Öztop, F., & Nayci, Ö. (2021). Does the digital generation
comprehend better from the screen or from the

paper?: a meta-analysis. International Online

Journal of Education and Teaching (IOJET), 8(2).

1206-1224.

Received : 17.10.2020

Revised version received : 20.01.2021

Accepted : 05.02.2021

DOES THE DIGITAL GENERATION COMPREHEND BETTER FROM

THE SCREEN OR FROM THE PAPER?: A META-ANALYSIS

Research article

Feyyaz Öztop

Sirnak University

feyyazoztop@gmail.com

Ömer Nayci

Sirnak University

nayciomer@gmail.com

Feyyaz Öztop is a lecturer at Sirnak University Program of Child Development. He is
currently continuing his doctorate education at the Gazi University Educational Sciences
Institute Primary Education Department.

Ömer Nayci is an Assistant Professor at Sirnak University Program of Child Development.
He received Ph.D. in the Department of Curriculum Development at Ankara University.

Material published and so copyrighted may not be published elsewhere without written permission of IOJET.

mailto:feyyazoztop@gmail.com
mailto:nayciomer@gmail.com
https://orcid.org/0000-0002-3462-145X
https://orcid.org/0000-0002-6087-6456

International Online Journal of Education and Teaching (IOJET) 2021, 8(2), 1206-1224.

1207

DOES THE DIGITAL GENERATION COMPREHEND BETTER FROM

THE SCREEN OR FROM THE PAPER?: A META-ANALYSIS

Feyyaz Öztop

feyyazoztop@gmail.com

Ömer Nayci

nayciomer@gmail.com

Abstract

This research was carried out to determine whether the comprehension level differs

according to the reading environment by examining the studies comparing the comprehension

level according to reading from the screen and from the paper by using meta-analysis method.

Meta-analysis method was used in this research. The data of the study were obtained from 12

studies (29 comparisons). All of these studies are studies in which Turkish texts are used in

reading. The analysis of the obtained data was carried out using the Comprehensive Meta-

Analysis software. In the interpretation of the studies, random effects model was taken as basis.

As a result of the analyses performed, a significant and medium effect size (g=-0.423, p=0.003)

was found in favor of reading from the paper. Therefore, this result shows that comprehension

achievement is higher in reading from the paper than reading from the screen. In addition, it

was determined that the effect sizes did not differ significantly according to the publication

year, grade level, text type and digital tool. It is recommended to perform studies on improving

the screen interface and screen reading skills.

Keywords: Digital generation, screen, paper, reading comprehension, meta-analysis.

1. Introduction

The process of change and transformation occurring in the context of technology has

also affected education as well as every aspect of life. It can be said that the developments in

technology have brought new orientations in education, especially in the learning and teaching

process, and this has brought along many new tools, equipment and applications. Mentioned

developments, the development of artificial intelligence as well as its application have had

significant effects in the field of education, and the developments in this context have led to

the emergence of many new possibilities (Wang, Yu, Hu, & Li, 2020). As a matter of fact,

today, educators and students widely use many applications related to artificial intelligence

(Taşçı & Çelebi, 2020). As a natural consequence of developments in question, the nature of

face-to-face teaching has changed, blended and online courses have become widespread

rapidly (Redmond, 2011). This has brought various dimensions to the interaction process that

takes place in the context of technology supported learning-teaching (Hillman, Willis, &

Gunawardena, 1994; Moore, 1989).

Interaction can be between individuals and technological tools as well as between

individuals. In the online environment, despite the difference in place, communication can be

provided in audio, written and visual form in case of need (Kaya, 2002), and among the

participants knowledge and experiences can be shared. These sharings are transferred to the

online environment so that information sharing can reach large masses. While these sharings

were performed with means such as books, magazines and encyclopedias, today in line with

the developments in information and communication technologies, they have started to be

transfered to the digital environment in different ways. Digital reading has also increased, as

information on demand has also become widespread (Kol & Schcolnik, 2000). It is seen that

mailto:feyyazoztop@gmail.com
mailto:nayciomer@gmail.com

Öztop & Nayci

1208

many concepts are used in the literature in the light of developments in information access and

acquisition of information from digital tools. One of these is the concept of "screen reading".

Screen reading is an active process in which the individual creates new meanings from

the information presented on the screen and structures it in his mind (Güneş, 2016). This

reading medium which is rapidly becoming widespread in every field differs from the paper

reading. In screen reading, a part of writing which is half of a paper page is presented one after

another on the screen, and the reader tries to understand this information which s/he gets piece

by piece. For this reason, screen reading is also called “piecemeal reading” (Güneş, 2009).

Screen reading is based on the continuous interaction of three components: reader, text

and environment. The most active component of this process is the reader. The reader performs

various operations such as seeing, perceiving, vocalizing, understanding, associating,

questioning, and structuring in the mind the writings on the screen (Güneş, 2010). Since the

reader, in other words, the learner who actively takes advantage of the process is the most

active component; information, media and technology skills have an important place among

the 21st century skills that s/he should have (Partnership for 21st Century Learning (P21),

2013). As a result of the active use of the mentioned skills in the teaching and learning process,

the learner acquires new information, accordingly s/he creates his cognitive synthesis and

continue his/her development in a universal, life oriented manner. For this, the individual

frequently encounters computers and its derivatives in the context of the screen.

It can be said that digital texts are preferred more than printed sources in terms of

facilitating access to information and accessing more resources apart from the traditional

understanding. This situation becomes more evident especially when physical access is limited.

Computer technology and software provided opportunities for transferring printed lecture

notes, articles and textbooks from a paper page to the computer screen. Thus, it provides the

learner with choice and flexibility in many subjects such as presenting a quality option in terms

of duplication of the text and sharing it with others (Spencer, 2006).

Reading from the screen is a type of reading that emerges due to the development of

computer and internet technology and changes as the computer and internet technology

develops (Duran & Alevli, 2014). Reading from the screen differs from reading from printed

text in terms of skills and processes. All the researchers conducted to date revealed that screen

reading has important differences from the paper reading in terms of eye movements,

comprehending and mental processes. These differences affect reading, understanding and

structuring in mind processes and require new skills and techniques (Güneş, 2010). Increasing

use of digital material for learning purposes has many effects on printed sources and screen

(Martin & Platt, 2001). Many studies have been conducted in the related literature. Especially

in the 1980s and early 1990s, researches on reading performance from monitors belonging to

image technologies focused mostly on speed and accuracy measurements and focused less on

comprehension (Garland & Noyes, 2004). Today, in the studies carried out in this field both at

home and abroad; it is seen that more focus is given on issues such as reading comprehension,

sense-making, reading speed and attitude towards text, reading skills, reading preferences and

behaviors, reading style, its effect on motivation, its effect on bodily movements (Başaran,

2014; Chen, Cheng, Chang, Zheng, & Huang, 2014; Ertuğrul, 2013; Kurata, Ishita, Miyata, &

Minami, 2017; Mangen, Walgermo, & Brønnick, 2013; Mizrachi, 2015; Rajanen, Salminen, &

Ravaja, 2016; Schilhab, Balling, & Kuzmicova, 2018; Uther, Ross, Randell, & Pye, 2019;

Yaman & Dağtaş, 2013).

