Reading linear texts on paper versus computer screen: Effects on reading comprehension

Reading

linear

texts

on

paper

versus

computer

screen:

Effects

on

reading

comprehension

Anne

Mangen

a,

*

,

Bente

R.

Walgermo

a

,

Kolbjørn

Brønnick

a,b

a

The

National

Centre

for

Reading

Education

and

Research,

University

of

Stavanger,

NO-4036

Stavanger,

Norway

b

Regional

Centre

for

Clinical

Research

in

Psychosis,

Division

of

Psychiatry,

Stavanger

University

Hospital,

PO

Box

8100,

NO-4068

Stavanger,

Norway

1.

Introduction

There

is

an

ongoing

transition

of

reading

from

print

to

screen

and

the

book

is

challenged

by

an

increasing

number

of

digital

reading

devices

(computers

and

laptops,

e-books,

tablet

devices,

smart

phones).

The

paradigm

of

reading,

in

particular

for

young

people,

is

increasingly

screen-based

rather

than

paperbound.

The

theoretical

and

pedagogical

implications

of

the

ongoing

digitization

for

reading

and

reading

comprehension

are

complex

and

multifaceted,

and

a

number

of

fundamental

research

questions

remain

at

best

partially

addressed:

how

and

to

what

extent

might

comprehension

of

linear,

narrative

and

non-narrative

texts

differ

when

texts

are

displayed

on

a

screen

as

compared

to

being

printed

on

paper?

Does

it

impact

students’

reading

comprehension

and

learning

to

read

geography,

science

and

history

texts

as

PDFs

on

a

computer

screen

instead

of

in

a

print

text

book?

In

the

Norwegian

school

system,

these

issues

have

become

highly

relevant

as

texts

are

increasingly

distributed

as

PDF

ﬁles

and

there

is

a

shift

toward

using

computer-presented

PDFs

in

realistic

test

situations.

To

what

extent

does

such

a

shift

affect

reading

comprehension?

This

is

the

topic

of

the

present

study.

There

exists

a

large

body

of

research

on

the

impact

and

effect

of

different

aspects

of

digital

textuality

on

reading

comprehension.

Many

studies

have

been

addressing

the

impact

of

hypertext

structure

on

cognitive

aspects

of

reading

and

comprehension.

A

recent

review

(DeStefano

&

LeFevre,

2007)

concluded

that

hypertext

structure

tends

to

increase

cognitive

demands

of

decision

making

and

visual

processing

and

this

additional

cognitive

load,

in

turn,

impairs

reading

comprehension

performance.

International

Journal

of

Educational

Research

58

(2013)

61–68

A

R

T

I

C

L

E

I

N

F

O

Article

history:

Received

24

May

2012

Received

in

revised

form

6

December

2012

Accepted

10

December

2012

Available

online

5

January

2013

Keywords:

Reading

comprehension

Screen

reading

Print

reading

Computers

in

education

A

B

S

T

R

A

C

T

Objective:

To

explore

effects

of

the

technological

interface

on

reading

comprehension

in

a

Norwegian

school

context.

Participants:

72

tenth

graders

from

two

different

primary

schools

in

Norway.

Method:

The

students

were

randomized

into

two

groups,

where

the

ﬁrst

group

read

two

texts

(1400–2000

words)

in

print,

and

the

other

group

read

the

same

texts

as

PDF

on

a

computer

screen.

In

addition

pretests

in

reading

comprehension,

word

reading

and

vocabulary

were

administered.

A

multiple

regression

analysis

was

carried

out

to

investigate

to

what

extent

reading

modality

would

inﬂuence

the

students’

scores

on

the

reading

comprehension

measure.

Conclusion:

Main

ﬁndings

show

that

students

who

read

texts

in

print

scored

signiﬁcantly

better

on

the

reading

comprehension

test

than

students

who

read

the

texts

digitally.

Implications

of

these

ﬁndings

for

policymaking

and

test

development

are

discussed.

ß

2012

Elsevier

Ltd.

All

rights

reserved.

*

Corresponding

author.

Tel.:

+47

51

83

32

00;

fax:

+47

51

83

32

50.

E-mail

addresses:

[email protected]

(A.

Mangen),

[email protected]

(B.R.

Walgermo),

[email protected]

(K.

Brønnick).

Contents

lists

available

at

SciVerse

ScienceDirect

International

Journal

of

Educational

Research

jo

u

r

nal

h

o

mep

age:

w

ww.els

evier.c

o

m/lo

c

ate/ijed

ur

es

0883-0355/$

–

see

front

matter

ß

2012

Elsevier

Ltd.

All

rights

reserved.

http://dx.doi.org/10.1016/j.ijer.2012.12.002

However,

not

all

digital

texts

are

hypertexts.

Hence,

digital

reading

does

not

necessarily

entail

increased

cognitive

load

caused

by

hypertext

features.

Compared

with

the

amount

of

research

on

hypertext

reading,

the

number

of

studies

speciﬁcally

addressing

the

potential

differences

between

sequential

and

continuous

reading

of

linear,

narrative

and

non-

narrative

texts

in

print

and

on

screen

is

small.

Moreover,

as

observed

by

Andrew

Dillon

(1992)

as

well

as

by

Graesser,

Millis,

and

Zwaan

(1997),

there

is

a

lack

of

empirical

studies

using

naturalistic

texts

(narrative

or

non-narrative)

which

are

longer

than

one

or

a

few

paragraphs.

Additionally,

several

early

studies

assessed

reading

comprehension

of

print

and

computer-

mediated

texts

with

the

intention

of

clarifying

factors

of

computer-presentation

of

text

which

facilitate

reading

comprehension

(e.g.,

deﬁnitions

of

key

concepts;

access

to

background

information;

technical

features

of

layout

and

organization)

(Reinking,

1988).

One

study

(Rice,

1994)

employed

two

measures

to

examine

reading

comprehension

of

short

texts

(142

words):

a

text

recall

measure

and

a

highlighting

task

(in

order

to

examine

reading

comprehension

in

a

so-called

‘‘interactive

mode’’).

The

ﬁndings

revealed

no

signiﬁcant

effect

for

presentation

mode

(screen

or

paper)

on

the

recall

measure.

However,

on

the

interactive

measure,

a

signiﬁcant

main

effect

was

found

for

presentation

mode,

with

paper

being

signiﬁcantly

better

than

computer.

The

authors

concluded

that

reading

comprehension

constructs

appears

to

be

the

same

between

computer

presentation

and

paper

presentation

of

text,

but

when

readers

engage

in

a

highlighting

task,

a

signiﬁcant

effect

is

found

for

presentation

medium

(paper

over

computer).

