[PDF] - The Human Ov erview Human can b e view ed as an PDF Document

SLIDE 1 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (1) The Human Ov erview Human can b e view ed as an information pro cessing system, for example, Card, Moran and New ell's Mo del Human Pro cessor. A simple mo del:

information

receiv ed and resp

nses

giv en via input{output c hannels

information

stored in memory

information

pro cessed and applied in v arious w a ys Capabiliti es

f

h umans in these areas are imp

rtan

t to design, as are individual dierences.

SLIDE 2 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (2) Input{Output c hannels Vision Tw

stages

in vision

ph

ysical reception

f

stim ulus

pro

cessing and in terpretation

f

stim ulus The ph ysical apparatus: the ey e

mec

hanism for receiving ligh t and transforming it in to electrical energy

ligh

t reects from

b

jects; their images are fo cused upside-do wn

n

retina

retina

con tains ro ds for lo w ligh t vision and cones for colour vision

ganglion

cells detect pattern and mo v emen t

SLIDE 3 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (3) In terpreting the signal Size and depth

visual

angle indicates ho w m uc h

f

eld

f

view

b

ject

ccupies

(relates to size and distance from ey e)

visual

acuit y is abilit y to p erceiv e ne detail (limited)

familiar
b

jects p erceiv ed as constan t size in spite

f

c hanges in visual angle | la w

f

size constancy

cues

lik e

v

erlapping help p erception

f

size and depth

SLIDE 4 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (4) In terpreting the signal (con t) Brigh tness

sub

jectiv e reaction to lev els

f

ligh t

aected

b y luminance

f
b

ject

measured

b y just noticeable dierence

visual

acuit y increases with luminance as do es ic k er Colour

made

up

f

h ue, in tensit y , saturation

cones

sensitiv e to colour w a v elengths

blue

acuit y is lo w est

8%

males and 1% females colour blind

SLIDE 5 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (5) In terpreting the signal (con t) The visual system comp ensates for mo v emen t and c hanges in luminance. Con text is used to resolv e am biguit y . Optical illusi

ns

sometimes

ccur

due to

v

er comp ensation.

SLIDE 6 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (6) Figure 1: The P

nzo

illusi

n

Figure 2: The Muller Ly er illusi

n

SLIDE 7 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (7) Reading Sev eral stages:

visual

pattern p erceiv ed

deco

ded using in ternal represen tation

f

language

in

terpreted using kno wledge

f

syn tax, seman tics, pragmatics Reading in v

lv

es saccades and xations. P erception

ccurs

during latter. W

rd

shap e is imp

rtan

t to recognition. Negativ e con trast impro v es reading from computer screen.

SLIDE 8 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (8) Hearing Pro vides information ab

ut

en vironmen t: distances, directions,

b

jects etc. Ph ysical apparatus:

uter

ear | protects inner and amplies sound

middle

ear | transmits sound w a v es as vibrations to inner ear

inner

ear | c hemical transmitters are released and cause impulses in auditory nerv e Sound

pitc

h | sound frequency

loudness

| amplitude

tim

bre | t yp e

r

qualit y

SLIDE 9 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (9) Hearing (con t) Humans can hear frequencies from 20Hz to 15kHz | less accurate distinguishi ng high frequencies than lo w. Auditory system lters sounds | can attend to sounds

v

er bac kground noise. F

r

example, the co c ktail part y phenomenon.

SLIDE 10 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (10) T

uc

h Pro vides imp

rtan

t feedbac k ab

ut

en vironmen t. Ma y b e k ey sense for someone who is visually impaired. Stim ulus receiv ed via receptors in the skin:

thermoreceptors

| heat and cold

no

ciceptors | pain

mec

hanoreceptors | pressure (some instan t, some con tin uous) Some areas more sensitiv e than

thers

e.g. ngers. Kinethesis | a w areness

f

b

dy

p

sition

aecting comfort and p erformance.

SLIDE 11 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (11) Mo v emen t Time tak en to resp

nd

to stim ulus: reaction time + mo v emen t time Mo v emen t time | dep enden t

n

age, tness etc. Reaction time | dep enden t

n

stim ulus t yp e:

visual

| 200ms

auditory

| 150 ms

pain

| 700ms Increasing reaction time decreases accuracy in the unskilled

p

erator but not in the skilled

p

erator.

SLIDE 12 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (12) Mo v emen t (con t) Fitts' La w describ es the time tak en to hit a screen target: M t = a + b l

g

2 (D =S + 1) where a and b are empirically determined constan ts, M t is mo v emen t time, D is Distance and S is Size. T argets in general should b e large as p

ssible

and the distances as small as p

ssible.