As a result of the increase in transition to studying on the screen in recent years, one of

the issues that researchers have emphasized is whether the reading comprehension is affected

by the change in the reading environment (Delgado & Salmerón, 2021; Halamish & Elbaz,

International Online Journal of Education and Teaching (IOJET) 2021, 8(2), 1206-1224.

1209

2019; Kazazoğlu, 2020; Lebedeva, Veselovskaya, & Kupreshchenko, 2020; Støle, Mangen, &

Schwippert, 2020; Zou & Ou, 2020). The most important purpose of reading is understanding

(Mellard, Fall, & Woods, 2010; Özen & Ertem, 2014). Determining the extent to which the

understanding is achieved is important in terms of both determining the distance taken by the

learner as a result of the education and teacher's seeing the level reached from the beginning of

the process (Ünal & Köksal, 2007). Although the skill of reading comprehension is a subject

that is emphasized in language lessons, it also has an important place in all academic and daily

life of individuals. It has been revealed that there are relationships between reading

comprehension skill in Turkish, social studies, science and technology, and mathematics

classes and success in classes (Yılmaz, 2011), critical thinking skills (Aloqaili, 2011; Hosseini,

Khodaei, Sarfallah, & Dolatabadi, 2012). It has been highlighted that reading is a basic skill on

which sound job opportunities and personal autonomy are based as well as academic success

(Calhoon, 2005). It is an undeniable fact that reading comprehension is of great importance for

a student or an adult whenever reading takes place, since the skill of reading comprehension is

used continuously to obtain information (Tamsi, Zuhri, & Kurniasih, 2013).

In this century, in which digital technology has become widespread in all areas of our

lives, individuals read from the screen rather than from printed sources. It is getting harder day

by day for individuals to leave the screen and return to paper due to habits. Therefore, the

importance of understanding what is read on the screen has increased even more. Thus, the

need to evaluate comprehension success in screen reading and to take steps in this direction

has increased. As a result of this, it has been a subject to be considered whether comprehension

success is higher when reading from screen or from printed material. Although many studies

have been conducted to enlighten this issue, many compilation and meta-analysis studies have

also been carried out due to the different results which have been obtained in different studies.

Clinton (2019), Delgado, Vargas, Ackerman, and Salmerón, (2018) and Kong, Seo, and Zhai

(2018) reached the conclusion that reading on paper is better than reading from the screen in

terms of reading comprehension. However, Wang, Jiao, Young, Brooks, and Olson (2008)

found that studies on the tests in paper-pencil form and computer form have no significant

effect on reading achievements according to the environment. In the study comparing multiple

choice tests applied on computer and on printed text by Kingston (2008), they emphasized that

the results differed, while some findings pointed to the advantages of printed text, the others

emphasized that the advantages of digital text were at the forefront. In this respect, it is stated

that the results of such studies are not always consistent. In another study by Singer and

Alexander (2017), the results showed that the environment played an effective role under

certain text and conditions or for certain readers. Besides, it is emphasized that in terms of the

studies included in the research; there is an increase in the studies conducted and in the variety

of digital devices used, the research environment and the choice of participants are constant

and in terms of text length and text manipulations there are three types of tendencies.

Although there are studies analyzing reading comprehension on screen and paper by

compiling comparative studies in the international arena, the absence of a meta-analysis study

on studies comparing the level of reading comprehension on screen and paper in Turkish texts

has led to the need for doing this research. This study conducted is considered significant in

terms of putting forth the difference in the comprehension level according to the reading

environment and guiding the steps to be taken and the studies to be carried out on this subject.

In this direction, the study was conducted in order to examine whether the comprehension level

differs according to reading from the screen and reading from the paper using the meta-analysis

method.

In the context of this purpose, answers to the following questions were sought within

the scope of the research:

Öztop & Nayci

1210

 Does the level of comprehension differ from reading on screen and reading on paper?

 Is there a significant difference between effect sizes according to publication year?

 Is there a significant difference between effect sizes according to grade level?

 Is there a significant difference between the effect sizes according to the type of text
used in reading?

 Is there a significant difference between the effect sizes according to the digital tool
used in reading?

2. Method

2.1. Research Design

In this research, meta-analysis method was used to examine whether the comprehension

level differs according to screen reading and paper reading or not. Meta-analysis is the

classification of similar studies about a subject, theme or field of study according to certain

criteria and the interpretation of these studies by gathering the quantitative findings (Dinçer,

2014). Meta-analysis is a way to generalize the results (Ellis, 2010). Meta-analysis is a

statistical analysis in which different study findings are combined and analysed again (Glass,

1976).

2.2. Data Collection Procedure

The data of the research were collected in April 2020. TÜBİTAK ULAKBİM (The

Scientific and Technological Research Council of Turkey / Turkish Academic Network and

Information Center), Web of Science, Ebschost, YÖK (Council of Higher Education) National

Thesis Center, Google Academic platforms and databases were scanned by using "Screen",

"Paper", "Printed", "Digital", "Reading", "Comprehension" and "Reading Comprehension"

keywords and their English meanings. The studies listed as a result of the scanning were

evaluated according to the determined inclusion criteria. The determined inclusion criteria can

be listed as follows;

 In the study, reading comprehension level according to reading from the screen
and reading from the paper (printed material) should be compared.

 Reading studies should be done by Turkish texts

 The study should be an article, a master's thesis or a doctoral thesis.

 The study should be conducted with students showing normal development.
Studies conducted with students having special needs are not included in the meta-

analysis.

 Studies should have the necessary statistical information for meta-analysis such as
sample size, average, standard deviation etc.

As a result of the examinations, 16 studies were determined in accordance with the

inclusion criteria. However, 4 of these studies were found to be articles produced from the

thesis. In this case, only articles were included in the meta-analysis and related theses were

eliminated. Finally, 12 studies were included in the meta-analysis. 10 of these studies are

articles and 2 of them are master's theses. 29 comparisons from these studies were examined.

2.3. Data Analysis

A coding form was created by the researchers for the analysis of the studies determined

according to the inclusion criteria. In this coding form, the publication year of the studies, grade

level, text used in reading, digital tool used in reading and statistical information are included.

