are lost
• In satellite communication:
o Time delay (about 2 seconds) disrupts conversation
o Turn-taking becomes difficult
o Leads to interruptions and awkward pauses
TEXT-BASED COMMUNICATION
Text-based communication is the main form of communication in many groupware systems, especially in
asynchronous systems.
• Even though other media like voice messages, graphics, and video clips may be used
• Text remains the dominant communication medium
This is because text is easy to store, send, and access.
Although people are familiar with text (like writing letters), text-based communication in groupware acts
as a substitute for speech, which creates several challenges.

1. TYPES OF TEXTUAL COMMUNICATION

There are four main types of text-based communication in groupware:

1. Discrete Communication
• Messages are sent individually (like email)
• There is no direct connection between messages
• Any relation between messages depends on text references
Example: Email conversations

2. Linear Communication
• Messages are arranged in sequence (usually time order)
• All messages appear in a single continuous transcript
This is similar to a conversation log.

3. Non-Linear Communication
• Messages are connected using hypertext links
• Allows multiple conversation threads
This helps manage complex discussions.

4. Spatial Communication
• Messages are arranged on a two-dimensional surface
• Example: electronic pin boards
Messages are organized visually instead of sequentially.

2. DIFFERENCE FROM FACE-TO-FACE COMMUNICATION

Text-based communication differs significantly from face-to-face communication because it lacks:
• Immediate feedback
• Non-verbal cues
• Real-time interaction
This leads to several challenges.

3. BACK CHANNELS AND AFFECTIVE STATE

Loss of Back Channels
In face-to-face communication:
• Back channels (nods, “uh-huh”, expressions) help coordination
In text-based communication:
• These back channels are completely absent
As a result:
• The speaker cannot check understanding while speaking
• Messages must be fully completed before response
Loss of Affective State
Text communication removes:
• Tone of voice
• Body language
These normally express emotions like:
• Happiness
• Anger
• Humor

Solutions Used by Users

Users create explicit symbols to express emotions:
• → happy
• → sad
• → humorous
• LOL → laughing
These are called emotional tokens or smilies.

Behavioral Changes
• People use stronger and more emotional language
• But they themselves feel less emotionally involved
This happens because:
• Communication is indirect
• Interaction is spread over time

4. GROUNDING CONSTRAINTS
Grounding is the process of achieving mutual understanding.
According to Clark and Brennan, communication depends on certain constraints:

1. Cotemporality
• Message is received as soon as it is sent
Weak in text communication (delays exist)

2. Simultaneity
• Participants can send and receive at the same time
Limited in text communication

3. Sequence
• Messages follow a clear order
Disturbed in text due to overlapping messages

Problems in Text Communication

• Messages may arrive out of order
• Conversations become confusing
• Participants may interpret messages differently
This makes it harder to establish common ground.

•
• are corrected

In Multi-User Communication
• Turn-taking becomes complex
5. TURN-TAKING IN TEXT COMMUNICATION
In Two-Party Communication
• Turn-taking usually works
• Minor breakdowns occur but are corrected

In Multi-User Communication
• Turn-taking becomes complex
• Difficult to decide:
o Who should speak next

6. CONTEXT AND DEIXIS

Problem of Context
Text-based communication lacks:
• Shared physical environment
• Immediate context
This makes understanding harder.

Deictic Reference
Deictic expressions include:
• “this”, “that”, “there”
In face-to-face communication:
• These are supported by gestures
In text-based systems:
• Users must describe objects fully

Solution: WYSIWIS
• “What You See Is What I See”
Ensures all participants see the same content.

Problems with Internal Context

• Messages may overlap
• References become unclear
Leads to confusion and extra effort to clarify.

