Media Technologies and Artificial Intelligence in

Education: Essential for the Future or Unwelcome

Innovation?
A look at general attitudes towards media technologies and artificial intelligence in
education, and the framework for innovation adaptation

Student Name: Timo van Leeuwen

Student Number: 447060

Supervisor: Peter Nikken

Master Media and Creative Industries

Erasmus School of History, Culture and Communication
Erasmus University Rotterdam

Master Thesis
June 28 2021

1
 Media Technologies and Artificial Intelligence in
Education: Essential for the Future or Unwelcome
Innovation?
A look at general attitudes towards media technologies and artificial intelligence in
education, and the framework for innovation adaptation

ABSTRACT
Since the role of education is of indescribable importance in societies, it is essential that education
stays up to date with its time. As of recent decades, media technologies (MT) have been increasingly
important in society and implemented in education, and now artificial intelligence (AI) is rapidly
developing and also seems poised to be majorly implemented in education. This study explores the
positives and negatives of both MT and AI in education, tries to assess general attitudes towards the
implementation of MT and AI in education, and tries to unravel key predictors that influence these
attitudes. Based on adaptation innovation frameworks such as the diffusion of innovation theory, the
technological acceptance model and the unified theory of acceptance of use of technology, certain
predictors are implemented in this research. The associations that perceived usefulness, perceived
ease of use, experience, mobile self-efficacy, age and Schwartz’ human values have with attitude
towards both MT and AI in education are researched. A self-administered online questionnaire was
carried out to test these predictors (N=165), with predominantly positive results. The overall attitude
towards MT and AI in education seemed neutral, however younger age groups were predominantly
positive towards both MT and AI, where older age groups were predominantly negative towards
both MT and AI. All predictors that were hypothesised to have associations with attitude towards MT
and AI, were at least partially significant. As single predictors, perceived usefulness, perceived ease of
use, experience, mobile self-efficacy, age and Schwartz’ human values all had strong significant
associations with both attitude towards MT and AI in education. In a wider model with all predictors
implemented, perceived usefulness and age were seen as the strongest predictors for attitude, with
perceived ease of use and conservatism also both having significant associations with both attitude
towards MT as AI in education. Further research on both the attitudes towards MT and AI in
education, as on the advancement of adaptation innovation frameworks is desirable, as it seems
unavoidable that MT and AI are (going to become) essential in societies. Because of this, refined and
future-proof frameworks to help with innovation adaptation could help with the implementation of
MT and AI, and future innovations.

KEYWORDS: Media technologies, artificial intelligence, education, innovation adaptation frameworks

2
Table of Contents
1. Introduction 4
2. Theoretical Framework 7
2.1. Media technologies in education 7
2.2. Artificial intelligence in education 10
2.3. Innovation and media technologies adaption frameworks 13
2.4. Direct technology-related predictors for attitude 15
2.5. Socio-demographic and indirect predictors 17
3. Methods 19
3.1. Choice of method 19
3.2. Research unit 20
3.3. Sampling 21
3.4. Operationalization 22
3.4.1. Schwartz’ human values 22
3.4.2. Experience 23
3.4.3. Mobile self-efficacy 23
3.4.4. Perceived usefulness and perceived ease of use 24
3.4.5 Attitude towards media technologies and AI in education 24
3.5. Survey design 25
3.6.Tool, reliability and validity 26
4. Results 26
4.1. Descriptive results 26
4.4.1. Sample descriptives 26
4.4.2. Scale descriptives 27
4.2. Effect of perceived usefulness and perceived ease of use 28
4.3. Effect of mobile self-efficacy and experience 30
4.4. Effect of age on attitude 30
4.5. Effect of Schwartz’ human values 32
4.6 Additional findings: Strongest predictors 33
5. Conclusion 36
5.1. Summary of findings 36
5.2. Theoretical implications 38
5.3. Limitations and future research 40
References 42
Appendix A: Survey 58

3
 1) Introduction

Education is, as Nelson Mandela famously put it, ‘the most powerful weapon which you can use to
change the world. (Nelson, 1990)’ A quote like this does not counter much resistance, as an
education system has, in every society, an incredibly important societal function, and is of
indescribable importance. The benefits of education are easily measurable, as there are countless of
direct links between education and for example improved risk perception (Torani, Majd, Maroufi,
Dowlati & Sheikhi, 2019), improved health (Kemp & Montez, 2020) and empowerment (Singh, 2016).
However, to write off education as a handful of beneficial consequences would be doing it short, as
education entails much more than this on both an individual and societal level. Education is the
foundation for the rest of your life, prepares you to individually and responsibly be a part of society,
and functions as a medium for personal, social, political, economic and cultural development
(Bhardwaj, 2016). Because of this essential role education has in our society, it is incredibly important
that an education system progresses along with its time and keeps preparing the youth for the world
awaiting them. However, in the past (few) decade(s), this has not always necessarily been the case.
Over recent years, education in countries that historically have performed well in the education
sector, has been slipping in overall quality, attainment and participation (Inspectie van het onderwijs,
2018; Busteed, 2020; Weale, 2020). Technology, and especially media technologies, have been so
rapidly developing over the last decades, that education systems are accused of not being able to
keeping up with their developments (Reams, 2017; Tsuboya-Newell, 2019). This is concerning
considering teachers, even back in 2013, almost unanimously claimed that media technologies have
massively impacted education (Purcell, Buchanan & Friedrich, 2013). Education systems, which
should prepare new generations for the world ahead of them, are thus now in danger of preparing
new generations for a time which has already past, and it is essential that they keep up with their
time to successfully keep fulfilling its societal role.

Two of the most important technological developments which have brought a

plethora of possible positives and negatives to education systems are media technologies and
artificial intelligence. Media technologies (MT) have become so ubiquitous in daily life that they are
practically unavoidable, especially amongst younger age groups (Anderson & Jiang, 2018). MT, which
are defined as ‘any hardware, software or tool that is used to compose, create, produce, deliver and
manage media’, refer in the context of this study only to hardware, so for example laptops, tablets,
mobile phones and other comparable hardware. Over the last few decades, these media
technologies have been increasingly used in education, through for example laptops and tablets, for
which usage has soared (Truong, 2020). Artificial intelligence (AI), a relatively newer concept, is now