SLIDE 13 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (13) Memory There are three t yp es

f

memory function.

Iconic Echoic Haptic Sensory memories

r

Short-term memory Working memory Long-term memory Attention Rehearsal

Sensory memory Buers for stim uli

iconic

| visual stim uli

ec

hoic | aural stim uli

haptic

| touc h stim uli Constan tly

v

erwritten. Information passes from sensory to STM b y atten tion. Selection

f

stim uli go v erned b y lev el

f

arousal.

SLIDE 14 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (14) Short-term memory (STM) Scratc h-pad for temp

rary

recall

rapid

access | 70ms

rapid

deca y | 200ms

limited

capacit y | 7 + =

2

digits

r

c h unks

f

information Recency eect | recall

f

most recen tly seen things b etter than recall

f

earlier items. Some evidence for sev eral elemen ts

f

STM | articulatory c hannel, visual c hannel etc. | in terference

n

dieren t c hannel do es not impair recall.

SLIDE 15 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (15) Long-term memory (L TM) Rep

sitory

for all

ur

kno wledge

slo

w access | 1/10 second

slo

w deca y , if an y

h

uge

r

unlimited capacit y Tw

t

yp es

episo

dic | serial memory

f

ev en ts

seman

tic | structured memory

f

facts, concepts, skills Information in seman tic L TM deriv ed from episo dic L TM.

SLIDE 16 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (16) Long-term memory (con t.) Seman tic memory structure

pro

vides access to information

represen

ts relationships b et w een bits

f

information

supp
rts

inference Mo del: seman tic net w

rk
inheritance

| c hild no des inherit prop erties

f

paren t no des

relationships

b et w een bits

f

information explicit

supp
rts

inference through inheritance

SLIDE 17 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (17) Long-term memory (con t.)

is a DOG is a colour: [brown/white, black/white, merle] instance is a HOUND is a BEAGLE instance SNOOPY friend of CHARLIE BROWN colour:[brown,black/white] size:small tracks has tail has four legs SHEEPDOG works sheep barks ANIMAL breathes moves is a COLLIE size: medium LASSIE film character colour:brown/white SHADOW colour: brown/white book character instance cartoon/book character

SLIDE 18 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (18) Long-term memory (con t.) Other mo dels

f

L TM F rames: Information

rganized

in data structure. Slots in structure are instan tiated with particular v alues for a giv en instance

f

data.

DOG

Fixed legs: 4 Default diet: carnivorous sound: bark Variable size: colour:

COLLIE

Fixed breed of: DOG type: sheepdog Default Variable colour: size: 65 cm

SLIDE 19 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (19) Long-term memory (con t.) Scripts: Mo del

f

stereot ypical information required to in terpret situation

r

language. Script also has elemen ts whic h can b e instan tiated with particular v alues.

dog needs medicine dog needs operation Tracks: paying examination waiting in room arriving at reception Scenes: vet examines Roles: diagnoses treats

wner brings dog in

pays takes dog out Props: examination table medicine instruments Result: dog better

wner poorer

vet richer dog ill vet open

wner has money

Entry conditions: Script for a visit to the vet

SLIDE 20 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (20) Long-term memory (con t.) Pro duction rules: Represen tation

f

pro cedural kno wledge. Condition{acti

n

rules | if condition is matc hed, rule res. L TM pro cesses Storage

f

information

information

mo v es from STM to L TM b y rehearsal

amoun

t retained prop

rtional

to rehearsal time: total time h yp

thesis
ptimized

b y spreading learning

v

er time: distribution

f

practice eect

structure,

meaning and familiarit y mak e information easier to remem b er

SLIDE 21 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (21) L TM pro cesses (con t.) F

rgetting
deca

y { information is lost gradually but v ery slo wly

in

terference | new information replaces

ld:

retroactiv e in terference

ld

ma y in terfere with new: proactiv e inhibition | so ma y not forget at all

memory

is selectiv e and aected b y emotion | can `c ho

se'

to forget Information retriev al

recall

| information repro duced from memory . Can b e assisted b y cues, e.g. categories, imagery

recognition

| information giv es kno wledge that it has b een seen b efore. Less complex than recall | information is cue.

SLIDE 22 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (22) Thinking: reasoning and problem solving Reasoning Deductiv e: deriv e logicall y necessary conclusion from giv en premises. E.g. If it is F rida y then she will go to w

rk

It is F rida y Therefore she will go to w

rk.

Logical conclusion not necessarily true: If it is raining then the ground is dry It is raining Therefore the ground is dry Human deduction p

r

when truth and v alidi t y clash.

