The Effect of Attention Cueing on Science Text Learning10.1007...The Effect of Attention Cueing on Science Text Learning Hui-Yu Yang1 • Che-Chang Chang2 Springer Science+Business

ORI GINAL RESEARCH

The Effect of Attention Cueing on Science Text Learning

Hui-Yu Yang1 • Che-Chang Chang2

� Springer Science+Business Media Dordrecht 2015

Abstract The present study examines how different types of attention cueing and cog-

nitive style affect learners’ comprehension of a cardiovascular system and cognitive load.

In this study, the learners were randomly assigned into one of following groups: no-cueing,

static-blood cueing, static-blood–static-arrow cueing, and dynamic-blood–dynamic-arrow

cueing. The results showed that attention cueing yielded similar test results but helped

reduce the learners’ cognitive load. No interaction effects between cognitive style and the

experimental conditions on the learners’ total score and cognitive load were observed. Both

high- and low-visual learners gained equal benefits from attention cueing. However, one

interaction effect in one subtest was observed implying that attention cueing can cause

interference among high-visual learners. Contrary to the hypothesis, the presence of

attention cueing did not enhance conceptual understanding.

Keywords Cognitive style � Cognitive load theory � Attention cueing

1 Introduction

Without proper guidance, learners’ attention might become distracted when learning an

unfamiliar subject in a multimedia environment. The presence of visual signals (i.e.,

arrows, distinctive color, flashing, etc.) is assumed to direct learners’ attention to the most

essential information, help organize that information into a coherent structure, and opti-

mize conceptual understanding (Mayer 2009).

& Hui-Yu [email protected]

Che-Chang [email protected]

1 Present Address: Foreign Language Department, Fujian University of Technology, Fuzhou, China

2 Business Management Department, Fujian University of Technology, Fuzhou, China

123

Tech Know LearnDOI 10.1007/s10758-015-9264-6

RE

TRA

CTE

DA

RTIC

LE

R

http://crossmark.crossref.org/dialog/?doi=10.1007/s10758-015-9264-6&domain=pdf

http://crossmark.crossref.org/dialog/?doi=10.1007/s10758-015-9264-6&domain=pdf

On the other hand, factors relevant to individual differences that might moderate

multimedia learning efficiency include cognitive style (Hegarty et al. 2003; Imhof et al.

2013; Plass et al. 1998) and prior knowledge (Imhof et al. 2013; Yang 2014a). It is

assumed that visual learners depend more on imagery and that if visual imagery is

unavailable, visual learners will not benefit so much from instructional materials (Chen

et al. 2008). However, the lack of visual imagery does not adversely influence verbal

learners’ learning (Plass et al. 1998). Besides, learners’ prior knowledge (Kriz and Hegarty

2007; Imhof et al. 2013) may interact with experimental treatment in affecting learning

efficiency.

According to the signaling principle, attention cueing is assumed to reduce extraneous

load (Mayer and Moreno 2010) and promote learning efficiency. This study expands upon

previous research conducted on the effects of attention cueing in multimedia learning with

the aim of addressing the questions of whether or not the presence of attention cueing can

reduce learners’ cognitive load and the ways in which different types of cognitive style and

attention cueing affect learning efficiency.

2 Literature Review

2.1 Cognitive Load Theory

The information processing that occurs in working memory involves: selection of relevant

words and images, organization of selected words and images, and integration of visual

and verbal information with prior knowledge (Horz and Schnotz 2010; Mayer 2009).

During information processing, poor instructional design may impose extraneous cognitive

load on learners and negatively influence their learning efficiency; however, such load may

be reduced by providing attention cueing (Mayer 2009). The presence of attention cueing is

assumed to direct learners’ attention to the target, thus minimizing the visual search

process, releasing more cognitive resources with which learners can engage in schema

construction and activation, and optimize learning efficiency (de Koning et al. 2009).

2.2 Signaling Principle

In terms of cognitive processing, attention cueing is classified into selection, organization,

and integration cues. Selection cues guide learners’ attention to the most essential infor-

mation in the representations (Crooks et al. 2012; de Koning et al. 2007, 2009). Organi-

zation cues assist learners in organizing the elements of the representations to better

facilitate text processing and improve retention (Crooks et al. 2012; de Koning et al. 2007,

2009). Integration cues help learners integrate the elements between and within the rep-

resentations into a coherent structure (de Koning et al. 2009). Therefore, the presence of

visual or dynamic contrast cues are assumed to draw learners’ attention to the target and

reduce their extraneous cognitive load (de Koning et al. 2009).

2.3 Relevant Studies About Attention Cueing in Multimedia Learning

Crooks et al. (2012) examined the effects of cueing and modality on a self-paced, com-

puter-based diagram depicting places of articulation in human speech. The learners were

classified into one of four conditions: (a) written text with arrow and color cueing,

H.-Y. Yang, C.-C. Chang

123

RE

TRA

CTE

DA

RTIC

LE

R

(b) written text without arrow cueing, (c) spoken text with arrow and color cueing, and

(d) spoken text without arrow cueing. The results indicated a reverse modality effect,

which can be explained more by perceptual than by cognitive resources view. The presence

or absence of cueing yielded no significant differences on the learners in terms of their test

performance and cognitive load.

In a study conducted by Tabbers et al. (2000), the learners were given one of four

treatments: (a) visual text without cues in the diagram, (b) visual text with cues in the

diagram, (c) audio text without cues in the diagram, and (d) audio text with cues in the

diagram. Providing cues did not reduce learners’ extraneous load but yielded similar

cognitive load under all conditions. Imhof et al. (2013) also explored the effects of arrow

cueing on learning fish locomotion patterns. The learning conditions included: (a) multiple

visualizations without arrows, (b) multiple visualizations with arrows, and (c) single

visualization with arrows. The first and third conditions were beneficial in facilitating

learning efficiency by comparing multiple pictures or making dynamic information

explicit. The second condition appeared to cause interference and hinder learning. The

ineffectiveness of cueing on animation might be due to interference caused by the

simultaneous highlighting of multiple elements without specificity (Moreno 2007).

Kriz and Hegarty (2007) also conducted a study to probe the effects of arrow cueing on

learning a flushing cistern. The learners who received arrow cues did not significantly

outperform those who did not receive arrow cues in comprehension and troubleshooting

tests. The authors suggest that presenting attention cueing may help learners focus their

attention only on the most essential elements, but without guaranteeing effective con-

ceptual understanding and mental model constructions of the visual representations.