4
also taking the world by storm, making its way in various essential sectors such as healthcare (Panch,
Mattie & Celi, 2019), entertainment (Sweichowski & Slezak, 2018), and more. Artificial intelligence,
which is defined as the ability of a computer to do tasks that are usually done by humans, in this
context strictly refers to software, that is run on the hardware that are MT. Famous everyday
examples of AI software, or software that utilizes AI, are is for example Google, Uber and Facebook
(Faggella, 2020). However, now there are innovative ways being found to also integrate AI into
various education systems. This increasing implementation of MT and AI into education provides
both opportunities and problems for education systems. Further MT and AI integration into
education systems could personalize learning more, and learning to work with these media
technologies and AI could prepare the youth for a future where these technologies seem
unavoidable. Additionally, that media technologies in education and their usage are on the up, is
apparent. The EdTech market, which is the market for educational technologies, for example media
technologies, has a market size value of 76.4 billion as of 2019, and is expected to grow by 18
percent annually for a market size of 285.2 billion in 227, the exponential growth partially explained
by the current covid-19 pandemic (Paykamina, 2021). While the market of AI in education is currently
worth only a fraction of this, at 1 billon, it is expected to grow much more exponentially at 45% a
year (Dukaninovska, 2020). Both MT and AI are poised to have a massive impact in many sectors of
society, but possibly education is one of the sectors where it will have most influence.

However, despite the positives both media technologies and AI can bring to education, and
despite how necessary it might be for education to fulfill its societal function, not everyone is very
happy about these developments. Numerous calls to action have been made to ban certain media
technologies, for example laptops, tablets and phones, from classrooms (Lieberman, 2017; Kurz,
2019; Truong, 2020; Criddle, 2021). Additionally, while artificial intelligence is not yet integrated to
such a degree that media technologies are, people are scared about the possible integration of AI in
education, and even AI in general (Ghafourifar, 2017; Pega, 2020). But it goes further than that, with
a sizeable portion of a recent survey even claiming they thought that AI would bring about the end of
the human race (Bucholz, 2019). What is interesting that most of these critiques and fears come from
older generations, parents teachers, generally the groups who use MT and AI less, but also have the
less knowledge about both MT and AI and are less competent in using them, compared to younger
generations and students. This digital skills and knowledge gap, which research shows is only getting
wider (Udeze & Oko, 2013; Milano, 2019), could be harmful to the integration of MT and AI in various
sector, but also towards other future innovations. We have seen this in the past, for example with
the integration of computers into our daily life. While computers are now pretty widely seen as
incredibly important in our lives and in society (Villalta, 2019), its integration was met with resistance

5
from predominantly older age groups, who still have a more negative attitude towards them (Broady,
Chan & Caputi, 2010; Lee, Czaja, Mozley, Sharit, Boot, Charness & Rogers, 2019). It is evident that not
everyone is on board with the revolution that MT and AI possibly are bringing to education, and it’d
be both interesting and relevant to unravel the reasons and motivators for this.

It is both scientifically as societally highly relevant to assess the attitudes people have
towards media technologies and AI in education, and what the reasons behind these attitudes are.
Scientifically, there is a lack of research thus far in attitudes towards AI and what the reasons behind
these attitudes are, considering it is still a relatively new phenomenon, and research on it is still
somewhat scarce. Research on the acceptation of MT is a lot more common, yet research on the
usage of MT in general education, and what people think about this, is surprisingly scarce.
Furthermore, it is scientifically relevant to research the several interrelating reasons for attitudes
towards innovations/technologies. This is the case because while current frameworks on innovation
adaptation do exist, they are arguably, and especially in the light of the covid-19 pandemic, outdated
and/or obsolete (van den Heuvel, 2020; Al-Emran & Granic, 2021). The societal relevance is also quite
sizeable, as there are three components as to why its societally important to research this
phenomenon. Firstly, as education has an immensely important societal function, it is important that,
as mentioned before, it stays up to date and prepares the youth for the future ahead. Considering
that AI in the US alone already has cost 60 million jobs (Kelly, 2020), is projected to cost 85 million
jobs by 2025, but also projected to create 97 million new jobs (Shalamanov, 2021), it is highly
important that the new generation are able to deal with this technology in able to participate in
society, what essentially the role is of education. Secondly, it is societally relevant to assess the
attitudes towards MT and AI considering that MT are now a ubiquitous presence in life, and AI is
poised to have a massive impact in societies too. Therefore, assessing attitudes towards these
technologies to see if specific action, for example regarding acceptation or education, needs to be
taken regarding these technologies is highly relevant. Lastly, it is societally relevant to extend the
existing frameworks on innovation adaptation, or establish a new framework for adaptation
innovation which is future-proof. Considering humanity, and their technology, is developing rapidly, a
framework to easily assess the key factors in innovation adaptation would help in adapting future
innovations and assessing where possible problems lie for these innovations. For example, if gender
or age proves to be highly significant in adapting innovation, a country with a high populations of
certain age groups would have to deal different with implementing these innovations than others,
considering the resistance it might bring. All things considered, the scientific and societal relevance
for assessing attitudes towards MT and AI in education is apparent, which brings to the following RQ
with the subsequent sub questions.

6
RQ 1: To what extent do people look positively towards the implementation of media technologies
and artificial intelligence into education ?

SQ 1: What direct technology-related factors are associated with attitudes towards media
technologies and artificial intelligence in education?

SQ 2: What socio-demographic factors are associated with attitudes towards media technologies and
artificial intelligence in education?

SQ 3: What factors are the most important predictors for attitudes towards media technologies and
artificial intelligence in education?

2) Theoretical Framework

2.1) Media Technologies in Education

Media technologies (MT) have, over the course of history, become increasingly significant in
education, to the point where they are now essential. MT are defined as ‘any hardware, software or
tool that is used to compose, create, produce, deliver and manage media’ (Spacey, 2019), and
include ‘traditional’ technologies such as VHS, cassette players, televisions and projectors. However,
for this research, the focus is solely on electronic and digital media technologies, which is hardware
that depends on electricity and uses screens, so this excludes most of the aforementioned traditional
MT. While electronic MT were absent in classrooms until the beginning of the 20th century (Domine,
2009), these types of traditional technologies were steadily introduced into education during the 20th
century, and have since been replaced by newer and, arguably, better media technologies. These
technologies, dubbed new media technologies, are defined as technologies that use digital computer
technology for distribution and exhibition (Manovich, 2002). The distinction between new media and
new media technologies must be made, as there are papers and research that use the terms
interchangeably. New media, also often referred to as web 2.0, refers to the development of the
internet into a fully interactive platform of applications and content. New media technologies then,
are the technologies used to access these applications and content. To illustrate, an online blog is a
form of new media, as it is a user-generated form of content accessible on the internet. A media
technology then, is for example the laptop or phone you use to access that content, so the term
media technologies strictly refers to hardware. Over the last 2 decades, since the uprising of new
media and new media technologies, we have seen these technologies implemented rapidly into
education. The usage of MT such as tablet, laptops and smartphones has skyrocketed across all levels

7
of education (Cavanagh, 2015; Hassler, Major & Hennesy, 2016). Furthermore, during 2020 and
2021, education has essentially become solely dependent on MT such as videocall technology, for
example the application Zoom. This is the case because the Covid-19 pandemic, in many countries,
prevented students from physically attending classes. MT have become a core element which are
highly embedded in all levels of education (Westera, 2015), and the expectations are that this trend
of growing importance will continue (Walsh, 2020; Hughes, 2021).