In

a

more

recent

study

(Wa

¨

stlund,

Reinikka,

Norlander,

&

Archer,

2005),

two

experiments

compared

production

(writing)

and

comprehension

performance.

In

the

ﬁrst

experiment,

subjects

read

10-page

PDF

documents

containing

ﬁve

different

texts

each

averaging

1000

words

and

followed

by

a

multiple-choice

test

(more

speciﬁcally,

the

READ-test

designed

to

measure

Swedish

language

reading

comprehension).

In

the

second

experiment,

subjects

read

short

newspaper

articles

with

a

mean

length

of

70

words

and

were

instructed

to

create

an

appropriate

headline

to

each

article

(requiring

rapid

acquisition

and

comprehension

of

verbal

material).

Time

allocated

to

reading

was

limited

in

both

experiments.

The

authors

found

that

in

both

experiments,

performance

in

the

computer

condition

was

inferior

to

that

of

the

paper

presentation

condition,

both

in

terms

of

writing

and

reading

comprehension.

Additionally,

subjects

in

the

computer

reading

condition

reported

higher

levels

of

experienced

stress

and

tiredness

than

those

reading

from

paper.

Hence,

Wa

¨

stlund

et

al.

concluded

that

reading

and

working

with

a

computer

results

in

a

higher

cognitive

workload

compared

with

paper

(Wa

¨

stlund

et

al.,

2005;

cf.

also

Wa

¨

stlund,

2007).

Noyes

and

Garland

have

conducted

a

number

of

studies

on

the

effect

of

presentation

medium

on

comprehension

performance

in

the

ﬁeld

of

ergonomics,

human

factors,

and

design

(Garland

&

Noyes,

2004;

Noyes

&

Garland,

2003,

2008).

In

one

study

(Noyes

&

Garland,

2003),

they

compared

reading

on

a

computer

(VDT

[video

display

terminal])

versus

reading

on

paper

while

measuring

study

and

reading

times,

factual

recall

and

comprehension.

The

texts

were

adapted

from

an

introductory

economics

course,

comprised

of

22

print

pages

and

were

matched

for

size,

color

and

resolution

across

media.

In

order

to

more

accurately

measure

the

potential

impact

of

presentation

medium

on

learning

performance,

Noyes

and

Garland

employed

an

additional

measure

in

conjunction

with

comprehension

scores,

namely

memory

awareness

ratings

in

terms

of

the

Remember–Know

learning

paradigm.

Indicating

the

manner

in

which

the

material

is

cognitively

processed,

assimilated

and

retrieved,

memory

awareness

measures

are

used

by

psychologists

to

gauge

the

nature

and

quality

of

recall

and,

by

implication,

learning.

Originally

developed

by

Tulving

(1985),

the

Remember–Know

paradigm

describes

two

main

types

of

retrieval

response,

Remember

versus

Know.

Knowledge

which

is

Remembered

is

typically

recollected

in

close

association

with

pertaining

to

the

learning

episode.

By

contrast,

knowledge

which

is

Known

is

recalled,

retrieved

and

applied

without

any

such

additional

contextual

associations

and

‘‘thus,

is

information

which

is

simply

based

on

a

certain

sense

of

just

knowing

or

familiarity

[.

.

.].’’

(Noyes

&

Garland,

2003,

p.

415)

One

important

implication

is

that

the

Remember

type

of

memory

is

thus

more

vulnerable

to

fading

with

time,

than

knowledge

which

is

Known.

By

implication,

it

is

assumed

that

knowledge

which

is

Known

is

indicative

of

better

learning

(Conway,

Gardiner,

Perfect,

Anderson,

&

Cohen,

1997;

Herbert

&

Burt,

2001).

In

their

experiment,

Noyes

and

Garland

(2003)

found

no

differences

between

the

two

media

(paper

versus

VDT)

in

terms

of

study

and

reading

times,

and

number

of

correct

answers.

On

the

qualitative

comprehension

measures,

however,

there

was

an

effect

of

presentation

medium,

with

Remember

frequencies

being

almost

twice

that

of

Know

frequencies

in

the

computer

group,

while

Remember

and

Know

response

levels

were

similar

in

the

Paper

group.

Noyes

and

Garland

conclude

that

‘‘characteristics

of

the

computer

screen

(refresh

rate,

high

levels

of

contrast

and

ﬂuctuating

luminance)

interfere

with

cognitive

processing

for

long-term

memory’’

(2003,

p.

420).

These

ﬁndings

were

replicated

in

a

later

study

(Garland

&

Noyes,

2004),

which

showed

lower

Know

levels

in

the

computer

condition

compared

with

the

paper

condition.

This

implies

that

knowledge

transition

from

the

episodic

memory

(indexed

by

Remember

responses)

to

the

semantic

memory

(indexed

by

Know

responses)

appears

to

be

dependent

on

the

nature

of

presentation

format

(screen

versus

paper):

‘‘The

knowledge

transition

was

much

more

rapid

for

those

learning

from

printed

material,

suggesting

less

interference

to

the

process

of

schematization,

and

consequently

more

readily

applied

knowledge.

Hence,

this

suggests

there

still

appears

to

be

a

beneﬁt

attached

to

learning

from

paper-based

rather

than

computer-based

materials’’

(Garland

&

Noyes,

2004,

p.

51).

One

important

implication

of

these

ﬁndings

is

that

knowledge

seems

to

be

better

assimilated

and

more

readily

retrieved

when

presented

in

paper

format.

In

one

of

the

few

studies

assessing

children’s

reading

performance

on

paper

and

screen

(Kerr

&

Symons,

2006),

the

researchers

examined

whether

children’s

reading

rate,

comprehension,

and

recall

were

affected

by

computer

presentation.

Sixty-ﬁfth

grade

students

read

two

expository

texts,

one

in

print

and

one

on

a

computer

screen.

The

texts

were

372

and

411

words

long

and

hence

required

a

limited

amount

of

scrolling

in

the

computer

condition.

Recall

and

comprehension

measures

entailed

using

lists

of

generated

questions

that

tested

recall

of

speciﬁc

information

from

the

texts

as

well

as

the

ability

to

A.

Mangen

et

al.

/

International

Journal

of

Educational

Research

58

(2013)

61–68

62

generate

the

required

inferences

for

appropriate

comprehension.

The

results

indicate

that

children

read

text

more

slowly

on

computers

than

on

paper,

and

they

recalled

more

of

the

information

that

they

had

read

from

the

computer

than

from

paper.

However,

the

children

were

more

efﬁcient

at

comprehending

the

texts

when

reading

from

paper

(comprehension

efﬁciency

calculated

as

a

product

of

accuracy

and

reading

rate).