International Online Journal of Education and Teaching (IOJET) 2021, 8(2), 1206-1224.

1211

In order to avoid confusion regarding the 5th grade due to the change in the education system

after 2012, the grade levels were described as primary education by its name in the old system

instead of elementary school - middle school. Manual calculation of effect sizes and

homogeneity is error-prone and time consuming (Cooper, 2017). So, all analyses of the study

were done using Comprehensive Meta-Analysis software. This software is a comprehensive

meta-analysis software allowing many tests required for meta-analysis to be performed.

Comprehensive Meta-Analysis software has many advantages such as being able to work with

many different data types, ease of use, automatic calculation of effect sizes, editing forest plot,

exporting graphics (Borenstein, 2005; Borenstein, Hedges, Higgins, & Rothstein, 2009).

In the study, analyses were carried out for evaluating publication bias. Publication bias

occurs when published study results do not represent the overall completed research results

(Littell, Corcoran, & Pillai, 2008; Rothstein, Sutton, & Borenstian, 2005). Meticulous selection

of the studies to be included and the definition of the hidden studies provide protection against

publication bias (Lipsey & Wilson, 2001). The more evidence is found about whether there is

bias or not, the decision to be made in this regard will be more accurate (Card, 2012). Hence,

it was decided whether the studies included in meta-analysis have publication bias or not by

using the Funnel chart, Rosenthal's Safe N method, Orwin's Safe N method, Begg and

Mazumdar rank correlation test and Egger regression test.

It is also important on which model the effect size of the studies will be interpreted. In

meta-analysis, two models are used: fixed effects and random effects model. In case of the

distribution of the effect sizes of the studies being homogeneous, the fixed effects model is

used, and in case of the distribution being heterogeneous, the random effects model is used

(Borenstein et al., 2009; Ellis, 2010). The fixed effects model is based on the assumption that

the effect size values of each study are the same, and the random effects model is based on the

assumption that the effect size values of the studies are different (Ellis, 2010). While

determining the model, Q, I2 and p values were examined for heterogeneity. According to the

findings, since it was determined that the heterogeneity was high, it was considered appropriate

to conduct moderator analyses. The characteristics of the studies with higher and lower effect

sizes can be determined by moderator analyses (Card, 2012).

In the studies comparing group averages, Cohen d, Hedges g and Glass Δ formulas are

commonly used for calculating the effect sizes. These three indexes provide information about

the magnitude of an effect relative to the standard deviation (Ellis, 2010). Hedges g calculates

average values by correcting them (Borenstein et al., 2009). Therefore, Hedges’ g coefficient

was used in the calculation of the effect sizes in this study and the confidence level was

accepted as 95% in the analyses. Reporting the effect sizes in a standard form allows studies

to be compared (Ellis, 2010). In this study, the effects were classified as follows: -0.15 ≤ g

<0.15 insignificant effect size, 0.15≤ g <0.40 small effect size, 0.40 ≤ g <0.75, medium effect

size, 0.75 ≤ g <1.10, large effect size, 1.10 ≤ g <1.45 very large effect size 1.45 ≤ g perfect

effect size (Thalheimer & Cook, 2002, as cited in Dinçer, 2014). In the context of this study,

as a result of the analysis, the effect size is interpreted in favor of reading from the paper when

a negative value is found, and in favor of reading from the screen when a positive value is

found.

3. Results

In this section, first descriptive information about the studies, then publication bias,

heterogeneity, general / individual effect size and moderator analysis findings are given place.

Descriptive information about the studies is presented in Table 1. Here, information about

Öztop & Nayci

1212

publication year, grade level, text type used in reading and digital tool variables used in reading

are given place.

Table 1. Descriptive Information About The Studies

Variable Frequency

Publication Year

2012 2

2013 12

2014 5

2015 1

2017 4

2018 2

2019 3

Grade Level

Primary Ed. 4. grade 6

Primary Ed. 5. grade 15

Primary Ed. 6. grade 3

Primary Ed. 8. grade 2

Primary Ed. 5, 6, 7 and 8. grade 3

Text Type Used in Reading

Informative 11

Narrative 11

Mixed 2

Unspecified 5

Digital Tool Used in Reading

Smart Phone 1

PC 24

Tablet PC 2

Unspecified 2

Total 29

According to Table 1, it is seen that the theses that were made in 2013 are the most

among the studies. The 2015 is the year with the least study. According to the grade level, it is

seen that the studies conducted with primary education 5th grade students are at the highest

level and the studies conducted with primary education 8th grade students are at the lowest

level. According to the type of text used in reading, it is seen that reading studies conducted

with narrative (f=11) and informative texts (f=11) are equal. While the number of studies

conducted with mixed type texts is 2, the number of studies conducted with unspecified text

type is 5. Among the studies, it is seen that most studies are the ones in which computers

(desktop or laptop) (f=24) are used for reading. It draws the attention that there are 2 studies

using a tablet computer in reading, 1 study using smart phones, and 1 study that does not specify

the digital tool used in reading.

International Online Journal of Education and Teaching (IOJET) 2021, 8(2), 1206-1224.

1213

Publication Bias

Figure 1. Funnel Chart

In the Funnel chart in Figure 1, the effect size distribution of the studies having a

symmetric appearance indicates that there is no publication bias, and having a asymmetric

appearance indicates that there is publication bias (Borenstein et al., 2009; Sterne, Becker, &

Egger, 2005). When Figure 1 is examined, since the effect sizes of the study display an almost

symmetrical distribution, it can be said that there is no publication bias. Also Rosenthal safe N

method was used to determine whether there is publication bias or not. Calculation findings for

Rosenthal safe N method are given in Table 2.

Table 2. Calculation Findings for Rosenthal Safe N Method

Z value for the observed studies -9,05569

P value for the observed studies 0,00000

Alpha 0,5

Direction 2

Z value for Alpha 1,95996

Number of observed studies 29

Safe N number 591

Rosental safe N method shows how many more studies having zero effect size should

be included in the meta-analysis in order to convert p value to meaningless value (p=0.05)

(Borenstein et al., 2009). This proposal of Rosental is based on the idea that researchers do not

want to publish studies of insignificant value, therefore they hide these studies (Becker, 2005).

As the N number obtained as a result of this calculation is very low, there can be hesitations

about the publication bias (Borenstein et al., 2009). When Table 2 is examined, it is seen that

the Rosenthal safe N number was found as 591 as a result of the analysis. This high value

supports that there is no publication bias.