The reason why MT have grown to be so important in education, is because there is a general
consensus that they aid with education. This is not a new insight, as research from back in the 1980s,
1990s, and even before that , already established that MT have a positive effect on education.
However, an effort must be made in clearly distinguishing two types of learning that MT enable
(Reeves, 1998). The first type of learning is learning from media, through for example instructional
media, integrated learning systems or other explanatory material (Seels, Berry, Fillerton & Horn,
1996; Reeves, 1998; Holden & Westfall, 2007) . You can think of for example how a teacher can
spend an hour trying to explain what a bee or any other animal looks like, but a video of a minute will
do a more effective job at doing the exact same thing. While older research, through predominantly
Richard Clark (1983; 1994; 2001), argued that there were no learning benefits to be gained from any
specific medium, and even that MT would never influence learning, countless research and the
current situation where MT are highly embedded in education has proven those assumptions wrong.
The second type of learning, originated by Robert Kozma (1991), is learning with media. This is
described as a complementary learning process where MT provide for example ‘cognitive tools and
constructivist learning environments’ (Reeves, 1998, p.4). Kozma argued that certain strengths of a
media technology, combined with methods or applications that take advantage of these strengths,
‘interact with and influence the ways learners represent and process information’ (Kozma, 1991,
p.179), and that because of this, learning is enhanced. Further research has since indicated that
different MT can indeed be implemented in various way to facilitate and enhance the learning
experience (Reeves, 1998; Singhal, Bagga, Goyal & Saxena, 2012; Ibáñez, Di Serio, Villarán & Kloos,
2014). What this means, is that for example the usage of tablets in a mathematics class can directly
lead to increased engagement, increased accuracy and a deeper understanding of mathematical
concepts (Murphy, 2016; Schacter & Jo, 2017), because the strengths of tablets help address
problems of disinterest, inaccuracy and misunderstanding, which are otherwise commonly seen in
mathematics classes. Countless further research with a wide variety in MT and level of education
exists to back up the general consensus regarding MT and education, which is that MT can enhance
education by allowing students to learn both from as with MT.

8
 However, the possible ineffectiveness or harmfulness of MT in education are also widely
researched and discussed. Some have even claimed that MT have ‘extensively plagued education’
(Khoshnnevisan, 2019, p.85). One of the biggest complaints about MT in education is that they are a
distracting factor, a claim backed up by an extensive amount of research showing for example
laptops, tablets and mobile phones to be a distraction in classes (Jackson, 2012; Goundar, 2014;
Taneja, Fiore & Fischer, 2015; Aaron & Lipton, 2018). While the argument could be made that
distraction in class is ever-present and has always been a problem (Kemp, 2008), various research
over the last decade shows that MT have been increasing the amount of distraction in education
(Douglas, Angel & Bethany, 2012; Spitzer, 2014), and a recent survey shows that over half of students
in class are continuously distracted by MT, despite the fact that they think MT in the classroom are
unavoidable (Hazelrigg, 2019). Because of the distracting function MT have, calls to action have been
made over the years to ban (certain) MT from education (Yamamoto, 2007; Rockmore, 2014; Reed,
2016; Selwyn and Aagaard, 2021), with France even banning mobile phones at certain levels of
education because of this back in 2018 (Hess, 2019). While it is debatable whether this is the right
solution to the problem, it is apparent that distraction is a consequential problem of the
implementation of MT in education.

The other major problem this research will touch upon is the combination of limited
effectiveness and our current overreliance on MT in education. As mentioned before, Richard Clark
strongly argued that MT would not have a specific effect on learning (1983). While there is some
research arguing for the limited effectiveness of MT, such as a study which found that physical note-
taking is actually more effective than digital note-taking (Mueller & Oppenheimer, 2014), the
research on both positive and negative effects of MT on education rebuke Clark’s statement
regarding complete ineffectiveness. However, Clark (1983) also states that pedagogy and the role of
the instructor in education (should) outweigh the importance of MT used, as he claims that MT are
‘mere vehicles that deliver instruction but do not influence student achievement any more than the
truck that delivers our groceries causes changes in our nutrition,” (Clark, 1983, p. 445). Van Lier
furthermore emphasized the supporting role MT should have, stating that MT ‘should not be cast as
an alternative to classroom teaching, or as replacing the teacher, but as a tool that facilitates
meaningful and challenging classroom work’ (Van Lier, 2003, p.2). Considering MT as supportive
towards the main pedagogic goal of education is extremely relevant nowadays, considering that the
current central role MT have gotten is, according to research, leading to a dehumanization of the
educational environments and a distortion of social interactions (Alhumaid, 2019). This trend of the
centralisation of MT in and dehumanization of education, has never been more apparent than
through the Covid-19 pandemic, wherein most education is held through MT. Both students (Rimer,

9
2020) as teachers (Cheung, 2021), have had mixed responses to this new learning environment.
However, almost two thirds of students felt that education is physical classrooms was better
(Chakraborty, Mittal, Gupta, Yadav & Arora, 2020), and many students and teachers named technical
problems (Wong, 2021a), a lack of social interaction with both fellow students as teachers (Wong,
2021b) and a lack of support and guidance (Schwartz, 2020) as major problems that plague online
education. The unarguable core and essential position MT now have in education is evidently causing
problems for both students as teachers, and the role of MT when physical classes become possible
again might need to be rediscussed.