These

ﬁndings

suggest

that

‘‘while

children,

if

given

enough

time,

may

be

able

to

comprehend

equal

amounts

of

information

from

paper

and

computer,

when

reading

time

is

accounted

for,

children

are

comprehending

less

efﬁciently

when

reading

from

computer’’

(Kerr

&

Symons,

2006,

pp.

13–14).

The

potential

effect

of

presentation

medium

on

reading

comprehension

is

also

an

issue

for

essay

marking

and

annotation.

In

several

countries

there

is

currently

a

shift

toward

assessors

marking

digitally

scanned

copies

on

screen

rather

than

the

original

paper

copies

traditionally

used

(Coniam,

2011;

Johnson,

Hopkin,

&

Shiell,

2011;

Johnson

&

Na

´

das,

2009;

Johnson,

Na

´

das,

&

Bell,

2010).

With

extended

essays

in

particular,

the

potential

effect

of

the

presentation

medium

on

reading

comprehension

becomes

an

issue.

The

ﬁndings

of

one

recent

study

addressing

this

issue

(Johnson

&

Na

´

das,

2009),

suggest

that

examiners

had

a

weaker

recall

of

essay

quality

on

screen

and

had

greater

difﬁculty

recollecting

the

location

of

details

in

these

texts.

Interview

data

support

the

suggestion

that

the

examiners’

comprehension

was

more

challenged

when

reading

from

screen

than

from

paper

(Johnson

&

Na

´

das,

2009).

In

the

present

study,

we

strived

for

a

more

natural

setting

than

that

of

a

laboratory,

employing

classrooms

in

a

test-like

situation,

hence

extending

within

this

ﬁeld

with

regard

to

ecological

validity.

The

stimulus

material

consisted

of

two

authentic

texts

that

are

both

quite

representative

for

their

respective

genres

(i.e.,

narrative

and

expository).

As

we

wanted

the

texts

to

be

as

close

to

identical

as

possible

across

presentation

media,

we

had

subjects

read

PDF

versions

on

computer

screens

and

paper,

respectively.

We

hypothesized

(1)

that

we

would

ﬁnd,

as

in

better

reading

comprehension

when

texts

were

read

on

paper,

and

(2)

that

the

expository

text

would

be

more

affected

by

reading

modality

than

the

narrative,

due

to

the

possible

higher

cognitive

load

introduced

by

the

topic

of

the

text.

An

additional

motivation

for

this

study

is

that

the

scoring

system

employed

is

planned

implemented

in

national

assessments

of

reading

in

Norway.

Hence,

the

study

reported

here

also

has

direct

implications

for

policymaking

and

test

development.

2.

Methods

2.1.

Participants

(n

=

72;

43%

females)

were

10th

grade

students

in

two

urban

primary

schools

in

Norway.

Age

was

not

systematically

assessed,

as

all

subjects

were

selected

to

be

15

or

16

years

old

as

shown

by

class/school

records.

Thus,

we

know

that

the

range

was

15–16

and

hence

mean

and

median

age

was

>15

and

<16.

The

sample

is

middle-class

Caucasian,

and

hence

homogenous

with

respect

to

socio-economic

status

and

ethnicity.

2.2.

Material

and

procedure

All

measures

were

administered

in

the

students’

respective

classrooms

and

carried

out

by

the

authors

and

a

research

assistant.

All

students

received

pretests

in

reading

comprehension,

word

reading

and

vocabulary.

2.2.1.

Reading-comprehension

measures

The

participants

received

a

reading-comprehension

pretest

consisting

of

one

narrative

and

one

expository

text.

Each

text

was

followed

by

multiple

choice

(MC)

items,

and

in

addition

a

few

short-answer

constructed

response

(CR)

questions

(12–20).

One

of

these

texts

was

designed

at

the

National

Centre

for

Reading

Education

and

Research

(The

Reading

Centre)

in

Norway

and

one

text

was

an

example

text

from

OECD

Programme

for

International

Student

Assessment

(PISA)

–

a

study

to

assess

reading

comprehension/literacy

among

15-year-olds

(Lie,

Kjærnsli,

Roe,

&

Turmo,

2001).

The

texts

contained

1400–

1600

words

and

included

graphical

and/or

pictorial

illustrations.

The

format

and

number

of

pages

for

each

text

were

identical

for

on-screen

and

on-paper

presentation.

For

all

texts

used

in

the

present

study

the

questions

were

designed

to

capture

ﬁve

different

aspects

of

reading

comprehension

in

line

with

the

ﬁve

comprehension

processes

used

for

item

development

in

PISA.

As

in

PISA,

it

was

not

possible

to

report

each

of

the

ﬁve

aspects

as

separate

subscales,

therefore

these

ﬁve

aspects

were

organized

into

three

broader

aspect

categories:

(1)

access

and

retrieve,

(2)

integrate

and

interpret

and

(3)

reﬂect

and

evaluate.

The

students

were

allowed

to

look

back

at

the

text

passages

while

answering

the

questions.

Those

who

read

digitally

were

able

to

scroll

up

and

down

the

pages,

and

change

between

the

PDF

and

answering

questions.

2.2.2.

Word

reading

Students’

word

reading

abilities

were

measured

using

a

word-chain

test

(http://www.udir.no/Vurdering/Kartlegging-

videregaende-opplaring/Lesing/).

This

is

a

screening

test

where

the

participants

are

instructed

to

segment

letters

into

their

constituent

words.

Four

words

are

combined

in

each

word

chain.

The

words

are

nouns,

verbs,

adjectives,

adverbs,

prepositions

or

numerals,

and

they

vary

in

length

from

two

to

six

letters.

Each

chain

consists

of

words

which

are

semantically

unrelated

(Anglicized

example:

housecarseafree).

There

are

74

word

chains

in

the

test,

and

the

students

have

a

time

limit

of

A.

Mangen

et

al.

/

International

Journal

of

Educational

Research

58

(2013)

61–68

63

4

min

to

solve

as

many

word-chains

as

possible.

The

reliability

coefﬁcient,

Cronbach’s

alpha,

was

.86

in

the

standardization

sample

(Høien

&

Tønnesen,

2008).

2.2.3.

Vocabulary

A

traditional

single-word-item

semantic

vocabulary

pretest

was

used

to

obtain

information

regarding

students’

vocabulary

(Buch-Iversen,

2010).

The

test

consisted

of

24

items.

Each

item

contained

one

stem-word,

and

the

students’

task

was

to

ﬁnd

the

stem-words

synonym

among

two

distractors.