Another method which was used to decide whether there is publication bias or not is

Orwin's Safe N-number method. Orwin's approach is related to determining the number of

studies having zero effect size required to reduce the average effect size to a certain level

(Lipsey & Wilson, 2001). Orwin's approach is different from Rosental's approach because

while Rosenthal's calculations are about the meaninglessness of the study, Orwin's calculations

are about reducing to a certain effect size (Becker, 2005). In other words, it can be said that it

is aimed to determine how many more studies should be included in the meta-analysis in order

Öztop & Nayci

1214

to reach a meaningless p value in Rosental's formula and a determined general effect size in

Orwin's formula. In this study, the number of studies that have to be included in the meta-

analysis in order to reduce the general effect to a minor effect level has been calculated. As a

result of the calculation, the Orwin's safe N number was determined to be 32. This is another

result that supports the lack of publication bias. In addition, Begg and Mazumdar rank

correlation test and Egger's regression tests were also applied for publication bias. The data

related to them are given below.

Table 3. Begg and Mazumdar Rank Correlation Test Findings

Begg and Mazumdar rank correlation test examines the relationship between effect and

variances (Vevea, Coburn, & Sutton, 2019). According to Begg and Mazumdar (1994), this

test is the complementary of the funnel chart. The statistically insignificant test results show

that there is no publication bias (Vevea et al., 2019). As seen in Table 3, findings obtained from

the Begg and Mazumdar rank correlation test performed (tau=-0.08 p=0.52362 p> 0.05)

indicate that the study does not have publication bias. It is seen that p value is not statistically

significant also as a result of the Egger's regression test (P=0.228). Therefore, this test offers

evidence indicating that there is no publication bias as well. When the performed test results

are evaluated in general, it can be said that there is no publication bias in this study.

Heterogeneity and Effect Size Calculations

Table 4. Effect Size of Studies according to Fixed Effects Model

When Table 4 is examined, it is seen that Q value is 367,379. It can be said that this

value is greater than the value 41,337 which corresponds to the 28 degrees of freedom in the

chi-square table, and also its effect sizes distribution has a heterogeneous feature because of P

value (P=0,000) being statistically significant. In addition, I2 value being found as 92.378%

indicates that there is a high degree of heterogeneity. In this direction, random effects model

was used to calculate the effect sizes in the research. The effect size of the studies according to

the random effects model is presented in Table 5.

Table 5. The Effect Size of the Studies according to the Random Effects Model

Kendall’s S statistic (P-Q) -34, 0000

Tau -0,08374

Z value for Tau 0,63777

P value 0,52362

Model
Effect

Size

Standard

Error

95% Confidence

Interval
Heterogeneity

Lower

Limit

Upper

Limit
Df(q)

Value
I2 p

Fixed effects -,306 0,038 -0,380 -0,231 28 367,379 92,378 0,000

Model
Effect

Size

Standard

Error

95% Confidence

Interval

Lower

Limit

Upper

Limit

Random Effects -0,423 0,143 -0,703 -0,143 0,003

International Online Journal of Education and Teaching (IOJET) 2021, 8(2), 1206-1224.

1215

When Table 5 is examined, the average effect size value calculated according to the

random effects model was found to be significant and as -0.423. The standard error of the effect

size is 0.143 and its lower limit is -0.703 and its upper limit is -0.143 in the 95% confidence

interval. These values indicate a moderate effect in favor of reading from the paper. The forest

plot regarding the effect size of the studies is given in Figure 2.

Figure2. The Effect Sizes of the Studies and Forest Plot

The black square at the bottom of the figure in Figure 2 shows the overall effect size.

The black squares having lines to their left and right represent the effect sizes of individual

studies. The horizontal lines to the left of the black squares indicate the confidence interval. It

is observed that the study with the lowest confidence interval is the study of Batluralkız (2008a,

2008b) and the study with the highest confidence interval is the study of Dündar and Akçayır

(2012). And on the far right side of the figure, the weights of the studies are located. It is seen

that the study with the highest weighting is the study of Ercan and Ateş (2015) and the study

with the lowest weighting is the study of Dündar and Akçayır (2012).

It is seen that the effect size values of the individual studies ranged between -2,689 and

0,891. It is seen that 19 of the studies have a negative positive effect size value and 10 of them

have a positive effect size value. It can be said that 5 of these studies have an insignificant (4

in favor of reading from the paper, 1 in favor of reading from the screen), 11 of them have a

small (5 in favor of paper reading, 6 in favor of screen reading), 4 of them have a medium (2

in favor of paper reading, 2 in favor of screen reading), 3 of them have a large (2 in favor of

paper reading, 1 in favor of screen reading), 1 of them have a very large (in favor of paper

Öztop & Nayci

1216

reading) and 5 of them have a perfect (in favor of paper reading) effect. It can be said that there

are more studies with high effect in favor of reading from the paper.

Moderator Analyses

Table 6. Effect Size Differences according to the Publication Year

Publication

Year

Number

of Studies

Effect

Size

%95 Confidence

Interval
Between-

Groups

Homogeneity

Value (QB)

p
Lower

Limit

Upper

Limit

2012 2 -0,002 -1,122 1,119

8,220 0,222

2013 12 -0,500 -0,946 -0,054

2014 5 -0,038 -0,710 0,634

2015 1 -1,144 -2,616 0,329

2017 4 -1,209 -1,974 -0,444

2018 2 -0,055 -1,088 0,978

2019 3 -0,010 -0,893 0,872

The first variable examined according to the effect sizes in the research is publication

year. When Table 6 is analysed, the effect sizes were found as -0,002 for 2012, -0,500 for 2013,

-0,038 for 2014, -1,144 for 2015, -1,209 for 2017, -0,055 for 2018 and -0,010 for 2019. As a

result of the test conducted to determine whether there is a significant difference between the

effect sizes of the studies according to the publication year, it was determined that the effect

sizes did not differ significantly according to the publication year (QB=8.220, p> 0.05).

Another moderator variable whose effect was examined in the study is the grade level variable.

The effect size differences of the studies according to the grade level are given in Table 7.

Table 7. Effect Size Differences according to the Grade Level

Grade Level
Number

of Studies

Effect

Size

%95 Confidence

Interval
Between-Groups

Homogeneity

Value (QB)

p
Lower

Limit

Upper

Limit

Prim. Ed. 4. grade 6 -0,839 -1,495 -0,183

2,959 0,565

Prim. Ed. 5. grade 15 -0,404 -0,826 0,018

Prim. Ed. 6. grade 3 -0,415 -1,314 0,485

Prim. Ed. 8. grade 2 0,049 -1,092 1,190

Prim. Ed. 5., 6., 7.

and 8. grade

3 -0,010 -0,945 0,924

As seen in Table 7, average effect sizes according to grade level were calculated as g=-

8.394 for primary education 4th grade students, g=-0,404 for primary education 5th grade

students, g=-0,415 for primary education 6th grade students, g=0,049 for primary education

8th grade students and g=-0,010 for primary education 5th, 6th, 7th and 8th grade students. As

a result of the test conducted to determine whether there is a significant difference between the

effect sizes of the studies according to the grade level, it was determined that the effect sizes

did not differ significantly according to the grade level (QB=2.959, p> 0.05). Nevertheless,

when the effect sizes are compared, it is seen that lower grades have a stronger effect in favor

of reading from the paper compared to upper grades. As the grade level increases, the situation

shifts in favor of reading from the screen. In the study, also the text type variable used in

International Online Journal of Education and Teaching (IOJET) 2021, 8(2), 1206-1224.