All things considered, trends show that MT have over time become increasingly
important in all levels of education, and have now claimed an essential role. This has been possible
due to the fact that MT evidently aid in the learning process, allowing students to learn both from as
with MT. Despite this, MT are also a distracting factor, and our current overreliance on MT combined
with the ideal assisting role MT should have, has also led to several problems in education today.
Despite the problems however, the general consensus is still that MT are helpful in education.
Because of this, this thesis hypothesises as follows:

H1: People are predominantly positive about the implementation of media technology in
education

2.2) Artificial Intelligence in Education

Artificial intelligence is a rapidly growing and spreading technology in all types of sectors, and
it is expected to have a big influence on education in the coming years. Artificial intelligence, better
known as AI, is ‘the simulation of human intelligence processes by machines’ (Tucci, 2020), meaning
that it makes it possible for machines, including digital devices and man-made technology, to learn
from experience and adjust to new inputs to perform human-like tasks. Early development dates
back to the 1950s, when Alan Turing published the work ‘Computing Machinery and Intelligence’,
first introducing the concept of artificial intelligence (Yampolskiy, 2013). Since then, it has developed
immensely to what it is now, and in today’s world it is already a helpful tool in for example
healthcare, for example through being able to ‘suggest courses of medication based on real-time
biomarkers’ (Panesar, 2019, p.18). While AI has thus far not been implemented into education as
much as in other sectors, like for example transportation or the aforementioned healthcare (Sekar,
2018), AI in education was already a market worth 1 Billion USD in 2019, and is expected to grow
annually by 45% (Dukadinovska, 2020). It must be noted that AI in education is already used more
than people might expect, for example on things such as data collection on attendance and

10
assignments (Luckin, Holmes, Griffiths & Forcier, 2016). Despite this quick growth and current usage,
which shows the demand, there is also doubt amongst people whether this rapid development of AI
is a good thing. Notable people, such as Stephen Hawking and Elon Musk, warned about the dangers
of this rapid development and even think that AI could replace humans altogether (Gall, 2018; Marr,
2018). A recent global survey by Pega (2020) amongst consumers furthermore showed this fear of
the rapid development, with almost a quarter of respondents indicating they fear the enslavement of
humanity as the result of the rapid development of AI. While there is evidently hesitance regarding
the development of AI, it is apparent that its development and implementation in different sectors is
proceeding rapidly, with education likely dealing with major implementation in the foreseeable
future.

The implementation of AI in education is beneficial, as its strengths aid both students in

learning as teachers in teaching. The biggest advantages to AI are, arguably, its reduction in human
errors, its precision and its efficiency (Kumar, 2019), and all three of these traits make AI useful in
aiding education. Perhaps the biggest positive about AI in education would be the efficiency created.
For teachers, who a study showed spend 50 hours a week working (Bryant, Heitz, Sanghvi, & Wagle,
2020), AI could assist in reducing their time spent on non-teaching tasks such as administrative work
and marking tests, which a survey showed teacher spend a significant amount of time on
(Southworth, 2019). For students, AI could focus on identifying ‘what a student does and doesn’t
know through diagnostic testing and then developing personalized curricula based on each student’s
specific needs’ (Kulkarni, 2019). An efficient personalized curriculum tailored to students’ needs,
could be an answer to the relative ineffectiveness of the traditional standardized approach to
education (Rouhiainen, 2019). The effectiveness of AI is evident, as for example a study showed that
the educational language application Duolingo, which relies on AI to identify what a user does and
does not know, is more effective in language learning than a full university semester language course
(Vesselinov & Grego, 2012). The effectiveness and precision of AI go hand in hand, as the precision of
the aforementioned system where student’s needs are tailored would increase the efficiency of
learning. Precision education requires ‘accurate predictions of academic performance based on early
observations of the learning process’ (Tempelaar, Rienties & Nguyen, 2021, p. 109), something which
AI can and already does provide (Buckingham & Luckin, 2019). Lastly, this precise and effective
scenario in which AI identifies students’ strengths and needs could also address the much criticised
issue with standardized testing. Standardized testing has been criticized heavily as they are supposed
to determine whether the students control the material they are supposed to learn, but the
standardizes testing eliminates certain subjects such as arts and social studies (Delgado, 2018),
because it ignored different needs different students have and because it undermines actual

11
teaching as both teachers and students predominantly care about scores for accountability reasons
(William, 2010; Aydeniz & Southerland, 2012; Shelton & Brooks, 2019). With a precise and effective
curriculum for students, the focus would be readjusted to making sure students acquire necessary
information and skills (Kulkarni, 2019). All things considered, AI could be a useful enhancer in
education to make students learn more effectively and focus more on their strengths and
weaknesses

However, the implementation of AI in education also brings some notable negatives, and just
as with media technologies, the biggest negatives arise when we over rely on AI. The biggest
negatives to AI are, arguably, that they do not possess social intelligence and common sense
reasoning (Piletic, 2018). As these inherently human traits are (yet) impossible for AI to mimic,
people fear that an implementation of AI in education will lead to a dehumanization of education.
The dehumanization of education as a consequence of (media) technology has long been a talking
point (Nissenbaum & Walker, 1998), and throughout the last two decades arguments have been
made that this has been constantly happening (Haslam, 2006; Alhumaid, 2019). The argument could
also be made that AI would only make education more of a human environment, as AI would
automate the most repetitive and physical tasks in education today. This would then lead to a
situation where uniquely human skills, such as for example empathy, sense-making and creativity will
be developed more (Uria-Recio, 2019). These types of skills are, besides the dehumanization of
education, another notable problem to some regarding AI implementation into education. While AI
has showed to be of aid in learning material such as (sign) language (Vesselinov & Grego, 2012;
Paudyal, Lee, Kamzin, Soudki, Banerjee & Gupta, 2019) and mathematics (Gadanidis, 2017), AI does
possess certain essential 21st century skills, which are, often indirectly, taught in education
nowadays. These skills, also called the 4 C’s, are (1) critical thinking, (2) creativity, (3) collaboration
and (4) communication, which are all deemed important skills, yet they are unique to humans
(Germaine, Richard, Koeller, Schuert-Irastorza, 2016). Just like with media technologies, AI could be
important as an aid in education, but considering the important pedagogic function of education, it is
not advisable to over rely on it as it is in no way a replacement of human teaching.