Additional activities accompanied with cues may assist learners engage in deep learning

(de Koning et al. 2009). de Koning and his colleagues conducted several studies by

decreasing the luminance of uncued subsystems to show their visual contrast with cued

subsystems in an animated cardiovascular system. In one study regarding presentation

speed (de Koning et al. 2011b), the learners showed similar performances on retention and

transfer tests regardless of cueing conditions and display speeds. Furthermore, those in

low-speed conditions experienced a higher cognitive load than did those in high-speed

conditions. This was probably due to the fact that in the low-speed conditions, learners had

to integrate and keep the information active in their working memory for a longer period of

time which generated a greater extraneous load as compared with the learners in the high-

speed conditions. In addition, the other two studies concerning the self- or instructional

explanations accompanied with attention cueing (de Koning et al. 2010b, 2011a) yielded

mixed results. Self- or instructional explanations accompanied with visual cueing seemed

to enhance the learners’ conceptual understanding of the causal relations of animated

cardiovascular system, yielded better performance, and reduced cognitive load. However,

in terms of efficiency, the effects of self- or instructional explanations were unclear. On the

other hand, in their other studies, they found positive effects of attention cueing on learning

efficiency as demonstrated by learners’ performances on transfer and inference tasks (de

Koning et al. 2007, 2010b, 2011a).

In another study, Boucheix et al. (2011) provided learners with four types of cues when

learning a piano mechanism by assigning them into groups involving: (a) entity cueing,

(b) localized coordinate cueing, (c) progressive path cueing, and (d) no cueing. The

learners in the localized coordinate cueing and progressive path cueing conditions out-

performed those in the entity cueing and non-cueing groups on dependent measures.

Cueing helped the learners have a better idea about the operation of a piano mechanism.


123

RE

TRA

CTE

DA

RTIC

LE

R

Amadieu et al. (2011) examined the effects of attention cueing on learning an animated

long-term potentiation. Learners received the presence or absence of zooming in the salient

element in an animation with repeated or unrepeated exposure respectively. The presence

of attention cueing helped guide the learners’ attention and comprehension of the salient

elements and ignore the peripheral, irrelevant elements. Repeated exposure to the cued

animation enhanced the learners’ understanding of causal relations and reduced their

perceived difficulties.

Kalyuga et al. (1999) compared the effects of conventional separate-diagram-and-text

and color-coded-diagram-and-text situations on learning an electrical circuit. The con-

ventional group was given an electrical circuit with a written text underneath, whereas the

color-coded-diagram-and-text group was presented with the same diagram and text but

with additional color cueing on the electrical elements in which unique coloring schemes

appeared when the learners clicked on the text. Those in the color-cueing condition showed

better test performance and lower cognitive load than did those in the conventional con-

dition. When dealing with split-attention diagrams where the text and diagrams are pre-

sented simultaneously, the text should be marked with color-cueing that draws the learners’

attention.

In sum, studies investigating the supposed benefits of visual attention cueing on mul-

timedia learning efficiency have demonstrated inconsistent results.

2.4 Cognitive Style: Visualizers Versus Verbalizers

Learners’ cognitive style have also been thought to moderate learning efficiency. Learners

can learn better when the presentation mode suits their learning style. Studies conducted by

Chen et al. (2008), Leutner and Plass (1998), and Plass et al. (1998) have indicated that

visual representations benefit visual learners more due to their strong visuospatial capa-

bilities in constructing mental models. However, Hegarty et al. (2003), Imhof et al. (2013),

Jones (2009), and Plass et al. (2003) have found that visual representations may not support

individual learner differences. On the other hand, Hoffler (2010) holds spatial-ability-as-

compensator hypothesis and suggests that dynamic visual representations can compensate

for low-visual learners’ weak visuospatial capabilities and support them more. In sum, the

debate over whether visual learners benefit more from visual representations is

controversial.

2.5 Statement of the Problem

The issue over whether providing attention cueing enhances learning efficiency (e.g.,

Amadieu et al. 2011; Boucheix et al. 2011; de Koning et al. 2007, 2010b, 2011a; Kalyuga

et al. 1999) or fails to optimize learning (e.g., Crooks et al. 2012; de Koning et al. 2010a,

2011b; Harp and Mayer 1998; Tabbers et al. 2000) remains controversial. Besides, pre-

vious studies used diagrams/animation alone, but the presentation of visual imagery

without verbal explanations may be insufficient for learners to understand abstract con-

cepts. Furthermore, the ability to construct mental imagery is relevant to one’s visuospatial

abilities as well as prior knowledge (Hegarty et al. 2003). To address the unresolved

questions, the research questions in the present study are as follows:

1. Do learners perform differently on dependent measures in different learning

conditions?


123

RE

TRA

CTE

DA

RTIC

LE

R

H1: Those who receive attention cueing can optimize their conceptual understanding and

perform better on dependent measures than those who do not receive attention cueing (e.g.,

Crooks et al. 2012; de Koning et al. 2009, 2011a).

H2: Those who receive dynamic contrast cues can enhance their conceptual under-

standing and perform better on dependent measures than those who only receive visual

contrast cues (de Koning et al. 2009).

H3: Those who receive relation cues can optimize their conceptual understanding and

perform better on transfer tasks than those who do not receive relation cues (de Koning

et al. 2011a).

2. Do learners in different conditions experience different cognitive loads?

H4: Those who receive attention cues will reduce their visual search processing and

experience lower cognitive loads (de Koning et al. 2009), whereas those who do not

receive attention cues will exert more cognitive resources in constructing mental imagery

and experience higher cognitive load.

3. Do cognitive style and experimental treatment affect learners’ performance and

cognitive load?

H5: High visual learners will perform well in either cued or uncued conditions, but low

visual learners are assumed to benefit more from cued conditions (e.g., Hegarty et al. 2003;

Yang 2014b). Furthermore, there may not be any significant interaction between learning

style and experimental treatment (Hegarty et al. 2003; Imhof et al. 2013; Yang 2014b).

3 Methodology

3.1 Participants

The participants were comprised of 169 undergraduates (male = 31, female = 138) with

an average age of 19 (M = 19.30, SD = 0.91) enrolled in college of humanities at a

science and technology university in southeastern China. None of them had the background

of biology, nor were they familiar with the material in the present study.

3.2 Data Collection Instruments

51 freshmen in college of humanities participated in the pilot study. The reliability of each

measurement was as follows:

3.2.1 Prior Knowledge Test (Cronbach’s a = 0.83)

A prior knowledge questionnaire with four statements was first administered to assess

participants’ background (de Koning et al. 2007, 2011a). The learners self-rated their

understanding of blood circulation by marking on a nine-point scale measuring their

responses to the statements such as ‘‘My understanding of a cardiovascular system is…’’

and ‘‘My interest in reading books and magazines about medical science is…’’. A one-way

ANOVA revealed no significant differences among the four groups, F(3,165) = 1.164,

p[ 0.05.