Anglicized

example:

sword

(stem)

–

bayonet

–

lance

–

sabre

(key).

2.2.4.

Main

survey

Four

weeks

after

the

pretests

were

carried

out,

the

students

received

another

reading

comprehension

test

consisting

of

one

narrative

and

expository

text.

This

test

was

also

designed

at

The

Reading

Centre

and

constructed

to

capture

the

three

aspect

categories

of

reading

comprehension

used

for

item

development

in

PISA.

This

time

the

reading-comprehension-test

was

organized

as

a

randomized

controlled

experiment,

and

the

learners

in

each

class

were

randomized

into

two

groups:

-

Group

1

read

the

texts

digitally/on

screen.

-

Group

2

read

the

texts

on

paper.

Both

groups

answered

questions

digitally/on

screen.

The

time

limit

was

1

h.

All

students

who

took

part

in

the

experiment

submitted

their

tests

within

this

time

limit.

The

reading

comprehension

measures

used

in

the

main

study

were

designed

at

The

Reading

Centre.

With

respect

to

reliability,

all

texts

used

in

this

study,

both

for

pretesting

and

in

the

main

survey

had

Cronbach’s

alpha

>.75.

2.3.

Instruments

The

computer

displays

were

15

‘‘LCD

monitors

operating

at

60

Hz,

at

a

resolution

of

1280



1024

pixels.

The

computers

were

familiar

to

the

subjects

as

it

was

the

computers

the

students

used

in

school

on

a

daily

basis.

The

digitally

presented

texts

were

presented

as

PDF-ﬁles,

read

using

Adobe

Reader

version

9.4

for

Windows

XP.

For

group

2

the

same

texts

were

printed

on

A-4

size

paper

(210

mm



297

mm).

The

font

was

black,

14

point,

Times

new

roman,

at

100%

scale.

The

students

had

Internet

access,

but

were

not

allowed

to

switch

to

any

other

Internet-based

activities

while

taking

the

test.

2.4.

Data

analysis

Normally

distributed

variable

are

presented

as

means

and

standard

deviations

for

descriptive

purposes.

A

sequential

regression

analysis

was

used

to

investigate

differences

between

the

print

reading

condition

and

the

screen

reading

condition.

To

control

for

vocabulary,

word

reading

skill

and

reading

comprehension,

the

variables

were

entered

in

the

ﬁrst

block

of

the

analysis,

followed

by

sex

and

a

binary

dummy

variable

representing

experimental

condition

(paper

versus

on-

screen

reading).

The

regression

analysis

was

checked

for

multicollinearity

using

the

approach

recommended

by

Belsley,

Kuh,

and

Welsch

(1980),

and

for

leverage,

heteroscedacity,

and

normality

of

residuals.

To

investigate

possible

relative

differences

between

effects

of

on-screen

reading

versus

paper

reading

for

the

narrative

versus

expository

text,

we

computed

standardized

scores

for

reading

comprehension

for

both

texts

in

order

to

rescale

the

variables

into

directly

comparable

units.

For

each

subject,

the

difference

score

between

the

standardized

comprehension

score

for

the

expository

text

and

the

narrative

text,

was

computed.

Finally,

differences

in

mean

pretest-scores

(vocabulary

and

word-chain)

and

differences

of

mean

difference

scores

for

reading

on-screen

versus

paper

were

analyzed

using

a

sequential

regression

analysis.

All

analyses

were

conducted

using

SPSS

version

18.03.

3.

Results

A

t-test

for

individual

samples

revealed

that

the

two

groups

did

not

differ

signiﬁcantly

from

each

other

on

the

pretests

(vocabulary:

t(70)

=

.41,

p

=

.88,

word-chain:

t(70)

=

.25,

p

=

.80,

both

t-tests

two-tailed)

indicating

that

the

students

who

read

the

text

on

paper

and

on

screen

were

similar

regarding

reading

comprehension/ability.

Table

1

lists

mean

scores,

standard

deviations

and

effect

sizes

on

the

pretests

for

the

two

groups

of

participants.

Cohen’s

d

effect

sizes

are

considered

as

small

(d



.3),

medium

(d



.5)

or

large

(d



.8)

by

convention

(Cohen,

1988).

Table

2

shows

the

results

of

the

sequential

regression

analyses.

The

model

was

considerably

improved

by

each

step

of

the

analysis,

as

shown

by

the

signiﬁcant

R

2

changes.

When

adding

reading

modality

as

a

predictor

in

block

three,

the

explained

variance

increased

by

4%,

p

=

.025.

The

total

explained

variance

was

42%

in

step

three

with

all

predictors

included

in

the

equation.

In

the

ﬁnal

model,

word-reading

ability

(

b

=

.144,

p

=

.043)

and

vocabulary

(

b

=

.551,

p

=

.009)

were

both

positive

predictors

of

reading-comprehension

scores.

After

controlling

for

the

variance

associated

with

the

variables

vocabulary,

word-reading,

reading

comprehension

and

gender,

the

reading

modality

variable

accounted

for

additional

variance.

Reading

A.

Mangen

et

al.

/

International

Journal

of

Educational

Research

58

(2013)

61–68

64

modality

was

found

to

be

statistically

signiﬁcant

(

b

=

.216,

p

=

.025)

indicating

that

students

who

read

texts

digitally

were

more

likely

to

receive

lower

scores

on

the

reading

comprehension

tests

compared

to

the

students

who

read

the

texts

on

paper.

There

were

no

signiﬁcant

difference

between

the

narrative

text

and

the

expository

text

with

regard

to

reading

modality,

F(1,70)

=

.142,

p

=

.707.

4.

Discussion

The

aim

of

the

present

study

was

to

investigate

the

potential

impact

of

the

reading

modality

on

certain

aspects

of

reading

comprehension.

Based

on

existing

research,

we

predicted

that

subjects

in

the

print

reading

condition

would

perform

better

on

the

reading

comprehension

test

than

subjects

in

the

computer

reading

condition.

Our

ﬁndings

provided

support

for

this

hypothesis.

Subjects

who

read

the

texts

on

paper

performed

signiﬁcantly

better

than

subjects

who

read

the

texts

on

the

computer

screen.

Our

second

prediction

was

that

comprehension

performance

would

be

more

negatively

affected

by

computer

presentation

for

the

expository

text

than

for

the

narrative

text.

This

hypothesis

was

not

supported

by

our

data,

as

there

were

no

differences

for

the

two

types

of

text

genre

(i.e.,

narrative

and

expository)

with

regard

to

reading

modality.

There

are

several

possible

explanations

why,

in

the

present

study,

subjects

in

the

print

condition

scored

signiﬁcantly

higher

on

the

comprehension

tests

than

those

in

the

screen

condition.