1217

reading was examined as moderator variable. The findings with regards to the effect size

differences according to the text type used in reading are given in Table 8.

Table 8. Effect Size Differences according to the Text Type Used in Reading

Text Type Used

in Reading

Number

of Studies

Effect

Size

%95 Confidence

Interval
Between-

Groups

Homogeneity

Value (QB)

p
Lower

Limit

Upper

Limit

Informative test 11 -0,632 -1,064 -0,199

5,451 0,066 Narrative text 11 0,056 -0,371 0,483

Mixed 2 -0,673 -1,677 0,332

As seen in Table 8, effect sizes according to the text type used in reading are examined

in three groups. Average effect sizes were found to be g=-0632 in informative text, g=0.056 in

narrative text and (-0.667) in mixed studies. As a result of the test conducted to determine

whether there is a significant difference between the effect sizes of the studies according to the

text type used in reading, it was determined that the effect sizes did not differ significantly

according to the text type used in reading (QB=5.451, p> 0.05). Nevertheless, when the effect

sizes are compared, it can be seen that informative texts have a medium effect (g=-0.632) in

favor of reading from the paper, and narrative texts have an insignificant effect (g=0.056) in

favor of reading from the screen. In the study, the digital tool variable used in reading was also

examined as a moderator variable. Findings with regards to the effect size differences

according to the digital tool used in reading are given in Table 9.

Table 9. Effect Size Differences according to the Digital Tool Used in Reading

Digital Tool Used

in Reading

Number

of Studies

Effect

Size

%95 Confidence

Interval
Between-

Groups

Homogeneity

Value (QB)

p
Lower

Limit

Upper

Limit

Smart Phone 1 -0,047 -1,779 1,685

1,577 0,455 PC 24 -0,527 -0,881 -0,174

Tablet PC 2 0,255 -1,014 1,523

When Table 9 is examined, the effect sizes according to the digital tool used in reading

were examined in three different groups. Average effect sizes were found as g=-0.047 for

smartphone, g=-0.527 for PC and g=0.255 for tablet PC. As a result of the test conducted to

determine whether there is a significant difference between the effect sizes of the studies

according to the digital tool used in reading, it was determined that the effect sizes did not

differ significantly according to the digital tool used in reading (QB=1,577, p> 0.05). However,

when the effect sizes are compared, it is seen that the computer has a medium effect in favor

of reading from the paper (g=-527), the smartphone has an insignificant effect in favor of

reading from the paper (g =-0.047) and the tablet has a small effect in favor of reading from

the screen (g=0.255).

4. Discussion, Conclusion and Recommendations

In this study which was conducted to examine whether the level of understanding

differs according to screen reading and paper reading by meta-analysis method, a significant

and medium (g=-0.423, p=0.003) effect size was found in favor of reading from the paper.

Öztop & Nayci

1218

Therefore, this result shows that the reading comprehension level is higher in reading from the

paper than reading from the screen. A number of meta-analysis studies conducted abroad also

set forth that the comprehension level in reading from the paper is higher, supporting the results

of this study (Clinton, 2019; Delgado et al., 2018; Kong et al., 2018). Therefore, it can be said

that the printed materials have a more positive effect on reading achievement compared to the

screen. Transferring the texts to the screen led to the role of many elements related to the screen

such as brightness, color, etc. to be deemed important during the reading process. In screen

reading, eye movements change, it becomes difficult to apply some reading techniques, and

the processes of understanding and structuring in the mind become difficult (Güneş, 2010).

While reading from the paper contributes more to focusing (Baron, 2015), screen reading is

more prone to being distracted and focusing problems can be seen. Wang et al. (2008), on the

other hand, determined in the meta-analysis study they carried out on paper-pencil and PC form

that there was no significant difference between reading achievements according to the

environment differing from this study.

In this study which examines whether the comprehension level differs according to

reading from the screen and reading from the paper, also analyses were conducted for some

sub-factors that may affect the effect sizes. One of the factors discussed is the publication year

variable. As a result of the analyses performed, it was determined that the effect size did not

differ significantly according to the publication year. These findings show that the publication

year does not significantly change the comprehension level according to the reading

environment. Kong et al. (2018) similarly determined in the meta-analysis studies they

conducted that the effect sizes did not differ significantly according to the year variable. Even

if the texts read from the past to present remain the same, changes in the environment in which

the texts are presented and in the reading habits of individuals in digital environment can be

seen. In this study, no significant difference was observed in reading comprehension according

to the environment in terms of years.

A different variable whose effect is examined is the grade level. As a result of the

analyses, it was determined that the effect size did not differ significantly according to the

grade level. Delgado et al. (2018) also found that effect sizes did not differ significantly

according to the grade level variable in meta-analysis studies. This finding shows that the grade

level does not significantly change the differentation in the comprehension level according to

the reading environment. Nevertheless, when the effect sizes are compared, it is seen that lower

grades have a stronger effect in favor of reading from the paper compared to the upper grades.

As the grade level increases, it is seen that there is a shift in favor of reading from the screen.

In other words, students at higher grade level have a higher level of reading comprehension on

the screen. It can be said that lower level students have difficulty in reading from the screen

due to the need for various high level mental skills such as intense attention, remembering, fast

recognition of words, fast combination of information and questioning (Güneş, 2010). To

support this idea, the study conducted by Helder, Van Leijenhorst, and van den Broek (2016)

at the primary school level found that lower grade students had poorer comprehension skills

and had more difficulty in coherence, in direct proportion to their age.

Another factor whose effect is observed is the text type used in reading. As a result of

the analysis, it was determined that the effect size did not differ significantly according to the

text type used in reading. This finding shows that the text type used in reading does not

significantly change the differentiation in the comprehension level according to the reading

environment. Nevertheless, when the effect sizes are compared, it is noteworthy that

informative texts have a medium effect in favor of reading from the paper, and narrative texts

have an insignificant effect in favor of reading from the screen. In their meta-analysis study,

Clinton, (2019) and Delgado et al. (2018) similarly found that informative texts have an effect

International Online Journal of Education and Teaching (IOJET) 2021, 8(2), 1206-1224.