123

RE

TRA

CTE

DA

RTIC

LE

R

3.2.2 Cognitive Style Measurement (Cronbach’s a = 0.86)

The learners’ cognitive styles were identified using the index of learning styles ques-

tionnaire, developed by Felder and Soloman (1997). The questionnaire comprised 44

alternative-choice questions. Only the visual and verbal scales in the index of learning

styles questionnaire were considered in the present study. The strength of the style was

indicated by an index ranging from 1 to 11 with 1 representing the lowest level and 11

representing the highest level. Learners with a rating at or above index 5 on the visual scale

were classified as high-visual learners. Those with a rating at index 1 on the visual scale

were classified as low-visual learners. A one-way ANOVA revealed no significant dif-

ferences among the four groups, F(3,165) = 0.509, p[ 0.05.

3.2.3 Retention Test (Cronbach’s a = 0.77)

The retention test, comprised of 13 multiple-choice questions, was designed to assess how

well the learners understood the instructional materials. Item 8 dealt with the structure of

the heart; items 3, 4, and 11 were relevant to heartbeats; items 2, 9, and 10 focused on the

functions of valves; items 1, 5, and 6 focused on contractions; and items 7, 12, and 13 dealt

with blood circulation. The students had to choose the best answer among the four answer

choices in each question (Fig. 1). Each correct answer was worth one point.

Point-biserial correlation was conducted to eliminate the weak test items which were

less reliable in discriminating between high and low level learners (Wu and Tu 2006).

Following item analyses, 2 items were removed and 11 items were retained.

3.2.4 Pictorial Recall Test (Cronbach’s a = 0.78)

The pictorial recall test, comprised of ten static pictures, was aimed at examining the

learners’ comprehension of the instructional materials. Each multiple-choice question was

comprised of one picture with four answer choices. Items 2, 9, and 10 dealt with the

contraction of ventricles; items 1, 4, 5, and 6 were concerned with how blood returns from

Fig. 1 Sample screenshot of the retention test


123

RE

TRA

CTE

DA

RTIC

LE

R

the body and collects in the atria; items 3, 7, and 8 dealt with the contraction of the atria.

The students had to choose the best answer among the four alternatives to describe the

picture (Fig. 2). Each correct answer was worth one point.

Point-biserial correlation was conducted to eliminate the weak test items. Following

item analyses, 3 items were removed and 7 items were retained.

3.2.5 Matching Test (Cronbach’s a = 0.79)

One static diagram concerning long and short loops was used to examine whether the

learners could apply what they learned and indicate where blood flows in the human body.

There were twelve general terms that needed to be matched with corresponding parts in the

diagram. Items 1–4 dealt with the structure of the heart; items 5–8 were relevant to body

parts; and items 9–12 dealt with how and where blood exchanges oxygen in the human

body. The students were required to match the correct term with a corresponding body part

in the diagram (Fig. 3). Each correct mark received one point and each ambiguous mark

received no point. Point-biserial correlation was conducted as item analyses and all items

were preserved.

3.2.6 Identification Test (Cronbach’s a = 0.96)

One static diagram regarding long and short loops was to examine whether the learners

could apply what they learned and mark the correct steps in the circulatory system on the

diagram. There were ten blanks that needed to be filled into show the steps in the blood

circulation process. The learners had to mark the steps from 1 to 10 on the diagram

(Fig. 4). Each correct mark was worth one point and ambiguous marks (i.e., random steps,

scribble, etc.) received no points. Point-biserial correlation was conducted as item analyses

and all items were preserved.

Fig. 2 Sample screenshot of the pictorial recall test


123

RE

TRA

CTE

DA

RTIC

LE

R

3.2.7 Cognitive Load Questionnaire (Cronbach’s a = 0.86)

A subjective cognitive load questionnaire was developed by the researcher, and it followed

the pattern in the subjective cognitive load rating scale (Paas et al. 1994). Item 1 dealt with

germane load (i.e., to understand the cardiovascular system, the mental effort I had

expended was …). Items 2–3 dealt with intrinsic load (i.e., the difficulty level of the

cardiovascular system is…). Items 4–6 dealt with extraneous load (i.e., how difficult was it

for you to concentrate on reading the text and viewing the pictures in the visual presen-

tations; and how clear and understandable were the text and pictures in the visual pre-

sentations). Items 7–10 dealt with performance load and included questions such as ‘‘How

Fig. 3 Sample screenshot of the matching test

Fig. 4 Sample screenshot of the identification test


123

RE

TRA

CTE

DA

RTIC

LE

R

difficult was it for you to answer the retention test?’’, ‘‘How difficult was it for you to

answer the pictorial recall test?’’, ‘‘How difficult was it for you to answer the matching

test?’’, and ‘‘How difficult was it for you to answer the identification test?’’.

A Pearson correlation analysis revealed a significant result on critical ratio and item-

total correlation. Bartlett’s test of sphericity was significant and Kaiser–Meyer–Olkin

Measure of Sampling Adequacy (KMO) was 0.87. The eigenvalue was greater than 1. The

total explained variance was greater than 50 %, implying that the construct validity of the

rating scale was good.

3.3 Instrumentation

The texts and pictures regarding a cardiovascular system were adopted from Knowledge—

Encyclopedia, published by Dorling Kindersley Inc. (2013). The text (359 words) and

pictures (concerning heartbeat cycles) were modified into PowerPoint slides. The

instructional materials included information about: (1) blood circulation, (2) heartbeats, (3)

the structure of the heart, (4) the functions of the valves, and (5) the heartbeat cycle. A time

counter was above each slide to control presentation time, and each slide was presented

only once for 40 s. The overall presentation lasted for 6 min.

In the fifth section, the written text was accompanied by three pictures illustrating: (1)

filling up the atria; (2) contraction of the atria; and (3) contraction of the ventricles. The

remaining sections contained written text only without pictorial illustrations.

Except for the introduction slide, the instructional materials comprised an average of

about 35 words on each slide. In the fifth section, three pictures depicted each of the three

steps involved in how blood circulates in and out of the heart. All the control and

experimental groups were shown these three pictures in section five of the instructional

materials. For the no-cueing group (NCG), the slides contained written text plus static

pictures without blood and arrow cues (Fig. 5). In the static-blood cueing group (SBG), the

slides contained written text plus static pictures along with static blood cues embedded in

the illustrations (Fig. 6). In the static-blood-and-static-arrow cueing group (SBSAG), the

Fig. 5 A sample screenshot used in the no-cueing group


123

RE

TRA

CTE

DA

RTIC

LE

R

slides contained written text and static pictures embedded with static blood and arrow cues

indicating the path and direction of blood flows (Fig. 7). In the dynamic-blood–dynamic-

arrow cueing group (DBDAG), the slides contained written text with static pictures, but

with dynamic blood and arrow cues indicating the movement path and direction of blood

flow (Fig. 8). In the DBDAG, the dynamic arrows and blood were triggered by clicking the

mouse and appeared gradually on the static diagrams to indicate how blood flows in and

out of the heart. In the DBDAG, when the dynamic feature was in a resting state, the

number and position of dynamic arrows and blood in the diagrams were the same as those

in the SBSAG, except that the dynamic cues in the DBDAG were played three times to

help learners capture the transiency of the animation.