Considering

the

length

of

the

stimulus

text,

and

the

fact

that

the

computer

condition

text

was

in

PDF

format,

the

difference

in

comprehension

performance

between

the

print

and

the

computer

group

could

be

issues

of

navigation

within

the

document.

When

reading

on

screen,

scrolling

is

inevitable

unless

the

text

is

within

the

screen

size.

Scrolling

is

known

to

hamper

the

process

of

reading,

by

imposing

a

spatial

instability

which

may

negatively

affect

the

reader’s

mental

representation

of

the

text

and,

by

implication,

comprehension

(Baccino,

2004;

Eklundh,

1992;

Piolat,

Roussey,

&

Thunin,

1997).

By

presenting

the

texts

as

PDFs,

we

intended

to

minimize

the

potentially

negative

effects

of

scrolling.

However,

the

scrolling

option

was

not

excluded

from

our

material.

We

have

no

data

showing

whether

or

not,

and

to

what

extent,

the

students

in

the

computer

condition

used

the

scrolling

option

when

reading

the

texts.

Hence,

it

cannot

be

eliminated

that

students

in

the

computer

condition

scrolled

during

reading,

and

that

this

scrolling

negatively

impacted

their

comprehension

performance.

Another

navigation

issue

is

the

ways

in

which

the

two

types

of

media

determine

and

restrict

one’s

access

to

the

texts

in

their

entirety.

Evidence

suggests

that

readers

often

recall

where

in

a

text

some

particular

piece

of

information

Table

1

Scores

on

the

pretests

for

the

two

groups

of

students

who

participated

in

the

experiment.

Reading

modality

N

Mean

SD

Effect

size

a

Reading

comprehension

Paper

25

19.5

6.3

.02

Computer

47

19.7

8.2

Vocabulary

Paper

25

18.4

3.4

.03

Computer

47

18.5

3.1

Word

reading

Paper

25

54.2

7.2

.06

Computer

47

53.5

12.7

a

Effect-sizes

were

calculated

using

Cohen’s

d.

Cohen’s

d

=

M

1



M

2

/

s

pooled

,

where

s

pooled

¼

ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ

½ð

s

12

þ

s

22

Þ=2

p

.

All

these

effect-sizes

are

small.

Table

2

Sequential

regression

with

reading

comprehension

scores

in

the

main

study

as

the

outcome.

Step

Variable

R

2

R

2

change

F

change

b

a

p-Value

*

1

.35

12.03

<.001

Vocabulary

.26

.019

Word

reading

.31

.005

Reading

comprehension

pretest

.19

.094

2

.37

.02

2.47

.121

Vocabulary

.27

.014

Word

reading

.26

.025

Reading

comprehension

pretest

.19

.098

Sex

.16

.121

3

.42

.04

5.24

.025

Vocabulary

.27

.010

Word

reading

.24

.029

Reading

comprehension

pretest

.19

.082

Sex

.17

.084

Reading

modality

.21

.025

Bold

values

pertain

to

the

omnibus

step-values

in

the

multiple

regression.

a

b

is

standardized.

*

p-Value

is

considered

statistically

signiﬁcant

at

p

<

.05.

A.

Mangen

et

al.

/

International

Journal

of

Educational

Research

58

(2013)

61–68

65

appeared

(e.g.,

toward

the

upper

right

corner

or

at

the

bottom

of

the

page)

(Piolat

et

al.,

1997;

Rothkopf,

1971;

Zechmeister

&

McKillip,

1972).

We

know

from

empirical

and

theoretical

research

that

having

a

good

spatial

mental

representation

of

the

physical

layout

of

the

text

supports

reading

comprehension

(Baccino

&

Pynte,

1994;

Cataldo

&

Oakhill,

2000;

Kintsch,

1998;

Piolat

et

al.,

1997).

For

instance,

Cataldo

and

Oakhill

(2000)

found

that

good

comprehenders

were

signiﬁcantly

better

than

poor

comprehenders

at

remembering

and

relocating

the

order

of

information

in

a

text,

hence

implying

a

relation

between

mental

reconstruction

of

text

structure

and

reading

comprehension.

To

this

effect,

the

ﬁxity

of

text

printed

on

paper

supports

reader’s

construction

of

the

spatial

representation

of

the

text

by

providing

unequivocal

and

ﬁxed

spatial

cues

for

text

memory

and

recall.

In

order

to

appropriately

respond

to

the

multiple

choice

questions,

the

subjects

in

our

study

were

required

to

locate,

access

and

retrieve

essential

pieces

of

information

in

the

two

texts.

It

is

known

that

comprehension

tasks

become

more

difﬁcult

when

the

information

required

to

complete

a

task,

such

as

answering

questions

in

a

reading

comprehension

assessment,

is

not

immediately

visible,

for

instance,

when

the

reader

has

to

integrate

information

occurring

at

locations

in

a

text

which

are

spatially

far

apart

(OECD,

2011).

Such

integration

requires

that

the

reader

has

constructed

a

solid

mental

representation

of

the

structure

of

the

text.

Even

with

the

relatively

short

texts

in

our

study

(4

pages),

it

is

not

unreasonable

to

assume

that

the

intangibility

of

the

digital

text

might

have

challenged

the

readers’

mental

reconstruction

of

the

physical

layout

of

the

text,

which

in

turn

might

have

impeded

their

overview

as

well

as

ability

to

access,

locate

and

retrieve

required

pieces

of

textual

information.

Readers

in

the

paper

condition

had

immediate

access

to

the

text

in

its

entirety.

This

access

is,

moreover,

built

on

both

visual

and

tactile

cues:

the

reader

can

see

as

well

as

tactilely

feel

the

spatial

extension

and

physical

dimensions

of

the

text,

as

the

material

substrate

of

paper

provides

physical,

tactile,

spatiotemporally

ﬁxed

cues

to

the

length

of

the

text

(Mangen,

2006,

2010;

Sellen

&

Harper,

2002).

Readers

in

the

computer

condition,

by

contrast,

were

restricted

to

seeing

(and

sensing)

only

one

page

of

the

text

at

any

given

time

of

reading.

Hence,

their

overview

of

the

organization,

structure

and

ﬂow

of

the

text

(Eklundh,

1992;

Piolat

et

al.,

1997)

might

have

been

hampered

due

to

limited

access

to

the

text

in

its

entirety.

As

noted

by

Kerr

and

Symons

(2006),

‘‘difﬁculties

in

reading

from

computers

may

be

due

to

disrupted

mental

maps

of

the

text,

which

may

be

reﬂected

in

poorer

understanding

and

ultimately

poorer

recall

of

presented

material’’

(p.