In short, AI has developed massively over the last few decades, and especially in the last few
years. It is, at the moment, being implemented predominantly in sectors that require a high amount
of precision and efficiency, but it is also expected to have a big influence on education in the
upcoming years or decades. The efficiency and precision that AI bring could aid in educating students
more personally and effectively, and how it could automate certain processes leading to a reduced
workload for teachers, are sizeable positives. However, as AI is not a replacement for teaching,

12
society should be aware for an over reliance on AI, as it misses certain essential skills unique to
humans. Considering the positives AI does have, however, this research hypothesises as follows:

H2: People are predominantly positive about the implementation of artificial intelligence in
education

2.3 Innovation and Media technology adoption Frameworks

Because the implementation of new media technologies in education is a complex matter

concerning both innovation and technology adaptation, on both a societal as a personal scale,
elements from multiple frameworks regarding innovation/technology adaptation have been adopted
in this thesis. The first framework from which elements have been adopted is the diffusion of
innovations theory (Rogers, 2003). The theory, which was originated by Everett Rogers in 1962
(Rogers, 1962), seeks to explain the societal adoption process of innovations through a five stage-
process. These stages are knowledge/awareness (1), persuasion (2), decision (3), implementation (4)
and confirmation/continuation (5) (Rogers, 2003). Rogers actually uses the words innovation and
technology as synonyms, as most diffusion research involves technological innovations (Sahin, 2006).
Furthermore, the theory is described as most appropriate for investigating adoption of innovation in
education and educational environments (Parisot, 1995; Medlin, 2001). The second framework from
which elements have been adopted is the technology acceptance model (TAM). The model, which
was originally proposed in 1986 by Davis (Davis, 1986), but ultimately finalized in 1996 (Venkatesh &
Davis, 1996), is a widely used model to understand and predict how technology will be accepted and
adopted. The model provides two predictors of attitude towards technology, namely perceived
usefulness and perceived ease of use (Venkatesh & Davis, 1996). Both of these predictors are
influenced by what are called external factors, such as age, gender and experience. However, those
two predictors are, in the TAM, the essential predictors. The third and final framework incorporated
into this thesis is the unified theory of acceptance and use of technology (UTAUT). The theory and
model, devised by Venkatesh et al. in 2003, is in essence a extended and unified version of eight
previous models of innovation adaptation, including the TAM (Venkatesh, Morris, Davis & Davis,
2003). The model shares the two main predictors of attitude towards innovation that the TAM also
has, but adds a third predictor for attitude, namely social influence, which is considered only an
external factor in the TAM. A fourth predictor which directly predicts usage of technology is also
included in the UTAUT, but since this thesis focuses on attitude towards new media technologies,
this is not as relevant. Venkatesh et al. (2003) found that while any of the eight previous models they
extended upon could only explain for between 17 and 53 percent of the variance within attitude

13
towards new technologies, the UTAUT could explain for 70 percent of this variance, meaning it is a
more extensive and complete framework in comparison to the TAM.

While all three theories and models provide useful elements which are used in this thesis and
have a lot of commonalities, all three theories and models also have significant differences and
shortcomings which is why not just one theory was drawn upon. First of all, all three theories agree
on the fact that an increase in positive attitude towards an innovation or technology leads to an
increase in usage (Venkatesh & Davis, 1996; Venkatesh et al., 2003; Rogers, 2003). Through
backwards reasoning, this would mean that the attitude towards media technologies would be more
positive than towards AI, considering media technologies are implemented more in education, and
more of an ubiquitous presence in everyday life. Furthermore, all three theories agree on two types
of barriers or predictors of technology acceptance, whether they are called first and second order
barriers, or internal and external predictors, or something else. In all three theories, there are
barriers/predictors directly related to the technology, and more indirect, often social,
barriers/predictors more related to the person itself and several socio-demographic factors. Lastly,
the theories agree on a lot of these predictors/barriers, which will be expanded upon in the
upcoming sections. However, all three theories also deal with criticism and notable differences
between them. The diffusion of innovation, theory, while widely implemented and seen as a
standard for analysing a society’s response to new innovations/technology has been criticised for
overlooking the importance of social factors and how all predictors of innovation acceptance are
interrelated (Lyytinen & Damsgaard, 2001; Lundblad, 2003). The TAM is also a widely used model to
predict technology acceptance on a more personal level, and is praised for being an easily applicable
model with an comprehensive amount of variables (Lim, 2018). However, its simplicity is also its main
criticism, with several claims that it hold little practical value. It is argued that a model so simple can
not be expected to explain such a complex social phenomenon and the model is criticised for being
able to explain less variance in attitude, compared to for example the UTAUT (Venkatesh et al., 2003;
Bagozzi, 2007; Benbasat & Barki, 2007; Chuttur, 2009). Also, similar to the diffusion of innovation
theory, the neglect of the social context and socio-demographic factors in the TAM is also criticised
(Bagozzi, 2007). Lastly, we have the UTAUT, which is in essence a more extensive version of the TAM
and other models, and is lauded for explaining more variance in attitude and usage of technology,
and incorporating the social context, and socio-demographic factors, as one of the main predictors in
attitude towards new technology. However, this model is criticised because it is too extensive,
leading to what some claim a chaotic model (Bagozzi, 2007), and it is impractical to apply generally
because of the many interdepending variables in the model (Van Raaij and Schepers, 2008, Li, 2020).
Because of this, the main role social context and the correlating socio-demographic variables it has in

14
attitude towards new technology are adapted, but most other variables are emitted from this
research.

In this research, elements from three prominent theories on innovation/technology

adaptation are implemented in order to try and establish the most important predictors for attitude
towards media technologies and AI in education. From the diffusion of innovation theory, the central
role of experience in predicting attitude is implemented, and its general societal outlook on
innovation adaptation. From the TAM, the simplicity of the model along with its two main predictors
are implemented. Lastly, from the UTAUT, the important role of social context along with several
socio-demographic factors as a main predictor rather than an external variable is implemented, in
order to see how important the social context on its own is. While the three theories differ on
significant points, major predictor for attitude towards new innovations/technologies are common,
and also the reasoning that more experience with a technology leads to a more positive attitude.
Because of this, and the fact that media technologies are more ubiquitous and have been used
longer than AI, this thesis hypothesises as follows:

H3: People are more positive about the implementation of media technologies in education, than
they are about the implementation of AI in education.

2.4 Technology-related predictors for attitude

While there are a lot of interconnected variables that come up in the aforementioned
theories, perceived usefulness is consistently one of the main predictors of attitude towards new
(media) technology. Perceived usefulness or performance expectancy is defined in the extended
UTAUT as the perceived ‘degree to which using a technology will provide benefits to consumers in
performing certain activities’ (Venkatesh, Thong & Xu, 2012, p.159). Perceived usefulness or
performance expectancy is, according to both the TAM and the UTAUT, influenced by many external
variables, which include image of the innovation/technology, its relevancy to one’s job, the
demonstrability of results and experience with the innovation/technology (Venkatesh & Davis, 1996;
Venkatesh et al., 2003; Venkatesh, Thong & Xu, 2012). This predictor has, according to the UTAUT,
been consistently the strongest predictor of attitude towards new technologies (Venkatesh et al.,
2003). The effect of perceived usefulness on attitude towards a certain (media) technology has been
widely explored and tested, for example regarding the attitude towards laptops (Moses, Wong,
Bakar, Mahmud, 2013), e-books (Letchumanan & Muniandy, 2013), online shopping (Ramayah &
Ignatius, 2005), and even educational technology (Hart & Laher, 2015). Because of the wide

15
theoretical support and empirical evidence, it is expected that perceived usefulness has a positive
correlation with attitude towards both media technologies and AI in education.