Fig. 6 A sample screenshot used in the static-blood cueing group

Fig. 7 A sample screenshot used in the static-blood–static-arrow cueing group


123

RE

TRA

CTE

DA

RTIC

LE

R

3.4 Experimental Procedures

The experiment was conducted during the students’ regular class period in a language

laboratory containing 60 student seats and a computerized teacher control system from

which the teacher could control the computer system and monitor all the students. The

researcher sat at the computer system to control the presentation, as well as turn on/off the

computer monitors. The overall system was controlled by the researcher, not by the stu-

dents themselves.

First, the researcher gave students instructions regarding: (1) how to answer prior

knowledge questionnaire, (2) how to answer the cognitive style measurement, (3) how to

participate in multimedia activities pertaining to a cardiovascular system, (4) how to

answer the retention, pictorial recall, matching, and identification tests, (5) what they were

not allowed to do during the tests, and (6) how to complete the cognitive load

questionnaire.

Prior to conducting the experiment itself, the students first filled out the prior knowledge

questionnaire. Secondly, they completed the index of learning styles questionnaire.

Thirdly, they received the instrumentation. Fourthly, they received the retention, pictorial

recall, matching, and identification tests sequentially on each student’s computer monitor

at their seat. The students needed to respond by writing down their answer choices on an

answer sheet. They were not permitted to return to previous questions (de Koning et al.

2011a) to reduce the possibility of making inferences from them. They were also not

allowed to return to previously-presented instructional materials, talk to their peers, or use

a dictionary while taking the tests. However, they were permitted to complete the tests at

their own rate. Finally, they completed a self-rated cognitive load questionnaire. After

completing the tests and questionnaires, they handed in their answer sheets and left the

laboratory. The data from the experimental and control groups were collected in separate

class periods.

Fig. 8 A sample screenshot used in dynamic-blood–dynamic-arrow cueing group


123

RE

TRA

CTE

DA

RTIC

LE

R

3.5 Data Analysis Instrument

The significant level was set at .05. A one-way ANOVA was used to analyze the differ-

ences in all the dependent measures among the four groups. A two-way multivariate

analysis of variance (MANOVA) was used to analyze interaction effects between exper-

imental treatment and cognitive style on all of the dependent measures. A Pearson cor-

relation was used to analyze the correlation between test results and cognitive load.

4 Results

4.1 Research Question 1: Do Learners Perform Differently on DependentMeasures in Different Learning Conditions?

The learners’ test performances are shown in Table 1. The results of the one-way ANOVA

indicated no significant differences among the four groups on the retention test,

F(3,165) = 0.573, p = 0.633; the pictorial recall test, F(3,165) = 1.986, p = 0.118; the

matching test, F(3,165) = 1.531, p = 0.208; the identification test, F(3,165) = 0.133,

p = 0.941; and the total score, F(3,165) = 0.802, p = 0.494.

4.2 Research Question 2: Do Learners in Different Conditions ExperienceDifferent Cognitive Loads?

The learners’ germane, intrinsic, extraneous and performance load ratings are shown in

Table 2. The results of the one-way ANOVA using Tukey HSD as a post hoc test indicated a

statistically significant difference regarding germane load, F(3,165) = 5.586, p = 0.001.

Those in the SBG (M = 7.51, SD = 1.74) had a higher germane load than did those in the

SBSAG (M = 6.19, SD = 2.04), p = 0.007; and DBDAG (M = 6.05, SD = 2.08),

p = 0.002. There was a statistically significant difference regarding intrinsic load,

F(3,165) = 3.097, p = 0.028. Those in the NCG (M = 13.05, SD = 2.79) had a higher

intrinsic load than did those in the DBDAG (M = 11.00, SD = 3.42), p = 0.018. There was

no statistically significant difference concerning extraneous load, F(3,165) = 0.939,

p = 0.423. There was no statistically significant difference in retention performance load,

F(3,165) = 0.944, p = 0.421. There was a statistically significant difference in pictorial

performance load, F(3,165) = 5.604, p = 0.001. Those in the NCG (M = 7.15, SD = 1.29)

had a significantly higher pictorial performance load than did those in the SBG (M = 6.14,

SD = 1.78), p = 0.039, SBSAG (M = 5.98, SD = 1.93), p = 0.011, and DBDAG

(M = 5.70, SD = 1.74), p = 0.001. There was a statistically significant difference regarding

matching performance load, F(3,165) = 3.59, p = 0.015. Those in the NCG (M = 7.40,

SD = 1.46) had a significantly higher matching performance load than did those in the

SBSAG (M = 6.19, SD = 2.09), p = 0.019; and DBDAG (M = 6.26, SD = 1.99),

p = 0.031. There was a statistically significant difference in identification performance load,

F(3,165) = 4.76, p = 0.0035. Those in the NCG (M = 7.35, SD = 1.88) had a significantly

higher identification performance load than did those in the SBSAG (M = 5.72, SD = 2.30),

p = 0.001. There was also a statistically significant difference in overall cognitive load,

F(3,165) = 3.858, p = 0.011. Those in the NCG (M = 64.88, SD = 9.34) had a signifi-

cantly higher overall cognitive load than did those in the SBSAG (M = 57.40, SD = 13.83),


123

RE

TRA

CTE

DA

RTIC

LE

R

Table

1T

est

per

form

ance

su

nd

erv

ario

us

con

dit

ion

s

Tes

tN

CG

3S

BG

4S

BS

AG

5D

BD

AG

6

M(N

7)

SD

M(N

)S

DM

(N)

SD

M(N

)S

D

Ret

enti

on

Hig

h1=

6.8

9(1

9)

2.1

8H

igh=

6.0

0(2

0)

1.8

1H

igh=

7.3

3(2

1)

2.2

2H

igh=

6.2

1(2

4)

2.0

9

Lo

w2=

6.1

4(2

1)

2.5

2L

ow

=6

.52

(23

)2

.15

Lo

w=

6.1

8(2

2)

2.5

6L

ow

=6

.11

(19

)2

.08

To

tal=

6.5

02

.36

To

tal=

6.2

81

.99

To

tal=

6.7

42

.44

To

tal=

6.1

62

.06

Pic

tori

alH

igh=

3.0

0(1

9)