1219

in favor of reading from the paper, whereas in narrative texts there is an effect that does not

matter for the two environments. As can be seen, reading the narrative texts from the screen or

paper does not affect the comprehension much. Since individuals encounter more narrative

texts from the beginning of school and are familiar with these texts, they may struggle when

they encounter informative texts. In addition, informative texts can be encountered as a text

type that has difficulties, as it reveals concepts and relationships that readers do not know (Coté,

Goldman, & Saul, 1998) and deals with more abstract concepts (Şahin, 2013). It can be said it

is expected that a better reading comprehension is realised in reading informative texts from

the paper due to its nature.

Finally, the effect sizes were examined according to the digital tool used in reading. As

a result of the analysis, it was determined that the effect sizes did not differ significantly

according to the digital tool used in reading. In the meta-analysis study of Kong et al. (2018),

similarly, the effect sizes did not differ significantly according to the digital tool. These

findings show that the digital tool used in reading does not significantly change the

differentiation in the comprehension level according to the reading environment. Nevertheless,

when the effect sizes are compared, it is seen that PC has a medium effect in favor of reading

from the paper, the smartphone has an insignificant effect in favor of reading from the paper

and the tablet has a small effect in favor of reading from the screen. It draws attention that

reading from tablets which is the digital device having the screen closest to the size of a book

has a higher comprehension level than reading from computers having large screens. Delgado

et al. (2018), in their meta-analysis study, found that hand devices are more effective in

comprehension of reading from the screen compared to the computers, supporting this finding.

Providing the flow of texts read on computers with the mouse can affect reading skills. Since

tablets offer readers an experience as if they were reading from a printed book due to having a

touch screen and an easy-to-use interface, (Chen et al., 2014), it can be said that there will be

less disconnection and reading skills will approach a level close to traditional reading. The text

being directly opposite the eye on the desktop or laptop computers and thus the eye not getting

the same angle which it gets while reading printed materials can be considered as another

possible reason for this result. For an effective and efficient reading, the monitor should be

placed 10-15 cm below eye level, in the manner to read by tilting the head slightly down

(Güneş, 2009).

This meta-analysis study, conducted with a limited number of studies, shows that the

use of printed materials in reading contributes more positively to comprehension success. As

screens enter our lives more and more, the importance of understanding what is read from the

screen is increasing. In order to increase the level of reading comprehension on the screen,

importance should be given to the development of screen reading skills and screen interface.

This study is thought to create an awareness for researchers, educators, digital content

developers and the society and contribute to the literature.

In line with the results obtained from the study, the following suggestions can be made:

 Studies in which only Turkish texts were used in reading were included in this meta-
analysis study. The meta-analysis of studies comparing the comprehension level in

foreign language education according to the reading environment can be made.

 Studies to improve screen reading skills can be carried out.

 Within the context of the screen interface, studies can be conducted on factors such
as screen size, text size, text font style, brightness that may affect reading from the

screen.

 Comparative studies can be conducted on groups that have been using digital devices
for a long time and who have been using them for a short time.

Öztop & Nayci

1220

 Comprehension levels of reading from the screen and from the paper can be
compared according to the type of comprehension tests (multiple choice, open-ended

etc.).

References

Aloqaili A. S. (2011). The relationship between reading comprehension and critical thinking:

A theoritical study. Journal of King Saud University-Languange and Translation, 24, 35-

41. doi: doi:10.1016/j.jksult.2011.01.001

*Aydemir, Z., Öztürk, E., & Horzum, M. B. (2013). Ekrandan okumanın 5. sınıf öğrencilerinin

bilgilendirici ve öyküleyici metin türünde okuduğunu anlama düzeylerine etkisi [The

Effect of Reading from Screen on The 5th Grade Elementary Students’ Level of Reading

Comprehension on Informative and Narrative Type of Texts]. Kuram ve Uygulamada

Eğitim Bilimleri, 13(4), 2263-2276.

Baron, N. S. (2015). Words onscreen: The fate of reading in a digital world. Oxford: Oxford

University Press.

*Başaran, M. (2014). 4. sınıf seviyesinde ekrandan ve kâğıttan okumanın okuduğunu anlama,

okuma hızı ve metne karşı geliştirilen tutum üzerindeki etkisi [Effect of reading on screen

and paper on elementary school 4th grade students' reading comprehension reading speed

and their attitudes]. Uşak Üniversitesi Sosyal Bilimler Dergisi, 7(2), 248-268.

*Baştuğ M., & Keskin, H. K. (2012). Okuma becerilerinin okuma ortamı açısından

karşılaştırılması: ekran mı kâğıt mı? [Comparison of reading skills in terms of the reading

environment: screen vs. paper] Atatürk Üniversitesi Sosyal Bilimler Enstitüsü Dergisi,

16 (3), 73-83.

*Batluralkız, Ç. (2018). Basılı ve dijital ortamlarda ortaokul öğrencilerinin okuduğunu

anlama becerilerinin karşılaştırılması [Comparing the reading comprehension skills of

primary school students ın print vs digital setting]. Unpublished master's thesis, Kütahya

Dumlupınar Üniversitesi, Kütahya.

Becker, B. J. (2005). Failsafe n or file-drawer number. In H. Rothstein, A. J. Sutton,

and M. Borenstein (Ed.), Publication bias in meta-analysis: Prevention, assessment and

adjustment (s. 111-125). Chichester: John Wiley and Sons Ltd

Begg, C.B., & Mazumdar, M. (1994) Operating Characteristics of a Rank Correlation Test for

Publication Bias. Biometrics, 50(4), 1088-1101.

Borenstein, M. (2005). Software for publication bias. In H. Rothstein, A. J. Sutton,

and M. Borenstein (Ed.), Publication bias in meta-analysis: Prevention, assessment and

adjustment (s. 193–220). Chichester: John Wiley and Sons Ltd

Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2009). Introduction to

meta-analysis. Chichester: John Wiley & Sons, Ltd.

Calhoon, M.B. (2005). Effects of a peer-mediated phonological skill and reading

comprehension program on reading skill acquisition for middle school students with

reading disabilities. Journal of Learning Disabilities, 38(5), 424–433.

Card, N. A. (2012). Applied meta-analysis for social science research. New York: The

Guilford Press.

Chen, G., Cheng, W., Chang, T. W., Zheng, X., & Huang, R. (2014). A comparison of reading

comprehension across paper, computer screens, and tablets: Does tablet familiarity

International Online Journal of Education and Teaching (IOJET) 2021, 8(2), 1206-1224.