5).

Admittedly,

the

potential

role

of

voluminosity

and

thickness

of

paper

versus

screens

is

of

limited

relevance

in

the

present

study

considering

the

length

of

the

texts

in

question.

A

suggestion

for

future

research

could

be

to

replicate

the

ﬁndings

of

our

study

and

have

subjects

read

longer

(and

potentially

more

complex)

texts,

as

it

might

be

assumed

that

challenges

pertaining

to

intratextual

navigation

(i.e.,

navigation

inside

a

textual

document)

might

increase

with

the

length

of

the

text.

The

response

mode

(in

this

case,

responding

to

multiple

choice-questions

online)

might

also

explain

why

subjects

reading

on

paper

performed

better

than

those

reading

on

computer

screens.

Whereas,

in

the

paper

condition,

the

task

of

responding

to

questions

involved

switching

between

the

two

media

(i.e.,

between

the

texts

on

paper

on

the

desk

and

the

questions

on

a

laptop

computer

screen),

subjects

in

the

computer

condition

had

to

switch

between

different

windows

on

the

computer,

one

displaying

the

text

to

be

read,

the

other

displaying

the

questions

to

be

answered.

As

found

also

in

a

(Wa

¨

stlund

et

al.,

2005),

the

dual-task

involved

in

switching

between

two

windows

displayed

on

the

same

computer

screen

is

likely

to

present

additional

cognitive

challenges

to

the

reader.

This

ﬁnding

is

an

illustrating

example

of

how

the

ease

of

multitasking

provided

by

the

computer

might

come

at

a

cognitive

cost,

something

which

is

evidenced

in

a

number

of

studies

(Bowman,

Levine,

Waite,

&

Gendron,

2010;

Fox,

Rosen,

&

Crawford,

2009;

Kirschner

&

Karpinski,

2010;

Lin,

2009;

Lin,

Robertson,

&

Lee,

2009;

Ophir,

Nass,

&

Wagner,

2009).

Another

potential

explanation

might

be

differences

at

a

metacognitive

level.

Metacognition,

or

the

ability

to

monitor

one’s

cognitive

performance,

has

been

shown

to

correlate

with

good

reading

comprehension

(Garner,

1987;

Paris,

Wasik,

&

Turner,

1991).

A

recent

study

by

Ackerman

and

Goldsmith

(2011)

is

particularly

interesting

in

this

regard.

Comparing

reading

performance

from

On-Screen

Learning

(OSL)

and

On-Paper

Learning

(OSL),

the

authors

found

that

under

ﬁxed

study

time

(Experiment

1),

test

performance

did

not

differ

signiﬁcantly.

However,

when

study

time

was

self-regulated

(Experiment

2),

poorer

performance

was

observed

in

screen

reading

than

in

paper

reading.

The

lower

test

performance

of

OSL

was

accompanied

by

signiﬁcant

overconﬁdence

with

regard

to

predicted

performance

(shorter

study

time

+

lower

level

of

actual

learning),

whereas

subjects

in

the

OPL

group

monitored

their

performance

more

accurately.

Ackerman

and

Goldsmith

(2011)

conclude

that

people

appear

to

perceived

the

medium

of

print

as

more

suitable

for

effortful

learning,

whereas

the

electronic

medium

(in

this

case,

a

computer)

is

better

suited

for

‘‘fast

and

shallow

reading

of

short

texts

such

as

news,

e-mails,

and

forum

notes

[.

.

.].

The

common

perception

of

screen

presentation

as

an

information

source

intended

for

shallow

messages

may

reduce

the

mobilization

of

cognitive

resources

that

is

needed

for

effective

self-regulation’’

(p.

29).

In

the

present

study,

we

did

not

measure

the

reading

time.

Hence,

we

cannot

say

if

subjects

in

the

computer

screen

condition

spent

more

time

reading

than

the

subjects

in

the

print

condition.

A

suggestion

for

future

studies

on

digital

reading

is

to

combine

objective

reading

time

measures

with

both

subjective

and

objective

measures

of

reading

performance.

Print

and

screen

media

also

differ

with

respect

to

visual

ergonomics.

Hence,

some

of

the

explanation

for

the

differences

between

screen

reading

and

print

reading

might

relate

to

the

different

lighting

conditions

in

the

two

modalities.

LCD

computer

screens

like

the

ones

used

in

this

study

are

known

to

cause

visual

fatigue

due

to

their

emitting

light.

In

contrast,

e-

book

technologies

based

on

electronic

ink,

such

as

Kindle

and

Kobo,

are

merely

reﬂecting

light

and

are

hence

more

reader

friendly

with

respect

to

the

visual

ergonomics.

Studies

by,

e.g.,

Garland

and

Noyes

(2004)

and

Noyes

and

Garland

(2003),

concluded

that

certain

features

of

the

LCD

screen,

such

as

refresh

rate,

contrast

levels

and

ﬂuctuating

light

interfere

with

cognitive

processing

and

hence

potentially

impair

long-term

memory.

LCD

screen

technologies

differ

from

both

print

and

more

print-like

screen

displays

based

on

electronic

ink

in

that

LCD

screens

as

found

in

most

computers

and

surf

tablets

(such

A.

Mangen

et

al.

/

International

Journal

of

Educational

Research

58

(2013)

61–68

66

as

the

iPad)

are

emitting

light,

whereas

e-book

technologies

based

on

electronic

ink,

such

as

the

Kindle,

are

reﬂecting

ambient

light.

Light-emitting

screens

are

known

to

cause

visual

fatigue

and,

more

speciﬁcally,

computer

vision

syndrome

(Baccino,

2004;

Blehm,

Vishnu,

Khattak,

Mitra,

&

Yee,

2005;

Yan,

Hu,

Chen,

&

Lu,

2008).

The

visuospatial,

perceptual

processes

of

reading,

involving

letter

detection

and

word

identiﬁcation,

depend

crucially

on

the

visual

legibility

of

the

text.

The

visual

legibility

of

the

text,

in

turn,

is

inﬂuenced

by

several

factors,

among

which

the

screen

resolution,

effects

of

backlighting,

and

luminance

contrast

(Dillon

&

Emurian,

1995;

Lee,

Ko,

Shen,

&

Chao,

2011).

The

deteriorating

effects

on

the

visual

processing

of

text

might

in

turn

have

negative

implications

for

higher-level

processes

such

as

comprehension.

It

is

impossible

to

determine

from

the

data

of

the

current

study

whether

visual

fatigue

contributed

to

the

poorer

reading

comprehension

performance

in

the

computer

condition.