7. PACE AND GRANULARITY

Pace
• Refers to the speed of interaction
Text communication is slower than speech:
• Email → seconds to hours
• Chat → slower due to typing

Granularity
• Refers to amount of information per message
Slower pace leads to:
• Larger messages
• More detailed communication

Effect on Interaction
• Reduced interactivity
• Less immediate feedback
• Harder to guide conversation direction
8. COPING STRATEGIES
To manage slow pace, users adopt strategies:
1. Multiplexing
• Multiple topics discussed in one message
• Several conversations handled simultaneously

2. Eagerness
• Messages include:
o Possible responses
o Future scenarios
Example:
“If this happens, I will do this…”

Advantages
• Reduces number of interactions
• Saves time

Disadvantages
• May ignore other possible discussion paths
• Can become confusing

9. REVIEWABILITY
Text communication has one major advantage:
• Messages can be stored and reviewed
This helps:
• Revisit earlier discussions
• Clarify misunderstandings
Unlike speech, which is difficult to review.

10. LINEAR TEXT VS HYPERTEXT

Linear Text
• Easy to read
• Provides continuous flow
• Better for understanding conversation

Hypertext
• Supports multiple conversation paths
• Reduces overlap confusion

GROUP WORKING
Group working is more complex than individual interaction because it involves:
• Multiple people
• Changing social relationships
• Dynamic roles and behaviors
Earlier, we discussed communication mainly between two people. However, in group working:
• The behavior of individuals is influenced by the group environment
• Relationships and roles are not fixed
• Group interaction is dynamic and constantly changing
This section focuses on:
• Groups that are actively working together
• Not long-term organizational structures, but short-term working interactions
1. GROUP DYNAMICS
Changing Roles in Groups
Unlike organizational roles (like manager or employee), group roles:
• Are not stable
• Can change during a task
• Can even change within a single session
Example:
In joint authoring:
• A person may act as:
o Author
o Co-author
o Commentator
• These roles keep changing over time

Problem with Fixed Roles in Systems

Some groupware systems assign formal roles, but:
• Real-world roles are socially defined and flexible
• Systems must allow roles to change dynamically
Otherwise:
• The system will not match real group behavior

Problem with Naming Roles

Even naming roles can create social issues:
Example:
• A person contributes ideas but does not write content
• System labels them as:
o “Commentator” or “Reviewer” instead of “Author”
This may:
• Hurt feelings
• Create conflict
• Cause social friction

Role Inversion in Groups

Sometimes, authority roles reverse depending on context
Example:
• A managing director visiting a coal mine
• Must follow the instructions of the site supervisor
Problem:
• It is difficult for:
o The supervisor to give orders to the director
o The director to accept a lower role
If systems enforce strict roles:
• This problem becomes worse

Changing Group Membership

Group composition is not fixed:
• Members may join or leave
• This causes major changes in group behavior
Example:
• If a dominant member leaves:
o Group may shift from:
▪ Leader-based → Democratic structure
Role of Groupware
Groupware systems can help by:
• Recording group history
• Allowing new members to:
o Review past discussions
o Understand context
Important requirement:
• New users must see all previous contributions, not just new ones

Subgroups in Group Work

Groups often:
• Split into subgroups for detailed work
• Then rejoin the main group
Systems must support:
• Independent subgroup work
• Sharing of results later

Example
Early versions of CoLab:
• Supported only a single shared screen
• Could not handle subgroups
Later versions:
• Allowed multiple groups to work separately

2. PHYSICAL LAYOUT

Importance of Physical Environment

The physical layout of a room affects:
• Communication
• Interaction
• Power relationships
This is especially important in:
• Meeting rooms
• Collaborative environments

Effect of Eye Contact and Visibility

From earlier discussion:
• Eye contact and gestures are important
• If monitors block view:
o Communication is affected
Therefore:
• Equipment should not obstruct visibility

Orientation of Equipment
The placement of computers influences interaction:
• If people face screens → less communication
• If people face each other → more discussion
Example:
• Inward-facing terminals encourage:
o Eye contact
o Better interaction
Power Positions in a Room
Physical layout affects authority and control
Traditional Meeting Room
• Power position → front of the room
• Near:
o Whiteboard
o Presentation area
Managers usually sit here

Electronic Meeting Room

In systems like Capture Lab:
• Screen can be controlled from anywhere
• Front position loses importance
New power position:
• Back of the room
Reason:
• Can observe everyone
• Others must turn to look at speaker
• Increases authority

User Adaptation Problem

Managers initially:
• Sit in traditional power seats
But later:
• Feel uncomfortable
• Move to new power positions
Shows that:
• Technology changes social behavior