2.3

1H

igh=

2.1

0(2

0)

1.7

4H

igh=

3.2

4(2

1)

2.1

4H

igh=

2.9

2(2

4)

2.0

6

Lo

w=

2.3

3(2

1)

2.1

3L

ow

=1

.96

(23

)1

.75

Lo

w=

2.9

5(2

2)

2.3

8L

ow

=1

.95

(19

)2

.05

To

tal=

2.6

52

.21

To

tal=

2.0

21

.73

To

tal=

3.0

92

.25

To

tal=

2.4

92

.00

Mat

chin

gH

igh=

3.2

1(1

9)

1.9

3H

igh=

2.1

0(2

0)

2.1

5H

igh=

2.8

1(2

1)

2.6

6H

igh=

3.3

3(2

4)

2.7

0

Lo

w=

3.5

2(2

1)

2.9

4L

ow

=2

.70

(23

)1

.58

Lo

w=

2.6

8(2

2)

2.8

5L

ow

=3

.42

(19

)3

.04

To

tal=

3.3

82

.45

To

tal=

2.4

21

.87

To

tal=

2.7

42

.73

To

tal=

3.3

72

.82

Iden

tifi

cati

on

Hig

h=

6.0

0(1

9)

4.0

3H

igh=

5.0

5(2

0)

3.7

3H

igh=

5.2

9(2

1)

4.5

5H

igh=

3.2

5(2

4)

3.8

8

Lo

w=

3.8

1(2

1)

4.1

8L

ow

=3

.70

(23

)4

.34

Lo

w=

3.6

8(2

2)

3.8

8L

ow

=5

.79

(19

)4

.26

To

tal=

4.8

54

.20

To

tal=

4.3

34

.08

To

tal=

4.4

74

.25

To

tal=

4.3

74

.20

To

tal

sco

reH

igh=

19

.11

(19

)6

.76

Hig

h=

15

.25

(20

)6

.30

Hig

h=

18

.67

(21

)8

.73

Hig

h=

15

.71

(24

)7

.99

Lo

w=

15

.81

(21

)7

.22

Lo

w=

14

.87

(23

)6

.31

Lo

w=

15

.50

(22

)8

.71

Lo

w=

17

.26

(19

)7

.29

To

tal=

17

.38

7.1

1T

ota

l=

15

.05

6.2

3T

ota

l=

17

.05

8.7

6T

ota

l=

16

.40

7.5

4

1H

igh

vis

ual

lear

ner

s,2

low

vis

ual

lear

ner

s,3

no

-cu

ein

gg

rou

p,

4st

atic

-blo

od

cuei

ng

gro

up,

5st

atic

-blo

od–st

atic

-arr

ow

cuei

ng

gro

up,

6dynam

ic-b

lood–dynam

ic-a

rrow

cuei

ng

gro

up

,7

num

ber

of

par

tici

pan

ts


123

RE

TRA

CTE

DA

RTIC

LE

R

Table

2C

og

nit

ive

load

su

nd

erv

ario

us

con

dit

ion

s

Lo

adN

CG

SB

GS

BS

AG

DB

DA

G

MS

DM

SD

MS

DM

SD

Ger

man

eH

igh=

7.2

11

.62

Hig

h=

7.1

01

.74

Hig

h=

6.1

91

.97

Hig

h=

6.2

12

.06

Lo

w=

6.5

71

.50

Lo

w=

7.8

71

.69

Lo

w=

6.1

82

.15

Lo

w=

5.8

42

.14

Intr

insi

cH

igh=

13

.62

3.3

4H

igh=

11

.90

2.9

9H

igh=

12

.24

3.3

2H

igh=

11

.17

3.4

9

Lo

w=

12

.52

2.1

4L

ow

=1

2.2

62

.88

Lo

w=

11

.09

3.4

4L

ow

=1

0.7

93

.43

Ex

tran

eou

sH

igh=

16

.68

4.0

8H

igh=

14

.00

4.8

8H

igh=

15

.57

4.2

1H

igh=

15

.57

4.4

1

Lo

w=

17

.10

2.9

3L

ow

=1

5.7

04

.50

Lo

w=

15

.73

4.1

5L

ow

=1

6.1

13

.70

Ret

enti

on

load

Hig

h=

6.2

61

.88

Hig

h=

5.7

01

.69

Hig

h=

6.2

42

.17

Hig

h=

5.7

91

.93

Lo

w=

6.0

51

.83

Lo

w=

5.6

11

.67

Lo

w=

5.8

21

.94

Lo

w=

5.3

21

.73

Pic

tori

allo

adH

igh=

7.3

21

.20

Hig

h=

6.2

02

.02

Hig

h=

6.1

42

.33

Hig

h=

5.8

81

.70

Lo

w=

7.0

01

.38

Lo

w=

6.0

91

.59

Lo

w=

5.8

21

.50

Lo

w=

5.4

71

.81

Mat

chin

glo

adH

igh=

7.6

31

.57

Hig

h=

6.7

02

.03

Hig

h=

6.4

32

.54

Hig

h=

6.5

42

.02

Lo

w=

7.1

91

.37

Lo

w=

6.5

21

.80

Lo

w=

5.9

51

.56

Lo

w=

5.8

91

.94

Iden

tifi

cati

on

load

Hig

h=

7.7

91

.75

Hig

h=

6.5

52

.01

Hig

h=

5.8

62

.63

Hig

h=

6.4

21

.84

Lo

w=

6.9

51

.94

Lo

w=

6.2

61

.96

Lo

w=

5.5

91

.59

Lo

w=

6.1

11

.70

To

tal

load

sH

igh=

66

.53

9.4

8H

igh=

59

.20

12

.60

Hig

h=

58

.67

14

.58

Hig

h=

57

.75

12

.95

Lo

w=

63

.38

9.1

8L

ow

=6

0.9

61

0.5

9L

ow

=5

6.1

81

3.3

1L

ow

=5

5.5

31

3.0

2


123

RE

TRA

CTE

DA

RTIC

LE

R

p = 0.027 and DBDAG (M = 56.77, SD = 12.88), p = 0.013. Those in the NCG reported

experiencing a higher cognitive load than did their counterparts in the SBSAG and DBDAG.

4.3 Research Question 3: Do Cognitive Style and Experimental TreatmentAffect Learners’ Performance and Cognitive Load?

A Pearson correlation was conducted to analyze the correlation between each subtest and

subcategory of cognitive load. There was no significant negative correlation between the

retention test and retention load (r = -0.125, p = 0.104). There was a significant negative

correlation between the pictorial test and pictorial performance load (r = -0.110,

p = 0.037). There was no significant negative correlation between the matching test and

matching performance load (r = -0.107, p = 0.464). There was a significant negative

correlation between the identification test and identification load (r = -0.173, p = 0.025).