Hence,

we

cannot

say

whether

the

visual

ergonomics

of

the

laptop

computer

screens

had

a

negative

impact

on

subjects’

reading.

In

future

studies,

employing

online

measures

including

eye

tracking

methods

would

be

a

way

of

appropriately

addressing

this

important

issue.

5.

Conclusion

The

results

of

this

study

indicate

that

reading

linear

narrative

and

expository

texts

on

a

computer

screen

leads

to

poorer

reading

comprehension

than

reading

the

same

texts

on

paper.

These

results

have

several

pedagogical

implications.

Firstly,

we

should

not

assume

that

changing

the

presentation

format

for

even

short

texts

used

in

reading

assessments

will

not

have

a

signiﬁcant

impact

on

reading

performance.

If

texts

are

longer

than

a

page,

scrolling

and

the

lack

of

spatiotemporal

markers

of

the

digital

texts

to

aid

memory

and

reading

comprehension

might

impede

reading

performance.

Furthermore,

our

results

suggest

that

implementing

both

reading

assessment

tasks

(i.e.,

text

reading

and

response

tasks)

in

the

same

medium

–

the

computer

–

leads

to

additional

cognitive

costs.

Hence,

the

ongoing

digitization

of

response

format

in

the

Norwegian

educational

assessment

system

warrants

extra

consideration

of

important

but

hitherto

largely

neglected

factors

potentially

inﬂuencing

assessment

outcomes,

such

as

challenges

pertaining

to

multitasking

in

a

digital

environment.

References

Ackerman,

R.,

&

Goldsmith,

M.

(2011).

Metacognitive

regulation

of

text

learning:

On

screen

versus

on

paper.

Journal

of

Experimental

Psychology:

Applied,

17(1),

18–

32.

Baccino,

T.

(2004).

La

lecture

e

´

lectronique:

De

la

vision

a

`

la

compre

´

hension.

Grenoble:

Presses

Universitaires

de

Grenoble.

Baccino,

T.,

&

Pynte,

J.

(1994).

Spatial

coding

and

discourse

models

during

text

reading.

Language

and

Cognitive

Processes,

9,

143–155.

Belsley,

D.

A.,

Kuh,

E.,

&

Welsch,

R.

E.

(1980).

Regression

diagnostics:

Identifying

inﬂuential

data

and

sources

of

collinearity.

New

York:

Wiley.

Blehm,

C.,

Vishnu,

S.,

Khattak,

A.,

Mitra,

S.,

&

Yee,

R.

W.

(2005).

Computer

vision

syndrome:

A

review.

Survey

of

Ophthalmology,

50(3),

253–262.

Bowman,

L.

L.,

Levine,

L.

E.,

Waite,

B.

M.,

&

Gendron,

M.

(2010).

Can

students

really

multitask?

An

experimental

study

of

instant

messaging

while

reading.

Computers

and

Education,

54(4),

927–931.

Buch-Iversen,

I.

(2010).

Betydningen

av

inferens

for

leseforsta

˚

else

[The

role

of

inferences

in

reading

comprehension].

Doctoral

thesis,

University

of

Stavanger,

Norway.

Retrieved

from

http://brage.bibsys.no/uis/handle/URN:NBN:no-bibsys_brage_13096.

Cataldo,

M.

G.,

&

Oakhill,

J.

(2000).

Why

are

poor

comprehenders

inefﬁcient

searchers?

An

investigation

into

the

effects

of

text.

Journal

of

Educational

Psychology,

92(4),

791–799.

Cohen,

J.

(1988).

Statistical

power

analysis

for

the

behavioral

sciences

(2

ed.).

Hillsdale,

NJ:

Lawrence

Earlbaum

Associates.

Coniam,

D.

(2011).

A

qualitative

examination

of

the

attitudes

of

Liberal

Studies

markers

towards

onscreen

marking

in

Hong

Kong.

British

Journal

of

Educational

Technology,

42(6),

1042–1054.

Conway,

M.

A.,

Gardiner,

J.

M.,

Perfect,

T.

J.,

Anderson,

S.

J.,

&

Cohen,

G.

(1997).

Changes

in

memory

awareness

during

learning:

The

acquisition

of

knowledge

by

psychology

undergraduates.

Journal

of

Experimental

Psychology:

General,

126(4),

393–413.

DeStefano,

D.,

&

LeFevre,

J.-A.

(2007).

Cognitive

load

in

hypertext

reading:

A

review.

Computers

in

Human

Behavior,

23(3),

1616–1641.

Dillon,

A.

(1992).

Reading

from

paper

versus

screens:

A

critical

review

of

the

empirical

literature.

Ergonomics,

35(10),

1297–1326.

Dillon,

T.

W.,

&

Emurian,

H.

(1995).

Reports

of

visual

fatigue

resulting

from

use

of

video

display

unit.

Computers

in

Human

Behavior,

11(1),

77–84.

Eklundh,

K.

S.

(1992).

Problems

in

achieving

a

global

perspective

of

the

text

in

computer-based

writing.

Instructional

Science,

21(1),

73–84.

Fox,

A.

B.,

Rosen,

J.,

&

Crawford,

M.

(2009).

Distractions,

distractions:

Does

instant

messaging

affect

college

students’

performance

on

a

concurrent

reading

comprehension

task?

Cyberpsychology

and

Behavior,

12(1),

51–53.

Garland,

K.

J.,

&

Noyes,

J.

M.

(2004).

CRT

monitors:

Do

they

interfere

with

learning?

Behaviour

and

Information

Technology,

23(1),

43–52.

Garner,

R.

(1987).

Metacognition

and

reading

comprehension.

Westport,

CT:

Ablex

Publishing.

Graesser,

A.

C.,

Millis,

K.

K.,

&

Zwaan,

R.

A.

(1997).

Discourse

comprehension.

Annual

Review

of

Psychology,

48,

163–189.

Herbert,

D.

M.

B.,

&

Burt,

J.

S.

(2001).

Memory

awareness

and

schematization:

Learning

in

the

university

context.

Applied

Cognitive

Psychology,

15,

617–637.

Høien,

T.,

&

Tønnesen,

G.

(2008).

Instruksjonshefte

til

ordkjedetesten.

[Manual

for

the

word-chain

test].

Bryne:

Logometrica.

Johnson,

M.,

Hopkin,

R.,

&

Shiell,

H.

(2011).

Extended

essay

marking:

Does

the

transition

from

paper

to

screen

inﬂuence

examiners’

cognitive

workload?

International

Journal

of

e-Assessment,

1(1),

107–124.

Johnson,

M.,

&

Na

´

das,

R.