SLIDE 23 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (23) Reasoning (con t.) Inductiv e: generalize from cases seen to cases unseen. E.g. all elephan ts w e ha v e seen ha v e trunks therefore all elephan ts ha v e trunks. Unreliable: can

nly

pro v e false not true. Ho w ev er, h umans are not go

d

at using negativ e evidence. E.g. W ason's cards.

7 4 E K

Ab ductiv e: reasoning from ev en t to cause. E.g. Sam driv es fast when drunk. If see Sam driving fast, assume drunk. Unreliable: can lead to false explanations.

SLIDE 24 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (24) Problem solving Pro cess

f

nding solution to unfamiliar task using kno wledge. Sev eral theories. Gestalt

problem

solving b

th

pro ductiv e and repro ductiv e

pro

ductiv e problem solving dra ws

n

insigh t and restructuring

f

problem

attractiv

e but not enough evidence to explain `insigh t' etc.

mo

v e a w a y from b eha viouralism and led to information pro cessing theories

SLIDE 25 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (25) Problem solving (con t.) Problem space theory

problem

space comprises problem states

problem

solving in v

lv

es generating states using legal

p

erators

heuristics

ma y b e emplo y ed to select

p

erators e.g. means-ends analysis

p

erates within h uman information pro cessing system e.g. STM limits etc.

largely

applied to problem solving in w ell dened areas e.g. puzzles rather than kno wledge in tensiv e areas

SLIDE 26 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (26) Problem solving (con t.) Analogy

no

v el problems are solv ed b y using kno wledge from a similar domain in new domain | analogical mapping

analogical

mapping ma y b e dicult if domains are seman tically dieren t Skill acquisition Skilled activit y c haracterized b y

c

h unking | lot

f

information is c h unk ed to

ptimize

STM

conceptual

rather than sup ercial grouping

f

problems | information is structured more eectiv el y

SLIDE 27 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (27) Skill acquisition (con t.) Mo del

f

skill acquisiti

n:

A CT* 3 lev els

f

skill

general

purp

se

rules to in terpret facts ab

ut

problem | kno wledge in tensiv e

sp

ecic task rules are learned | rely

n

kno wn pro cedures

rules

are ne-tuned | skille d b eha viour Mec hanisms for mo ving b et w een these

pro

ceduralization | lev el 1 to lev el 2

generalization

| lev el 2 to lev el 3

SLIDE 28 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (28) Skill acquisition { pro ceduralization Lev el 1: IF co

k[t

yp e, ingredien ts, time] THEN co

k

for: time co

k[casserole,

[c hic k en,carrots,p

tato

es], 2 hours] co

k[casserole,

[b eef, dumpling, carrots], 2 hours] co

k[cak

e, [our, sugar,butter, egg], 45 mins] Lev el 2: IF t yp e is casserole AND ingredien ts are [c hic k en,carrots,p

tato

es] THEN co

k

for: 2 hours IF t yp e is cak e AND ingredien ts are [our,sugar,butter ,eggs] THEN co

k

for: 45 mins

SLIDE 29 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (29) Skill acquisition { generalizati

n

Lev el 2: IF t yp e is casserole AND ingredien ts are [c hic k en,carrots,p

tato

es] THEN co

k

for: 2 hours IF t yp e is casserole AND ingredien ts are [b eef,dumplings,carrots] THEN co

k

for: 2 hours Lev el 3: IF t yp e is casserole AND ingredien ts are ANYTHING THEN co

k

for: 2 hours

SLIDE 30 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (30) Errors and men tal mo dels T yp es

f

error

slips

| c hange to asp ect

f

skilled b eha viour can cause slip

incorrect

understanding | h umans create men tal mo dels to explain b eha viour. If wrong (dieren t from actual system) errors can

ccur.

SLIDE 31 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (31) Individual dierences

long

term | sex, ph ysical and in tellec tual abiliti es

short

term | eect

f

stress

r

fatigue

c

hanging | age Ask: will design decision exclude section

f

user p

pulation?

SLIDE 32 Human{Com puter In teraction, Pren tice Hall A. Dix, J. Finla y , G. Ab

wd

and R. Beale c

1993

The Human Chapter 1 (32) Cognitiv e Psyc hology and In teractiv e System Design Some direct applications. E.g. blue acuit y is p

r

so blue should not b e used for imp

rtan

t detail. Ho w ev er, application generally requires

understanding
f

con text in psyc hology

understanding
f

particular exp erimen tal conditions A lot

f

kno wledge has b een distille d in

guidelines

| see Chapters 4 and 5

cognitiv

e mo dels | see Chapter 6

exp

erimen tal and analytic ev aluation tec hniques | see Chapter 11