There was a significant negative correlation between total score and overall cognitive load

(r = -0.202, p = 0.008). As learners’ cognitive load decreased, their test performances

increased, and vice versa.

A two-way multivariate variance (MANOVA) was conducted to examine the interactive

effects between the experimental condition and cognitive style on the four subtests and the

total score (Table 3). The ANOVA source of variation results indicated no interaction on

the retention test, F(3,161) = 1.174, p = 0.321, partial g2 = 0.021; on the pictorial recall

test, F(3,161) = 0.352, p = 0.788, partial g2 = 0.007; or on the matching test,

F(3,161) = 0.161, p = 0.923, partial g2 = 0.003.

However, interaction effects were found on the identification test, F(3,161) = 2.887,

p = 0.037, partial g2 = 0.051. The one-way ANOVA and the follow-up contrasts com-

paring both high- and low-visual learners in the four experimental conditions showed that

the high-visual learners in the NCG (M = 6.00, SD = 4.03) significantly outperformed

their counterpart in the DBDAG (M = 3.25, SD = 3.88), t(80) = 2.207, p = 0.030.

However, the low-visual learners in the DBDAG (M = 5.79, SD = 4.26) scored higher

than their counterparts in the NCG (M = 3.81, SD = 4.18), SBG (M = 3.70, SD = 4.34),

and SBSAG (M = 3.68, SD = 3.88), but did not reach the significance level, p[ 0.05.

There were no interaction effects on the total score, F(3,161) = 1.033, p = 0.380,

partial g2 = 0.019. In all the above MANOVA analyses, no main effects in regard to

learning styles and experimental conditions were observed.

In addition, a two-way multivariate variance (MANOVA) was conducted to examine

the interactive effects between the experimental conditions and cognitive style in each

subcategory and overall cognitive load (Table 4). The ANOVA source of variation results

indicated no interaction effects in regard to germane load, F(3,161) = 1.108, p = 0.347,

partial g2 = 0.020. However, a main effect of the experimental treatment was statistically

significant, F(3,161) = 5.477, p = 0.001, partial g2 = 0.093. Results of post hoc test

using Tukey HSD indicated that those in the SBG (M = 7.51, SD = 1.74) had a signifi-

cantly higher germane load than did those in the SBSAG (M = 6.19, SD = 2.04),

p = 0.007, and DBDAG (M = 6.05, SD = 2.08), p = 0.002.

There were no interaction effects in regard to intrinsic load, F(3,161) = 0.540,

p = 0.655, partial g2 = 0.010. However, a main effect of the experimental treatment was

statistically significant, F(3,161) = 3.182, p = 0.026, partial g2 = 0.056. Results of post

hoc test using Tukey HSD indicated that those in the NCG (M = 13.05, SD = 2.79) had a

significantly higher intrinsic load than did those in the DBDAG (M = 11.00, SD = 3.42),

p = 0.018.


123

RE

TRA

CTE

DA

RTIC

LE

R

There were no interaction effects in regard to extraneous load, F(3,161) = 0.188,

p = 0.905, partial g2 = 0.003. There were no interaction effects regarding performance

load, F(3,161) = 0.081, p = 0.970, partial g2 = 0.002. However, a main effect of the

experimental treatment was statistically significant, F(3,161) = 4.352, p = 0.006, partial

g2 = 0.075. Results of post hoc test using Tukey HSD indicated that those in the NCG

(M = 28.05, SD = 4.93) had a significantly higher performance load than did those in the

SBSAG (M = 23.91, SD = 7.24), p = 0.015, and DBDAG (M = 23.81, SD = 5.99),

p = 0.012.

There were no interaction effects regarding the overall cognitive load, F(3,161) = 0.501,

p = 0.682, partial g2 = 0.009. However, a main effect of the experimental treatment was

statistically significant, F(3,161) = 3.103, p = 0.028, partial g2 = 0.055. Results of post

hoc test using Tukey HSD indicated that those in the NCG (M = 64.87, SD = 9.34) had a

significantly higher overall cognitive load than did those in the SBSAG (M = 57.40,

SD = 13.83), p = 0.028 and DBDAG (M = 56.77, SD = 12.88), p = 0.014.

5 Discussion and Conclusions

First, the results did not support hypothesis one. The total score and those in each subtest

indicated that those who received attention cueing failed to promote learning efficiency and

did not outperform those who did not receive attention cueing. The results contradicted the

Table 3 MANOVA on style 9 group on dependent measures

Wilke’s k(Sig.)

Source Dependentvariable

Type III sumof squares

df MS F Sig. g2p

.921 (.363) Group Retention 9.308 3 3.103 .632 .595 .012

Pictorial 25.605 3 8.535 2.018 .114 .036

Matching 29.493 3 9.831 1.538 .207 .028

Identification 6.557 3 2.186 .129 .943 .002

Total 139.956 3 46.652 .837 .475 .015

.958 (.148) Style Retention 5.784 1 5.784 1.179 .279 .007

Pictorial 11.167 1 11.167 2.640 .106 .016

Matching 1.981 1 1.981 .310 .578 .002

Identification 17.864 1 17.864 1.055 .306 .007

Total 73.375 1 73.375 1.317 .253 .008

.912 (.255) Group 9 style Retention 17.279 3 5.760 1.174 .321 .021

Pictorial 4.470 3 1.490 .352 .788 .007

Matching 3.080 3 1.027 .161 .923 .003

Identification 146.622 3 48.874 2.887 .037* .051

Total 172.641 3 57.547 1.033 .380 .019

Error Retention 161 4.906

Pictorial 161 4.230

Matching 161 6.392

Identification 161 16.930

Total 161 55.715

* p\ 0.05; ** p\ 0.01; *** p\ 0.001


123

RE

TRA

CTE

DA

RTIC

LE

R

explanations of perceptual and cognitive resources (Crooks et al. 2012) but somewhat

echoed the results of previous studies (e.g., Crooks et al. 2012; de Koning et al. 2011b;

Kriz and Hegarty 2007; Moreno 2007). One possible explanation for this discrepancy is

that merely providing attention cueing may only direct learners’ attention to the essential

information without guaranteeing that learners can construct accurate mental representa-

tions and enhanced conceptual understanding (de Koning et al. 2009; Harp and Mayer

1998; Kriz and Hegarty 2007). Secondly, providing verbal text may be sufficient for

learners to construct mental imagery (Plass et al. 2003), with the addition of visual rep-

resentations with attention cueing being redundant. When the verbal and visual repre-

sentations presented the same information, the learners applied cognitive resources to

process both the visual and verbal information and left the remaining resources unavailable

for helpful information processing. Therefore, the information presented might be redun-

dant (Hegarty et al. 2003; Imhof et al. 2013).