(2009).

Marginalised

behaviour:

Digital

annotations,

spatial

encoding

and

the

implications

for

reading

comprehension.

Learning,

Media

and

Technology,

34(4),

323–336.

Johnson,

M.,

Na

´

das,

R.,

&

Bell,

J.

F.

(2010).

Marking

essays

on

screen:

An

investigation

into

the

reliability

of

marking

extended

subjective

texts.

British

Journal

of

Educational

Technology,

41(5),

814–826.

Kerr,

M.

A.,

&

Symons,

S.

E.

(2006).

Computerized

presentation

of

text:

Effects

on

children’s

reading

of

informational

material.

Reading

and

Writing,

19(1),

1–19.

Kintsch,

W.

(1998).

Comprehension:

A

paradigm

for

cognition.

Cambridge:

Cambridge

University

Press.

Kirschner,

P.

A.,

&

Karpinski,

A.

C.

(2010).

Facebook

1

and

academic

performance.

Computers

in

Human

Behavior,

26(6),

1237–1245.

Lee,

D.-S.,

Ko,

Y.-H.,

Shen,

I.-H.,

&

Chao,

C.-Y.

(2011).

Effect

of

light

source,

ambient

illumination,

character

size

and

interline

spacing

on

visual

performance

and

visual

fatigue

with

electronic

paper

displays.

Displays,

32(1),

1–7.

Lie,

S.,

Kjærnsli,

M.,

Roe,

A.,

&

Turmo,

A.

(2001).

Godt

rustet

for

framtida?

Norske

15-a

˚

ringers

kompetanse

i

lesing

og

realfag

i

et

internasjonalt

perspektiv

[Well

prepared

for

the

future?

Reading

and

writing

skills

of

Norwegian

15-year

olds

in

an

international

perspective].

Oslo:

University

of

Oslo,

Institutt

for

Lærerutdanning

og

Skoleutvikling.

Lin,

L.

(2009).

Breadth-biased

versus

focused

cognitive

control

in

media

multitasking

behaviors.

PNAS:

Proceedings

of

the

National

Academy

of

Sciences

of

the

United

States

of

America,

106(37),

15521–15522.

Lin,

L.,

Robertson,

T.,

&

Lee,

J.

(2009).

Reading

performances

between

novices

and

experts

in

different

media

multitasking

environments.

Computers

in

the

Schools,

26(3),

169–186.

A.

Mangen

et

al.

/

International

Journal

of

Educational

Research

58

(2013)

61–68

67

Mangen,

A.

(2006).

New

narrative

pleasures?

A

cognitive-phenomenological

study

of

the

experience

of

reading

digital

narrative

ﬁctions.

Doctoral

thesis,

Norwegian

University

of

Science

and

Technology

[NTNU],

Trondheim,

Norway.

Retrieved

from

http://urn.kb.se/resolve?urn=urn:nbn:no:ntnu:diva-1833.

Mangen,

A.

(2010).

Point

and

click:

Theoretical

and

phenomenological

reﬂections

on

the

digitization

of

early

childhood

education.

Contemporary

Issues

in

Early

Childhood11(4).

Noyes,

J.

M.,

&

Garland,

K.

J.

(2003).

VDT

versus

paper-based

text:

Reply

to

Mayes,

Sims

and

Koonce.

International

Journal

of

Industrial

Ergonomics,

31,

411–423.

Noyes,

J.

M.,

&

Garland,

K.

J.

(2008).

Computer-

vs.

paper-based

tasks:

Are

they

equivalent?

Ergonomics,

51(9),

1352–1375.

OECD.

(2011).

PISA

2009

results:

Students

on

line:

Digital

technologies

and

performance

(Vol.

VI,

pp.

).

OECD.

Ophir,

E.,

Nass,

C.,

&

Wagner,

A.

D.

(2009).

Cognitive

control

in

media

multitaskers.

PNAS:

Proceedings

of

the

National

Academy

of

Sciences

of

the

United

States

of

America,

106(37),

15583–15587.

Paris,

S.

G.,

Wasik,

B.

A.,

&

Turner,

J.

C.

(1991).

The

development

of

strategic

readers.

In

Barr,

R.,

Kamil,

M.

L.,

Mosenthal,

P.

B.,

&

Pearson,

P.

D.

(Eds.),

Handbook

of

reading

research

(Vol.

II,

pp.

609–640).

New

York:

Longman.

Piolat,

A.,

Roussey,

J.-Y.,

&

Thunin,

O.

(1997).

Effects

of

screen

presentation

on

text

reading

and

revising.

International

Journal

of

Human–Computer

Studies,

47(4),

565–589.

Reinking,

D.

(1988).

Computer-mediated

text

and

comprehension

differences:

The

role

of

reading

time,

reader

preference,

and

estimation

of

learning.

Reading

Research

Quarterly,

23(4),

484–498.

Rice,

G.

E.

(1994).

Examining

constructs

in

reading

comprehension

using

two

presentation

modes:

Paper

vs.

computer.

Journal

of

Educational

Computing

Research,

11(2),

153–178.

Rothkopf,

E.

(1971).

Incidental

memory

for

location

of

information

in

text.

Journal

of

Verbal

Learning

and

Behavior,

10,

608–613.

Sellen,

A.

J.,

&

Harper,

R.

H.

R.

(2002).

The

myth

of

the

paperless

ofﬁce.

Cambridge,

MA:

MIT

Press.

Tulving,

E.

(1985).

Memory

and

consciousness.

Canadian

Psychology,

26,

1–12.

Wa

¨

stlund,

E.

(2007).

Experimental

studies

of

human–computer

interaction:

Working

memory

and

mental

workload

in

complex

cognition.

Go

¨

teborg:

Department

of

Psychology,

Gothenburg

University.

Wa

¨

stlund,

E.,

Reinikka,

H.,

Norlander,

T.,

&

Archer,

T.

(2005).

Effects

of

VDT

and

paper

presentation

on

consumption

and

production

of

information:

Psychological

and

physiological

factors.

Computers

in

Human

Behavior,

21,

377–394.

Yan,

Z.,

Hu,

L.,

Chen,

H.,

&

Lu,

F.

(2008).

Computer

vision

syndrome:

A

widely

spreading

but

largely

unknown

epidemic

among

computer

users.

Computers

in

Human

Behavior,

24(5),

2026–2042.

Zechmeister,

E.

B.,

&

McKillip,

J.

(1972).

Recall

of

place

on

the

page.

Journal

of

Educational

Psychology,

63(5),

446–453.

A.

Mangen

et

al.

/

International

Journal

of

Educational

Research

58

(2013)

61–68

68