The relation between perceived ease of use or effort expectancy and a positive attitude
towards technology is also established both through theory as empirical evidence. Perceived ease of
use is defined as "the degree to which a person believes that using a particular system would be free
of effort’ (Davis, 1993, p. 320). Similar to perceived usefulness, perceived ease of use is also
influenced by many external factors. Age and gender are presumed to have a sizeable influence on it,
but what perhaps is more important are the grouped external variables of ‘usability’. Under these
variables of usability fall perspicuity, which is the extent to which a technology is clearly expressed
and presented, and dependability, which is the extent to which a technology is dependable (Mlekus,
Bentler, Paruzel, Kato-Beiderwieden & Maier, 2020). Much like perceived usefulness, perceived ease
of use is also supported by both theories and empirical evidence, for example researching its effect
on attitude towards 3G (Suki & Suki, 2011), mobile banking (Raza, Umer & Shah, 2017), and once
again educational technology (Joo, Park & Lim, 2018). Because of this, it is expected that perceived
usefulness too, has a positive correlation with attitude towards both media technologies and AI in
education.

The last directly related predictor for attitude towards media technologies and AI in
education is experience, meaning the amount of interaction one has had with media technologies
and AI during their education. While this was initially seen as an external predictor, only indirectly
influencing attitude towards innovation, the revised UTAUT does include experience and habit as
direct predictors for attitude (Venkatesh, Thong & Xu, 2012), but the definition of both are for this
topic slightly outdated, and a mix of the two is better suited for this research. Experience in the
revised UTAUT is defined as ‘the passage of time from the initial use of a technology by an individual’
(Venkatesh, Thong & Xu, 2012, p.161). However, according to this, older adults, who have been
exposed to media technologies for longer than younger adults have, would have more experience
with this than younger adults. Data showing media technology usage and competence however
shows that younger adults are more competent with media technologies and also use them a lot
more (Smith & Anderson, 2018). Habit then, is explained as both prior behaviour (Kim and Malhotra,
2005), but also as ‘the extent to which an individual believes the behaviour to be automatic
(Venkatesh, Thong & Xu, 2012, p. 161). While predominantly the prior behaviour is relevant to this
study, the perceived automatic behaviour is not applicable since most adults are not anymore in
their respective education system, thus restricting media usage in this field from becoming a habit.
Experience then, in this research, is seen as the amount of meaningful experience one has had with a
certain media technology. Previous research has shown that increases in experience lead to for

16
example better attitudes towards personal computers (Thompson, Higgins and Howell, 1994), is a big
influence on direct attitudes and behaviour (Haselhuhn, Pope, Schweitzer & Fishman, 2012), and
more specifically on attitudes towards media technologies (Mlekus et al., 2020). Because of this, this
research hypothesizes that direct experience with media technologies and AI during one’s education
leads to a better attitude towards media technologies and AI in education.

This chapter discussed three predictors, directly related to one’s relationship with MT and AI
in education, that are expected to influence the attitudes towards MT and AI in education. Perceived
usefulness, perceived ease of use and direct experience all have theoretical and empirical evidence
showing that an increase of any of these predictors, likely leads to an increase in positive attitudes
towards MT and AI in education. Hence, this research hypothesizes as follows:

H4: People with higher perceived usefulness of MT and AI in education, have a more positive
attitude towards MT and AI in education.

H5: People with higher perceived ease of use of MT and AI in education, have a more positive
attitude towards MT and AI in education.

H6: People with more direct experience with MT and AI in education, have a more positive
attitude towards MT and AI in education.

2.5 Socio-demographic and indirect predictors

Besides the predictors directly related to MT and AI in education, there are several other
sociodemographic and indirect predictors, amongst which age is a prominent and important one. In
the TAM and the UTAUT, age is seen as an external predictor, which moderates the effect of the
direct predictors mentioned in the previous chapter (Davis, 1993, Venkatesh et al., 2003). However,
based on previous research and present data, this research hypothesises that age is also a direct
predictor of attitudes towards MT and AI in education. Both media technology usage as competence
are significantly lower amongst older adults, and that older adults need a lot more training to
become competent with media technologies (Broady, Chan & Caputi, 2010; Kim & Choudhury, 2020).
In the theories adapted in this thesis, a lower usage is the result of a worse attitude towards those
technologies, so it being a fact older adults use less MT, this would have to be the result of a worse
attitude towards those technologies. But the influence of age goes further than just usage and
competences with media technologies and AI. Research shows that as we get older, we grow more
conservative (Tilley, 2015), we become less open to change (Schwaba, Luhmann, Denissen, Chung &

17
Bleidorn, 2018), become more set in our ways and less open to new experiences (Mühlig-Versen,
Bowen & Staudinger, 2012). Because of this data and research, it is expected that age is a negative
predictor of attitudes towards MT and AI in education.

Besides age, one’s personality and accompanying values are also hypothesized as predictors
for one’s attitude towards MT and AI in education. While it is difficult to quantify and recognize
global values, especially in intercultural research, theory does exist which identifies universal values,
recognised throughout all major cultures. The theory of basic human values, which is develop by
Shalom Schwartz in 1992, identifies 10 distinct motivational values, grouped in four higher order
groups (Schwartz, 1992), out of which two are relevant to this research. The theory, which was
further refined in 2012 (Schwartz, 2012; Schwartz, Cieciuch, Vecchione, Davidov, Fischer, Beierlein,
Ramos, Verkasalo, Lönnqvist, Demirutku, Dirilen-Gumis & Konty, 2012), has been widely used in
cross-cultural research (Berry, Poortinga, Pandey, Segall & Kâğıtçıbaşı, 1997) and the ten universal
values are recognised throughout all major cultures (Schwartz, 2012). Two of the four higher-order
groups which are relevant for innovation adoption are openness to change and conservation.
Openness to change, which contains the values of self-direction and stimulation, ‘emphasizes the
readiness for new ideas, actions, and experiences. (Schwartz et al., 2012, p. 668)’. This directly
contradicts the higher order group of conversion, which emphasizes self-restriction, order and
avoiding change (Schwartz et al., 2012). Schwartz’ human values, and especially openness to change
and conservation, have been widely used to explain adopting innovation (Isomursu, Ervasti, Kinnula
& Isomursu, 2011; Barbarossa, De Pelsmacker & Moons, 2017). Because of this, it is expected that
openness to change and conservation, as described in Schwartz’ theory, both influence one’s attitude
towards MT and AI in education.