Secondly, hypothesis two was not supported. The learners who received dynamic

contrast cueing did not significantly outperform those who received visual contrast cues.

The results somewhat echoed the results of previous studies (e.g., Hegarty et al. 2003;

Tversky et al. 2002) in which animation was not superior to static diagrams in promoting

learning efficiency. Possibly learners focused more on the salient dynamic-blood and

dynamic-arrow cues and less on the verbal text, resulting in limited integration of the

visual and verbal representations. A second possible explanation is that the transiency of

animation caused interference (e.g., Hegarty et al. 2003). The learners had to visually

Table 4 MANOVA on style 9 group on cognitive load

Wilke’sk (Sig.)

Source Dependentvariable

Type III sumof squares

df MS F Sig. g2p

.799 (.000) Group Germane 57.934 3 19.311 5.477 .001*** .093

Intrinsic 94.991 3 31.664 3.182 .026* .056

Extraneous 40.851 3 13.617 .931 .427 .017

Perform 503.120 3 167.707 4.352 .006** .075

Total 2183.619 3 727.873 3.103 .028* .055

.949 (.078) Style Germane .157 1 .157 .044 .833 .000

Intrinsic 13.536 1 13.536 1.360 .245 .008

Extraneous 12.932 1 12.932 .884 .349 .005

Perform 88.320 1 88.320 2.292 .132 .014

Total 72.457 1 72.457 .309 .579 .002

.970 (.964) Group 9 style Germane 11.726 3 3.909 1.108 .347 .020

Intrinsic 16.123 3 5.374 .540 .655 .010

Extraneous 8.244 3 2.748 .188 .905 .003

Perform 9.356 3 3.119 .081 .970 .002

Total 352.702 3 117.567 .501 .682 .009

Error Germane 161 3.526

Intrinsic 161 9.950

Extraneous 161 14.633

Perform 161 38.536

Total 161 234.594

* p\ 0.05; ** p\ 0.01; *** p\ 0.001


123

RE

TRA

CTE

DA

RTIC

LE

R

switch back and forth between the verbal text and animation which may have caused them

to miss some information (Hegarty et al. 2003; Johnson and Mayer 2010).

Thirdly, hypothesis three was not supported. Those who received relation cues generally

did not significantly outperform those who were not presented with relation cues in the

matching and identification tests. Besides, by specifically analyzing learners’ performances

in each item of each subtest, those receiving relation cues could more easily understand

how blood flows between the heart and the rest of the body, as well as where blood absorbs

oxygen from the lungs. However, they were still unable to indicate the steps in blood

circulation. The presence of relation cueing did not optimize conceptual understanding.

The possible explanation was that the diagrams in the test were completely different from

the pictures in the instructional material, so the learners perhaps felt difficult to transfer

what they had learned in their attempt to figure out the flows of blood through the body.

Fourthly, hypothesis four was supported. Those in the NCG had significantly higher

intrinsic, germane, performance and overall cognitive loads than did those in the SBSAG

and DBDAG. Attention cueing slightly reduced the learners’ extraneous load, but helped

reduce even more their intrinsic, germane, and performance loads. The results are some-

what in line with the studies of Amadieu et al. (2011) and Kalyuga et al. (1999), which

showed that the presence of visual or dynamic contrast cues can help reduce learners’

cognitive loads.

Additionally, those in the NCG did not report having a higher cognitive load when

taking the retention test, which simply required them to recall what they had learned

without requiring them to convert texts into images. However, they generally reported

having a higher cognitive load when answering the pictorial recall, matching, and iden-

tification tests, all of which involved pictures. The pictures used in the dependent measures

were completely different from those in the instructional materials. The learners had no

images to retrieve from the instructional materials when answering the imagery-based

questions which required them to convert verbal texts into mental images. Therefore,

performing these tasks was evidently more mentally demanding.

Finally, the results partially supported hypothesis five. There were no interaction effects

between the experimental treatment and cognitive style in regard to the total score and

cognitive load, implying that both the high- and low-visual learners benefited equally well

and experienced similar mental load from attention cueing. However, further examination

of the learners’ performance on each subcategory of the tests reveals that attention cueing

caused interference for the high-visual learners but probably helped compensate for the

low-visual learners. The high-visual learners in the NCG significantly outperformed those

in the DBDAG on the identification test. The low-visual learners in the DBDAG had higher

scores on the identification test compared to those in the other three conditions but the

difference did not reach the significance level. Besides, the low-visual learners in the

DBDAG had dramatically superior score on the identification test than the high-visual

learners in the same condition. Probably the dynamic contrast cues providing external

representations helped the low-visual learners build up mental models (Hoffler 2010),

develop greater conceptual understanding, and perform better (Hoffler and Leutner 2011;

Hoffler 2010; Mayer 2009). On the other hand, since the high-visual learners had strong

cognitive abilities that better enabled them to construct mental animation, the dynamic

contrast cues were likely redundant and caused interference. Only one interactive effect

indicating visual representation with dynamic cueing was redundant for the high-visual

learners, but whether this result can be generalized to explain other forms of visual learning

is uncertain. These findings are largely consistent with those of Hegarty et al. (2003) and

Imhof et al. (2013), in which no interactive effects between experimental treatment and


123

RE

TRA

CTE

DA

RTIC

LE

R

cognitive style were found. There was almost no evidence to indicate that attention cueing

favors high-visual learners.

In sum, regardless of learners’ cognitive style, the presence of attention cueing yielded

similar effects among all the learners, while reducing their cognitive load.

6 Limitations for the Present Research

The present study was unable to incorporate eye-tracking techniques to trace learners’

visual searching processes while reading the visual representations and answering the tests.

Besides, what the learners were thinking and what silent self-explanations (de Koning et al.

2010b, 2011a) were made while answering the questions were unable to be recorded.

Acknowledgments This work was partially funded by the Higher Education Research and DevelopmentCenter of FJUT (GB-G-14-31). The researchers are also grateful to partners for providing the learningequipment and instrumentation.

References

Amadieu, F., Marine, C., & Laimay, C. (2011). The attention-guiding effect and cognitive load in thecomprehension of animations. Computers in Human Behavior, 27, 36–40.

Boucheix, J.-M., Low, R. K., Putri, D. K., & Groff, J. (2011). Cueing animation: Dynamic signaling aidsinformation extraction and comprehension. Learning and Instruction, 25, 71–84.

Chen, N. S., Hsieh, S. W., & Kinshuk, A. (2008). Effects of short term memory and content representationtype on mobile language learning. Language Learning and Technology, 12(3), 93–113.