1221

matter?. Journal of Computers in Education, 1(2-3), 213-225. doi: 10.1007/s40692-014-

0012-z

Clinton, V. (2019). Reading from paper compared to screens: A systematic review and meta-

analysis. Journal of Research in Reading, 42(2), 288–325. doi:10.1111/1467-

9817.12269

Cooper, H. (2017). Research synthesis and meta-analysis: A step-by-step approach. Thousand

Oaks: Sage Publications.

Coté, N., Goldman, Susan R., & Saul, E. U. (1998). Students making sense of Informational

text relations between processing and representation. Discourse Processes, 25 (1), 1-53.

*Çınar, M., Doğan, D., & Seferoğlu, S. S. (2019). The effects of reading on pixel vs. paper: a

comparative study. Behaviour & Information Technology,

doi:10.1080/0144929X.2019.1685594

Delgado, P., & Salmerón, L. (2021). The inattentive on-screen reading: Reading medium

affects attention and reading comprehension under time pressure. Learning and

Instruction, 71, 101396. doi:10.1016/j.learninstruc.2020.101396

Delgado, P., Vargas, C., Ackerman, R., & Salmerón, L. (2018). Don't throw away your printed

books: A meta-analysis on the effects of reading media on reading comprehension.

Educational Research Review, 25, 23-38. doi: 10.1016/j.edurev.2018.09.003

Dinçer, S. (2014). Eğitim bilimlerinde uygulamalı meta-analiz. Ankara: Pegem Akademi.

*Duran, E., & Alevli, O. (2014). Ekrandan okumanın sekizinci sınıf öğrencilerinde anlamaya

etkisi [The effect of reading on screen on comprehension of eight grade

students]. Okuma Yazma Eğitimi Araştırmaları, 2(1), 1-11.

*Dündar, H., & Akçayır, M. (2012). Tablet vs. paper: the effect on learners' reading

performance. International Electronic Journal of Elementary Education, 4(3), 441-450.

Ellis, P. D., (2010). The essential guide to effect sizes: Statistical power, meta-analysis, and

the interpretation of research results. Cambridge: Cambridge University Press.

*Ercan, A. N., & Ateş, M. (2015). Ekrandan okuma ile kâğıttan okumanın anlama düzeyi

açısından karşılaştırılması [The comparison of reading from paper and screen in terms of

comprehension levels]. Turkish Studies, 10 (7), 395-406.

doi:10.7827/TurkishStudies.8032

*Ertuğrul, B. (2013). İlköğretim beşinci sınıflarda ekrandan okumanın anlamaya etkisi [The

effect of screen reading on understanding in primary fifth class]. Unpublished master's

thesis, Uşak Üniversitesi, Uşak.

Garland, K. J., & Noyes, J. M. (2004). CRT monitors: Do they interfere with learning?.

Behaviour & Information Technology, 23(1), 43-52.

doi:10.1080/01449290310001638504

Glass, G. V. (1976). Primary, secondary, and meta-analysis of research. Educational

Researcher, 5(10), 3-8. doi:10.2307/1174772

Güneş, F. (2009). Ekran okumada verimlilik. Kalkınmada Anahtar Verimlilik Gazetesi, (Milli

Prodüktivite Merkezi Aylık Yayın Organı), August, 248,26-28.

Güneş, F. (2010). Öğrencilerde ekran okuma ve ekranik düşünme [Thinking based on screen

and screen reading of students]. Mustafa Kemal Üniversitesi Sosyal Bilimler Enstitüsü

Dergisi, 7(14), 1-20.

http://dergipark.org.tr/tr/pub/oyea

Öztop & Nayci

1222

Güneş, F. (2016). Kâğıttan ekrana okuma alanındaki gelişmeler [From paper to screen

developments in the field of reading]. Bartın Üniversitesi Eğitim Fakültesi Dergisi, 5(1),

1-18. doi:10.14686/buefad.v5i1.5000155474

Halamish, V., & Elbaz, E. (2019). Children‘s reading comprehension and meta comprehension

on screen versus on paper. Computers and Education, 103737. doi:

10.1016/j.compedu.2019.103737

Helder, A., Van Leijenhorst, L., & van den Broek, P. (2016). Coherence monitoring by good

and poor comprehenders in elementary school: Comparing offline and online measures.

Learning and Individual Differences, 48, 17–23. doi:10.1016/j.lindif.2016.02.008

Hillman, D. C., Willis, D. J., & Gunawardena, C. N. (1994). Learner-interface interaction in

distance education: An extension of contemporary models and strategies for

practitioners. The American Journal of Distance Education, 8(2), 30-42.

Hosseini, E., Khodaei, F. B., Sarfallah, S., & Dolatabadi, H. R. (2012). Exploring the

relationship between critical thinking, reading comprehension and reading strategies of

English University students. World Applied Sciences Journal, 17, 1356-1364.

Kaya, Z. (2002). Uzaktan eğitim. Ankara: Pegem Yayıncılık.

Kazazoğlu, S. (2020). Is printed-text the best choice? A mixed-method case study on reading

comprehension. Journal of Language and Linguistic Studies, 16(1), 458.

Kingston, N. M. (2008). Comparability of computer-and paper-administered multiple-choice

tests for K–12 populations: A synthesis. Applied Measurement in Education, 22(1), 22-

37. doi:10.1080/08957340802558326

Kol, S., & Schcolnik, M. (2000). Enhancing screen reading strategies. CALICO Journal, 18(1),

67-80.

Kong, Y., Seo, Y.S., & Zhai, L. (2018). Comparison of reading performance on screen and on

paper: A meta‐analysis. Computers & Education, 123, 138–149. doi:

10.1016/j.compedu.2018.05.005

Kurata, K., Ishita, E., Miyata, Y., & Minami, Y. (2017). Print or digital? Reading behavior and

preferences in Japan. Journal of the Association for Information Science and

Technology, 68(4), 884-894. doi:10.1002/asi.23712

*Kuru, O., Kaşkaya, A., & Calp, Ş. (2017). İlkokul 4. sınıf öğrencilerinin kağıttan ve ekrandan

okuduğunu anlama becerilerinin sınanması; öğretmen ve öğrenci görüşleri [Comparison

of the reading skills of 4th grade primary school students from the paper and from the

screen, the viewpoints of teachers and students]. Sosyal Bilimler Dergisi, 7(13), 70-84.

Lebedeva, M. Y., Veselovskaya, T. S., & Kupreshchenko, O. F. (2020). Features of perception

and understanding of digital texts: interdisciplinary view. Perspektivy naukii

obrazovania – Perspectives of Science and Education, 46 (4), 74-98. doi:

10.32744/pse.2020.4.5

Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Thousand Oaks: Sage

Publications.