The last indirect predictor which is hypothesized to influence attitudes towards MT and AI in
education, is mobile self-efficacy. Mobile self-efficacy, a term which stems from Bandura’s research
on self-efficacy (Bandura, 1977; Bandura, 2010), is the amount to which an individual believes in
his/her capabilities to effectively use mobile devices (Keith, Babb, Furner & Abdullat, 2011), for
example MT such as laptops, phones and tablets. There is numerous existing research that shows
that people who have a higher amount of self-efficacy, are more positive towards (using) for example
mobile learning (Yang, 2012; Yorganci, 2017), but also for teacher’s usage of technological devices in
classes (Kwon, Ottenbreit-Leftwich, Tari, Khlaif, Zhu, Nadir & Gok, 2019). Because of the existing
research, it is expected that mobile self-efficacy too effects one’s attitude towards MT and AI in
education.

18
 This chapter covered three major predictors, seen as indirect or external predictors in the
frameworks that were adapted in this research, but which this research hypothesises to be direct
predictors of one’s attitude towards the usage of MT and AI in education. The three predictors of
age, values, and mobile self-efficacy are , based on existing research, all predicted to influence
attitudes towards MT and AI in education, and there this research hypothesises as follows:

H7: People of higher age have a less positive attitude towards MT and AI in education.

H8: People with a higher reported score for openness to change, have a more positive attitude
towards MT and AI in education.

H9: People with a higher reported score for conservatism have a less positive attitude towards MT
and AI in education.

H10: People with more mobile self-efficacy have a more positive attitude towards MT and AI in
education.

3) Thesis methodology

3.1) Choice of method

For this research, a quantitative approach using self-administered online questionnaires was carried
out. Quantitative methods were best suited for this research, as one of the two main goals was to
quantify opinions about media technologies and AI in education, and generalize them to a certain
degree. The other main goal of this research was to test the influence of certain predictors on the
attitude towards media technologies and AI in education, based on hypothesises coming forth out of
theoretical and empirical support. Both of these main goals fit can only be achieved through
obtaining a large number of data, thus explaining the choice for quantitative methods.

Online self-administered questionnaires were chosen as the means to get this large amount
of data, because they are ‘a very useful tool that allow large populations to be assessed with relative
ease’ (Jones, Baxter & Khanduja, 2013). Furthermore, they are easy to conduct, and can be quickly
distributed with global reach, which in the context of the current pandemic, is a big advantage
(Andrade, 2020). There are also downsides to the choice for this method, as for example the
increased difficulty to generalise outcomes (Andrade, 2020), and the increased difficulty in engaging
respondents for a longer amount of time (Guin, Baker, Mechling & Ruyle, 2012). However, with a
careful sampling process opted to select a generalizable sample, and through keeping the online
survey at a low completion time, these disadvantages were not seen as significant problems in this

19
research. The online self-administered questionnaire implemented in this research asks about
people’s general opinions on both MT and AI in education, their experience and perceived
competence with MT and AI (in education), and people’s motivational domains through Schwartz’
human values. This research acknowledges there might be some age bias regarding one’s opinion of
MT and AI in education, considering it would have been impossible for most older age groups to have
experience with those technologies in education. Because of this, these age groups might have found
it harder to relate as to how useful and easy to use MT and AI are in education, and it might also
capture their general attitude towards MT and AI and not necessarily their attitude towards MT and
AI in education.

3.2) Research Unit

The research units in this research are adults in the Netherlands, the United Kingdom, Ireland, the
United States, Canada and Australia. These 6 countries are chosen not only because they are, to a
certain degree, comparable on an economic and cultural level (Nijman, Muller & de Blij, 2016), but
also because their education systems and the relation those have with media technology and AI are
relevant. While there are many differences between the education systems of the specified
countries, they are in their basics and essence comparable if we look at quality. While there is no
absolute way to quantify which countries have the best education systems, in various research over
the last few years, all 6 specified countries consistently score highly, with all countries except Ireland
even ranking in the top 10 on a regular basis (Ireland, 2020). Furthermore, all countries can, arguably,
be considered frontrunners on the acceptance and usage of both media technology and AI in their
education system. The countries specified all use educational media technologies to a relative high
extent (Hamidi, Ghorbandordinejad, Rezaee & Jafari, 2011; Trucano, 2014), EdTech companies
predominantly originate in these countries (IPRAN, 2020), and these countries also rank quite highly
on the ‘AI readiness index’, which is an index based on how ready governments are to implement AI
into various sectors (IDRC, 2020).

A conscious decision was made to not include other countries which are comparable on a
cultural and economic level and are also frontrunners on media technologies and AI in education.
Countries such as Germany, Norway, Denmark and more all have similar quality education systems to
the ones from the countries in the research unit, yet were not included in the research unit. The
main reason for this is in order to be able to offer all respondents a survey in their own domestic
language. This is the case because the usage of the English language in cross-national research
obscures national differences (Harzing, 2005), but also because the average person from non-english

20
countries should not be assumed to speak English, despite that in a lot of these countries the English
proficiency is rather high (Clark, 2019).

3.3) Sampling

The sampling for this research existed out of two separate stages. The first sampling stage
entailed nonprobability sampling through chain/snowball sampling, while the second stage took
place through the crowdsourcing platform Prolific. The first stage produced about a 125 respondents,
which happened through volunteers who sent it through to family and friends. Initially, a group of
between 40-50 people were reached out to, mostly coming from the Netherlands, Australia, the
United Kingdom and Canada, but also people coming from both the United states and Ireland. This
sample consists out of friends, (ex-)colleagues and acquaintances, from current or former
(international) work, travel and study of the researcher. The order of reaching out to this group of
respondents was based on nationality, making sure to have a sample from every country as large as
possible before sampling respondents from other countries, to keep the amount of respondents
from different countries comprehensible. This initial group of the first sampling stage then
distributed it further amongst their friends and families. These respondents, of which there were
approximately 80, was more diverse, ranging more in age groups and education level. Reaching out
to this the first stage of the sampling stage took approximately 9 days, between the 17th of May and
the 26th of May. By implementing this strategy for the first stage of sampling, the costs of the study
were limited and a large sample of predominantly younger respondents was collected. While this
kind of nonprobability sampling is not ideal for the representativeness and generalizability of the
sample, as the initial volunteers are mostly highly educated, this approach was necessary to ensure a
large international sample without high costs. The first part of the sample consisted out of
predominantly younger people, and most respondents came out of the Netherlands, Australia, the
United Kingdom and Canada.