Crooks, S. M., Cheon, J., Inan, F., Ari, F., & Flores, R. (2012). Modality and cueing in multimedia learning:Examining cognitive and perceptual explanations for the modality effect. Computers in HumanBehavior, 28, 1063–1071.

de Koning, B. B., Tabbers, H. K., Rikers, R. M. J. P., & Paas, F. (2007). Attention cueing as a means toenhance learning from an animation. Applied Cognitive Psychology, 21, 731–746.

de Koning, B. B., Tabbers, H. K., Rikers, R. M. J. P., & Paas, F. (2009). Towards a framework for attentioncueing in instructional animations: Guidelines for research and design. Educational PsychologyReview, 21, 113–140.

de Koning, B. B., Tabbers, H. K., Rikers, R. M. J. P., & Paas, F. (2010a). Attention guidance in learningfrom a complex animation: Seeing is understanding? Learning and Instruction, 20, 111–122.

de Koning, B. B., Tabbers, H. K., Rikers, R. M. J. P., & Paas, F. (2010b). Learning by generating vs.receiving instructional explanations: Two approaches to enhance attention cueing in animations.Computers & Education, 55, 681–691.

de Koning, B. B., Tabbers, H. K., Rikers, R. M. J. P., & Paas, F. (2011a). Improved effectiveness of cueingby self-explanations when learning from a complex animation. Applied Cognitive Psychology, 25,183–194.

de Koning, B. B., Tabbers, H. K., Rikers, R. M. J. P., & Paas, F. (2011b). Attention cueing in aninstructional animation: The role of presentation speed. Computers in Human Behavior, 27, 41–45.

Dorling Kindersley. (2013). Knowledge—Encyclopedia. London: Dorling Kindersley Inc.Felder, R. M., & Soloman, B. A. (1997). Index of learning styles questionnaire. Retrieved from http://www.

engr.ncsu.edu/learningstyles/ilsweb.htmlHarp, S. F., & Mayer, R. E. (1998). How seductive details do their damage: A theory of cognitive interest in

science learning. Journal of Educational Psychology, 90(3), 414–434.Hegarty, M., Kriz, S., & Cate, C. (2003). The roles of mental animations and external animations in

understanding mechanical system. Cognition and Instruction, 21(4), 325–360.Hoffler, T. N. (2010). Spatial ability: It’s influence on learning, with visualizations—A meta-analytic

review. Educational Psychology Review, 22, 245–269.Hoffler, T. N., & Leutner, D. (2011). The role of spatial ability in learning from instructional animations—

Evidence for an ability-as-compensator hypothesis. Computers in Human Behavior, 27, 209–216.


123

RE

TRA

CTE

DA

RTIC

LE

R

http://www.engr.ncsu.edu/learningstyles/ilsweb.html

http://www.engr.ncsu.edu/learningstyles/ilsweb.html

Horz, H., & Schnotz, W. (2010). Cognitive load in learning with multiple representations. In J. L. Plass, R.Moreno, & R. Brunken (Eds.), Cognitive load theory (pp. 229–252). New York, NY: CambridgeUniversity Press.

Imhof, B., Scheiter, K., Edelmann, J., & Gerjets, P. (2013). Learning about locomotion patterns: Effectiveuse of multiple pictures and motion indication arrows. Computers & Education, 65, 45–55.

Johnson, C. I., & Mayer, R. E. (2010). Applying the self-explanation principle to multimedia learning in acomputer-based game-like environment. Computers in Human Behavior, 26, 1246–1252.

Jones, L. C. (2009). Supporting student difference in listening comprehension and vocabulary learning withmultimedia annotation. CALICO Journal, 26(2), 267–289.

Kalyuga, S., Chandler, P., & Sweller, J. (1999). Managing split-attention and redundancy in multimediainstruction. Applied Cognitive Psychology, 13, 351–371.

Kriz, S., & Hegarty, M. (2007). Top–down and bottom–up influences on learning from animation. Inter-national Journal of Human–Computer Studies, 65, 911–930.

Leutner, D., & Plass, J. L. (1998). Measuring learning styles with questionnaires versus direct observation ofpreferential choice behavior in authentic learning situations: The visualizer verbalizer behaviorobservation scale (VV-BOS). Computers in Human Behavior, 14(4), 543–557.

Mayer, R. E. (2009). Multimedia learning (2nd ed.). Cambridge, MA: Cambridge University Press.Mayer, R. E., & Moreno, R. (2010). Techniques that reduce extraneous cognitive load and manage intrinsic

cognitive load during multimedia learning. In J. L. Plass, R. Moreno, & R. Brunken (Eds.), Cognitiveload theory (pp. 131–152). New York, NY: Cambridge University Press.

Moreno, R. (2007). Optimizing learning from animations by minimizing cognitive load: Cognitive andaffective consequences of signaling and segmentation methods. Applied Cognitive Psychology, 21,765–781.

Paas, F., van Merrienboer, J. J. G., & Adam, J. J. (1994). Measurement of cognitive load in instructionalresearch. Perceptual Motor and Skills, 79(1), 419–430.

Plass, J. L., Chun, D. M., Mayer, R. E., & Leutner, D. (1998). Supporting visual and verbal learningpreferences in a second-language multimedia learning environment. Journal of Educational Psy-chology, 90(1), 25–36.

Plass, J. L., Chun, D. M., Mayer, R. E., & Leutner, D. (2003). Cognitive load in reading a foreign languagetext with multimedia aids and the influence of verbal and spatial abilities. Computers in HumanBehavior, 19, 221–243.

Tabbers, H. K., Martens, R. L., & van Merrienboer, J. J. G. (2000). Multimedia instructions and cognitiveload theory: Split-attention and modality effects. Paper presented at the ACET 2000 in Long Beach,California. Online Retrieved on May, 22, 2014.

Tversky, B., Morrison, J. B., & Betrancourt, M. (2002). Animation: Can it facilitate? International Journalof Human–Computer Studies, 57, 247–262.

Wu, M. L., & Tu, J. T. (2006). SPSS and statistical application analysis. Taipei: Wu Nan.Yang, H.-Y. (2014a). The effects of advance organizers and subtitles on EFL learners’ listening compre-

hension skills. CALICO Journal, 31(3), 345–373.Yang, H.-Y. (2014b). Does multimedia support individual differences? EFL learners’ listening compre-

hension and cognitive load. Australasian Journal of Educational Technology, 30(6), 699–713.


123

RE

TRA

CTE

DA

RTIC

LE

R

Documents

The Effect of Attention Cueing on Science Text Learning10.1007...The Effect of Attention Cueing on Science Text Learning Hui-Yu Yang1 • Che-Chang Chang2 Springer Science+Business