Littell, J. H., Corcoran, J., & Pillai, V. (2008). Systematic reviews and meta-analysis. New

York: Oxford University Press

Mangen, A., Walgermo, B.R., & Brønnick, K. (2013). Reading linear texts on paper versus

computer screen: Effects on reading comprehension. International Journal of

Educational Research, 58, 61-68. doi: 10.1016/j.ijer.2012.12.002

International Online Journal of Education and Teaching (IOJET) 2021, 8(2), 1206-1224.

1223

Martin, L. A., & Platt, M. W., (2001). Printing and screen reading in the medical school

curriculum: Guttenberg vs. the cathode ray tube. Behaviour & Information Technology,

20, 143 – 148. doi: 10.1080/01449290110048043

Mellard, D. F., Fall, E., & Woods, K. L. (2010). A path analysis of reading comprehension for

adults with low literacy. Journal of Learning Disabilities, 43, 154–165.

Mizrachi, D. (2015). Undergraduates' academic reading format preferences and behaviors. The

Journal of Academic Librarianship, 41(3), 301-311. doi:10.1016/j.acalib.2015.03.009

Moore, M.G. (1989) Editorial: Three types of interaction. American Journal of Distance

Education, 3(2), 1-7. doi: 10.1080/08923648909526659

*Özen, M., & Ertem, İ. S. (2014). Metinleri ekrandan okumanın anlam kurma üzerine etkisi [

The effect of screen reading of texts on meaning making]. International Journal of Social

Science, 24, 319-350. doi:10.9761/JASSS2161

Partnership for 21st Century Skills (P21). (2013). Framework for 21st century learning.

Retrieved April 5, 2020, from https://www.battelleforkids.org/networks/p21

Rajanen, D., Salminen, M., & Ravaja, N. (2016, Jan.). Reading a newspaper on print versus

screen: a motivational perspective. In 2016 49th Hawaii International Conference on

System Sciences (HICSS) (pp. 630-637). IEEE.

Redmond, P. (2011). From face-to-face teaching to online teaching: Pedagogical transitions.

In G. Williams, P. Statham, N. Brown, & B. Cleland (Eds.), Changing demands,

changing directions. Proceedings of the ascilite conference (pp. 1050–1060). Hobart

Tasmania, Australia.

Rothstein, H. R., Sutton, A., & Borenstian, M. (2005). Publication bias in meta-analysis. In

H. Rothstein, A. J. Sutton, & M. Borenstein (Ed.), Publication bias in meta-analysis:

Prevention, assessment and adjustment (s. 1-7). Chichester: John Wiley and Sons Ltd

Schilhab, T., Balling, G., & Kuzmicova, A. (2018). Decreasing materiality from print to screen

reading. First Monday, 23 (10) 1-12.

Singer, L. M., & Alexander, P. A. (2017). Reading on paper and digitally: What the past

decades of empirical research reveal. Review of educational research, 87(6), 1007-1041.

Spencer, C. (2006). Research on learners’ preferences for reading from a printed text or from

a computer screen. Journal of Distance Education, 21(1), 33-50.

Sterne, J.A.C., Becker, B. J., & Egger, M. (2005). The funnel plot. In H. Rothstein, A. J. Sutton,

and M. Borenstein (Ed.), Publication bias in meta-analysis: Prevention, assessment and

adjustment (s. 75-98). Chichester: John Wiley and Sons Ltd.

Støle, H., Mangen, A., & Schwippert, K. (2020). Assessing children’s reading comprehension

on paper and screen: A mode-effect study. Computers & Education, 151,1-13. doi:

https://doi.org/10.1016/j.compedu.2020.103861

Şahin, A. (2013). The effect of text types on reading comprehension. Mevlana International

Journal of Education (MIJE), 3(2), 57-67.

Tamsi, R. K., Zuhri, F., & Kurniasih, E. (2013). The implementation of skimming and scanning

strategies in teaching reading narrative text for the tenth grade students of SMAN 21

SURABAYA. Retain, 1(2), 1–7.

Taşcı, G., & Çelebi, M. (2020). Eğitimde yeni bir paradigma: “yükseköğretimde yapay zekâ”

[“A new paradigm in education: "artificial intelligence in higher education”]. OPUS

Öztop & Nayci

1224

Uluslararası Toplum Araştırmaları Dergisi, 16 (29), 2346-2370.

doi:10.26466/opus.747634

Uther, M., Ross, K., Randell, J., & Pye, R. (2019, July). Digital vs. hard copy? A preliminary

study of reading style in children using touch screen and paper books. In International

Conference on Human-Computer Interaction (pp. 495-502). Springer, Cham.

Ünal, E., & Köksal, K. (2007). Okuduğunu anlama ve sorular. Üniversite ve Toplum: Bilim,

Eğitim ve Düşünce Dergisi, 7(4), 1-12.

Vevea, J. L., Coburn, K., & Sutton, A. (2019). Publication bias. In H. Cooper, L. V. Hedges,

J. C. Valentine (Ed.), The Handbook of Research Synthesis and Meta-Analysis (s. 383-

429). New York: Russell Sage Foundation.

Wang, S., Jiao, H., Young, M., Brooks, T., & Olson, J. (2008). Comparability of computer-

based and paper-and-pencil testing in K-12 reading assessments: A meta-analysis of

testing mode effects. Educational and Psychological Measurement, 68(1), 5–24.

doi:10.1177/0013164407305592

Wang, S., Yu, H., Hu, X., & Li, J. (2020). Participant or spectator? Comprehending the

willingness of faculty to use intelligent tutoring systems in the artificial intelligence era.

British Journal of Educational Technology, 51(5), 1657-1673. doi:10.1111/bjet.12998

*Yaman, H., & Dağtaş, A. (2013). Ekrandan okumanın okuduğunu anlamaya etkisi [İmpact of

reading from screen on reading comprehension]. Okuma Yazma Eğitimi Araştırmaları, 1

(2), 64-79.

Yılmaz, M. (2011). İlköğretim 4. sınıf öğrencilerinin okuduğunu anlama seviyeleri ile türkçe,

matematik, sosyal bilgiler ve fen ve teknoloji derslerindeki başarıları arasındaki ilişkinin

belirlenmesi [Determining among to correlation of 4th grade students’ success in turkish,

maths, social seciences and science and technology lessons of reading comprehension

levels]. Dumlupınar Üniversitesi Sosyal Bilimler Dergisi, 29(29), 9-13.

Zou, X.-L., & Ou, L. (2020). EFL reading test on mobile versus on paper: a study from

metacognitive strategy use to test-media impacts. Educational Assessment, Evaluation

and Accountability. 1-20. doi:10.1007/s11092-020-09320-0

*Studies included in meta-analysis.