User Modeling

Predicting thoughts and actions GOMS

___________________________________________________________________________________________________ SCHOOL OF INTERACTIVE ARTS + TECHNOLOGY [SIAT] | WWW.SIAT.SFU.CA

Agenda
User

modeling

Fitts Law GOMS

Feb 24, 2011

IAT 334

User Modeling
Idea:

If we can build a model of how a user works, then we can predict how s/he will interact with the interface
Predictive modeling

Many

different modeling techniques exist

Feb 24, 2011

IAT 334

User Modeling 2 types

Stimulus-Response
Hicks law Practice law Fitts law

Cognitive human as interperter/predictor based on Model Human Processor (MHP)

Key-stroke Level Model
Low-level, simple

GOMS (and similar) Models

Higher-level (Goals, Operations, Methods, Selections) Not discussed here
Feb 24, 2011 IAT 334 4

Power Law of Practice

Tn = T1n-a
Tn to complete the nth trial is T1 on the first trial times n to the power -a; a is about .4, between .2 and .6 Skilled behavior - Stimulus-Response and routine cognitive actions
Typing speed improvement Learning to use mouse Pushing buttons in response to stimuli NOT learning
IAT 334 5

Feb 24, 2011

Power Law of Practice

How

to use it?

Use measured T1 on the first trial

Predict whether usability criteria will be met How many trials?

Predict how many practice iterations needed to reach usability criteria

Feb 24, 2011

IAT 334

Hicks Law
Decision

time to choose among n equally likely alternatives

T = Ic log2(n+1) Ic ~ 150 msec

Feb 24, 2011

IAT 334

Hicks Law
How

to use it?

Menu selection Choose among 64 choices:

Single 64-item menu 2-level menu: 8 choices at each level 2-level menu: 4 choices then 16 choices

Feb 24, 2011

IAT 334

Fitts Law
Models

movement times for selection (reaching) tasks in one dimension Basic idea: Movement time for a selection task
Increases as distance to target increases Decreases as size of target increases

Feb 24, 2011

IAT 334

Fitts Experiment: 1D
d w

Feb 24, 2011

IAT 334

10

Fitts: Index of Difficulty

ID - Index of difficulty

ID = log2 (d/w + 1.0)

bits result

distance to move

width (tolerance) of target

ID is an information theoretic quantity

Based on work of Shannon larger target => more information (less uncertainty)

Feb 24, 2011

IAT 334

11

Fitts formula
MT

- Movement time

MT = k1 + k2*ID MT = k1 + k2 *log2 (d/w + 1.0)

MT

is a linear function of ID

k1 and k2 are experimental constants

Feb 24, 2011

IAT 334

12

Run empirical tests to determine k1 and k2 in MT = k1 + k2* ID Will get different ones for different input devices and device uses

MT ID = log2(d/w = 1.0)
Feb 24, 2011 IAT 334 13

What about 2D
h

x w rect: one way is ID = log2(d/min(h, w) + 1)

Should take into account direction of approach

Feb 24, 2011

IAT 334

14

Design implications
Menu

item size Icon size Put frequenlty used icons together Scroll bar target size and placement
Up / down scroll arrows together or at top and bottom of scroll bar

Feb 24, 2011

IAT 334

15

GOMS
One

of the most widely known

Assumptions

Know sequence of operations for a task Expert will be carrying them out
Goals,

Operators, Methods, Selection Rules

IAT 334 16

Feb 24, 2011

GOMS Procedure
Walk

through sequence of steps Assign each an approximate time duration -> Know overall performance time
(Can

be tedious)

Feb 24, 2011

IAT 334

17

Limitations
GOMS

is not for

Tasks where steps are not well understood Inexperienced users

Why? Good

example: Move a sentence in a document to previous paragraph

IAT 334 18

Feb 24, 2011

Goal
End

state trying to achieve Then decompose into subgoals

Select sentence Moved sentence Cut sentence Move to new spot Paste sentence Place it

Feb 24, 2011

IAT 334

19

Operators
Basic

actions available for performing a task (lowest level actions) move mouse pointer, drag, press key, read dialog box,

Examples:

Feb 24, 2011

IAT 334

20

Methods
Sequence

of operators (procedures) for accomplishing a goal (may be multiple) Select sentence

Example:

Move mouse pointer to first word Depress button Drag to last word Release
Feb 24, 2011 IAT 334 21

Selection Rules
Invoked

method

when there is a choice of a

Example:

Could cut sentence either by menu pulldown or by ctrl-x

Feb 24, 2011

IAT 334

22

Further Analysis
GOMS

is often combined with a keystroke level analysis

Assigns times to different operators Plus: Rules for adding Ms (mental preparations) in certain spots

Feb 24, 2011

IAT 334

23

Example
Move Sentence
1. Select sentence Reach for mouse Point to first word Click button down Drag to last word Release H P K P K 0.40 1.10 0.60 1.20 0.60 3.90 secs

2. Cut sentence Press, hold ^ Press and release x Release ^

or

Point to menu Press and hold mouse Move to cut Release

3. ...
Feb 24, 2011 IAT 334 24

Keystroke-Level Model
Simplified GOMS KSLM - developed by Card, Moran & Newell, see their book

The Psychology of Human-Computer Interaction, Card, Moran and Newell, Erlbaum, 1983

Skilled users performing routine tasks Assigns times to basic human operations experimentally verified Based on MHP - Model Human Processor

Feb 24, 2011 IAT 334 25

User Profiles
Attributes:

attitude, motivation, reading level, typing skill, education, system experience, task experience, computer literacy, frequency of use, training, color-blindness, handedness, gender,
Novice,
Feb 24, 2011

intermediate, expert
IAT 334 26

Motivation

User
Low motivation, discretionary use Low motivation, mandatory High motivation, due to fear High motivation, due to interest

Design goal
Ease of learning Control, power

Ease of learning, robustness, control Power, ease of use

Feb 24, 2011

IAT 334

27

Knowledge & Experience

Experience task system

low
high

low
high

Design goals
Many syntactic and semantic prompts Efficient commands, concise syntax Semantic help facilities
Lots of syntactic prompting

low
high

high
low

Feb 24, 2011

IAT 334

28

Job & Task Implications

Frequency of use
High - Ease of use Low - Ease of learning & remembering

Task implications
High - Ease of use Low - Ease of learning

System use
Mandatory - Ease of using Discretionary - Ease of learning

Feb 24, 2011

IAT 334

29

Modeling Problems
1.

Terminology - example

High frequency use experts - cmd language Infrequent novices - menus

Whats

frequent, novice?

Feb 24, 2011

IAT 334

30

Modeling Problems (contd.)

2.

Dependent on grain of analysis employed

Can break down getting a cup of coffee into 7, 20, or 50 tasks That affects number of rules and their types

Feb 24, 2011

IAT 334

31

Modeling Problems (contd.)

3.

Does not involve user per se

Dont inform designer of what user wants

4.

Time-consuming and lengthy

Feb 24, 2011

IAT 334

32