3. DISTRIBUTED COGNITION
Traditional cognition focuses on:
• Thinking inside the human mind
Distributed cognition suggests:
• Thinking happens across:
o People
o Tools
o Environment
Represented by Distributed cognition

Key Concept
Cognition is not limited to the brain, but includes:
• External objects
• Interactions with others

Examples
• Notes on paper
• Whiteboard diagrams
• Computer systems
These act as extensions of memory

Relation to Traditional Models

DIFFERENCES BETWEEN TASK ANALYSIS AND OTHER TECHNIQUES
I’ve written it in clear, full sentences with detailed points exactly from your text, without adding extra
content.

Task analysis is a process used to:

• Study how people perform their work
• Understand:
o What tasks they do
o What objects they interact with
o What knowledge they require
It mainly focuses on real-world activities, not just computer interaction.

DIFFERENCES BETWEEN TASK ANALYSIS AND OTHER TECHNIQUES

1. Scope of Task Analysis vs Other Techniques

• Task analysis has a very wide scope.
• It includes:
o Activities involving computers
o Activities NOT involving computers
Example:
In word processing task analysis, it includes:
• Fetching documents from a filing cabinet
• Changing printer ribbon
• Handling disks
This means:
• Task analysis studies the entire work environment, not just software interaction

Comparison
• Like traditional systems analysis:
o It is not limited to computer-based tasks
• However:
o Task analysis gives more importance to the user

2. Main Difference in Emphasis

The main difference between task analysis and older techniques is:
Task analysis emphasizes the user
• Focus is on:
o User activities
o User needs
o User behavior
Whereas:
• Traditional techniques focus more on:
o System structure
o Technical aspects

3. Task Analysis vs Goal-Oriented Cognitive Models

Task analysis has similarities with models like
GOMS model
Similarity
• Both can:
o Represent tasks as hierarchies
o Break tasks into sub-tasks

Key Difference
Task Analysis
• Focuses on:
o External observable behavior
• Includes:
o Real-world actions
o Physical activities
Example:
• Fetching a document from a cabinet

Goal-Oriented Models (like GOMS)

• Focus on:
o Internal cognitive processes
• Concerned with:
o Mental steps
o Decision making
These models use:
• Smaller, detailed steps (fine granularity)

4. Observable Behavior vs Mental State

Task analysis mainly:
• Observes what users do externally
Focus:
• Actions
• Tasks
• Interactions

Different Views
• Some practitioners say:
o Task analysis should only focus on:
▪ What users do (objective behavior)
▪ Not why they do it
• However:
o In practice, some inference about:
▪ User goals and intentions is included
Example:
• Task names often reflect user goals

5. Conceptual Model vs Observation

• Some task analysis approaches:
o Try to build a conceptual model of the user
This means:
• Understanding:
o How users think about tasks
o How they organize knowledge
This is especially seen in:
• Knowledge-based techniques
6. Level of Detail (Granularity)
Task Analysis
• May produce:
o Low-level task breakdown
• But:
o This is usually the final output
Used for:
• Designing interfaces
• Structuring dialogs

Cognitive Models
• Goal hierarchy is:
o The main focus
• Further analyzed for:
o Complexity
o Learnability
o Performance

7. Role in Design Life Cycle

Task analysis and cognitive models are used at different stages.
Task Analysis
• Used at the beginning
• During:
o Requirements gathering
Purpose:
• Understand user tasks
• Define system requirements

Cognitive Models
• Used at the end
• During:
o Evaluation
Purpose:
• Analyze:
o Usability
o Efficiency
o Performance

TASK DECOMPOSITION (Hierarchical Task Analysis – HTA)

I’ve written it in clear language, full sentences, and detailed points exactly from your text, without adding
extra content.
Task decomposition is a fundamental part of task analysis in which:
• A large task is broken down into smaller subtasks
• These subtasks represent:
o The steps required to complete the task
o The order in which actions are performed
Example:
The task “clean the house” can be decomposed into:
• Get the vacuum cleaner
• Attach the required parts
• Clean rooms
• Empty the dust bag (if needed)
• Put equipment away
HIERARCHICAL TASK ANALYSIS (HTA)
A common method for task decomposition is:
Hierarchical Task Analysis (HTA)

Outputs of HTA
HTA produces two main outputs:
1. Hierarchy of tasks and subtasks
o Shows how tasks are broken down
2. Plans
o Describe:
▪ The order of tasks
▪ Conditions under which tasks are performed

Representation of HTA
• Tasks are shown using:
o Indentation or numbering
• This shows the hierarchical relationship
Example:
• Task 0 → Main task
• Task 1, 2, 3 → Subtasks
• Task 3.1, 3.2 → Sub-subtasks

Role of Plans
Plans explain:
• Which tasks are performed
• When they are performed
• In what order
Important point:
• Not all subtasks:
o Need to be performed
o Are performed in fixed order

Example (Dust Bag)

• Task: Empty dust bag
• Condition: Only when bag is full
This shows:
• Tasks can be conditional

Flexibility in Task Order

Some tasks:
• Can be done in any order
Example:
• Cleaning rooms:
o Order is flexible
o Only dirty rooms need cleaning

Making Plans More Specific

Plans can include detailed rules:
Example:
• Clean hall → daily
• Clean bedroom → weekly
PROCESS OF TASK DECOMPOSITION
Task decomposition is an iterative process

Steps Involved
1. Start with a main task
2. Ask:
o What subtasks are needed?
3. Break each subtask further
4. Repeat until tasks are detailed enough

Sources of Information
To identify tasks, we use:
• Observation
• Expert knowledge
• Documentation

STOPPING RULES (When to Stop Decomposition)

Task decomposition cannot go on forever, so we use stopping rules.

1. Based on Purpose
The level of detail depends on:
• The goal of analysis
Example:
• For system automation:
o Focus on monitoring and actions
• For training manuals:
o Focus on decision-making steps

2. P × C Rule (Very Important)

P × C = Probability × Cost
• If:
o Probability of error × Cost of error is low
Then:
• Stop decomposition

Meaning
• Simple tasks:
o Do not need further breakdown
• Unless:
o They are critical

3. Motor Actions
• Stop when tasks involve:
o Complex physical actions (e.g., mouse movement)
Reason:
• Further breakdown is not useful

4. Internal Cognitive Tasks

• Do not expand:
o Pure thinking or mental processes
Exception:
• If designing a decision support system
REPRESENTATION OF TASK HIERARCHY
Task hierarchy can be shown in:
• Text form
• Diagram form
Diagrams:
• Easier to understand visually

REFINING THE TASK HIERARCHY

After creating the hierarchy:
• It must be checked and improved

1. Detecting Errors and Omissions

• Present tasks to experts
• Missing steps can be identified
Example:
• Tea-making task missed:
o “Warm the pot”

2. Finding Implicit Steps

Example:
• “Turn off gas” exists
• But “turn on gas” is missing
Must be added
3. Balancing the Hierarchy
• If structure is uneven:
o Add new levels
Example:
• Add “make pot” as a higher-level task

FINAL TASK STRUCTURE (Example: Tea Making)

Top-level tasks:
• Boil water
• Empty pot
• Make pot
• Wait
• Pour tea

TYPES OF PLANS IN HTA

HTA includes different types of plans:

1. Fixed Sequence
• Tasks performed in a fixed order
2. Optional Tasks
• Performed only if needed
Example:
• Add sugar (optional)

3. Waiting for Events

• Task depends on time or condition
Example:
• Wait for water to boil
4. Cycles (Repetition)
• Tasks repeated until condition met
Example:
• Fill cups until no cups left

5. Time Sharing
• Tasks can be done simultaneously
Example:
• Boiling water and preparing pot

6. Discretionary Tasks
• User decides whether or not to perform
Example:
• Clean rooms based on need

7. Mixed Plans
• Combination of multiple plan types

IMPORTANT CHARACTERISTICS
• Task decomposition is:
o Not straightforward
o Requires skill
• Different analysts:
o May produce different results
• There is:
o No single correct answer
The result depends on:
• Purpose of analysis