The second stage of the sampling, which occurred through Prolific, produced approximately
40 people, selected through selection criteria. Prolific is a crowdsourcing platform, through which
people can sign up to fill in questionnaires, for which they get a small monetary compensation per
questionnaire filled in. Prolific enables high quality data collection from a big diverse population.
Target participants can be targeted accurately through selecting certain selection criteria for
respondents. This part of the sampling process, and the accompanying selection criteria, had the
purpose of making the sample more representative regarding predominantly age, but also
nationality. The first part of the sampling produced an insufficient amount of respondents in the
highest age category, and also from Irish and American nationality. Because of this, the second part

21
of the sampling and its selection criteria were used to create a more coherent, representative and
generalizable sample. There are some methodological issues with selecting predominantly older
people through a crowdsourcing platform. The biggest issue with this is that respondents who fill in
the survey through prolific, must be familiar with this website, the internet and have to be able to
use media technologies to be able to participate. Therefore, a decent amount of new media
technologies competence must be acknowledged, and considering that this thesis hypothesises that
competency with media technology results in a more positive attitude towards media technologies
and AI in education, the issue of representativeness for older respondents occurs. While this is a
shortcoming that this research acknowledges, there was no alternative feasible strategy to generate
respondents from all age groups, thus this research accepts this negative and pays attention to it in
discussing the results.

3.4) Operationalisation

3.4.1) Schwartz’ human values

Measuring universal human motivational domains has traditionally been difficult, considering
the high amount of differences between cultures nationally, but also within nations. The theory of
Schwartz’ human values has brought this back to 7 main motivational domains recognized across all
cultures. To measure Schwartz’ human values, two separate types of scales already exist. The first
one of these, is the Schwartz’ values survey (SVS), where respondents have to attach a score to all
the human values, given to them in the survey, which there are 30 of (Schwartz, 1992; Schwartz,
2006). However, this type of scale has been criticized as it is too explicit, too incomprehensible and
too long, leading to a bad response rate (Lindeman & Verkasala, 2005). The alternative to this, and
from which this research takes elements, is the portrait values questionnaire (PVQ). The PVQ was
created to reduce the complexity of the scale, making it easier for people to fill in, and not directly
ask for one’s values but implicitly ask for them (Knoppen & Saris, 2009; Schwartz, 2012). While
originally intended for children, the PVQ ‘works equally well with adults in representative national
samples. (Schwartz, 2012, p.11)’. An 8-item scale on a 7 point Likert scale (1 = strongly disagree, 7 =
strongly agree) was created to measure openness to change, and an 8-item scale on a 7 point Likert
scale (1 = strongly disagree, 7 = strongly agree)was created to measure conservatism. For openness
to change, 8 items from the PVQ as seen in the European social survey (ESS), (Knoppen & Saris, 2009;
Bilksy, Janik & Schwartz, 2011), were adopted, coming from the subscales of self-direction and
stimulation. Statements in this scale included statements such as ‘I am always looking for new things
to try’ and ‘It is important to me to see life as an adventure’. The Cronbach’s alpha for this scale
measure openness to change reported at .858. For conservatism, 8 items from the same PVQ as seen

22
in the ESS were adopted, coming from the subcategories of security, conformity and tradition.
Statements in this scale included statements such as ‘I believe people should be satisfied with what
they have’ and ‘I find it important that the established order is protected’. The cronbach’s alpha for
this scale measuring conservatism reported at .872.

3.4.2) Experience

While there are some scales to measure parts of one’s experience with predominantly MT,
such as for example the spatial presence experience scale (SPES) (Hartmann, Wirth, Schramm,
Klimmt, Vorderer, Gysbers, Böcking, Ravaja, Laarni, Saari, Gouveia & Sacau, 2015), a conscious
decision was construct a scale as simple as possible, simply asking respondents to which degree they
have used the three most well-known and widely implemented examples of MT and AI in education.
A three-item scale was developed for this research, asking respondents for their experience with
certain technologies/applications during their time in education on a 5 point Likert scale (1 = never, 5
= most of the time). For MT, the three technologies that were chosen were laptops, tablets and
interactive whiteboards. While there is no official data on what is used most in education, research
and articles often mention laptops and tablets as main media technologies and reports show their
usage is going up (Truong, 2020). The interactive whiteboard was chosen instead of for example a
mobile phone, because mobile phones might be too familiar to the other two, and interactive
whiteboards, either through original display or through a beamer, are widely used in education
(Chade, 2021). The Cronbach’s alpha for this scale reported at .833. For measuring experience with
AI, it was important to pick three application of AI that would be easily understandable for everyone,
considering data shows a lot of people actually do not know what AI exactly is (Mozilla, 2019; Pega,
2020). The three applications that were picked were learning applications such as Duolingo,
automatic grading systems and online custom quizzes and tests based on strengths and weaknesses,
as these are, while still relatively new, some of the more used application of AI in education (Marr,
2018; Fagella, 2019; Gupta, 2020). The Cronbach’s alpha for measuring experience with AI during
one’s education reported at .879.

3.4.3) Mobile self-efficacy

At the point of writing, no widely used and validated scale exists to measure mobile self-
efficacy. For this research, a scale was constructed to measure one’s amount of mobile self-efficacy.
A six-item scale was constructed, based on Bandura’s research on self-efficacy, and other research on
the effects of mobile self-efficacy and new media literacy (Bandura, 1986; Ozturk, Bilighan, Nusair &
Okumus, 2016; Koc & Barut, 2016; Nikou & Economides, 2017; Chao, 2019). The scale, which focuses
on one’s belief in themselves to meaningfully and effectively interact with media technologies, draws

23
from the new media literacy scale too. It does this because mobile media technologies nowadays are
a ubiquitous element in everyday life (Sigerson & Cheng, 2018), and it is therefore important to test
whether someone believes (s)he can effectively and meaningfully use mobile media technologies
rather than just use them. The scale, which respondents answered on a 5 point Likert scale (1 =
strongly disagree, 5 = strongly agree), asked questions such as whether the respondent was
proficient in creating accounts/profiles on websites/applications and whether they were proficient in
finding the online information they want or require. Furthermore, because people, and then
especially the lowest performers, overestimate their own digital literacy skills (Mahmood, 2016), the
question are phrased as simple as possible, in a manner that it does not seem condescending or bad
if someone were to fill in a negative answer. For example, the scale asked whether the respondents
stays up to date with changes in the media, and whether the respondent would sometimes find it
difficult to find the online information they want. The Cronbach’s alpha for the scale reported at
.879.

3.4.4) Perceived usefulness and perceived ease of use

As perceived usefulness (PU) and perceived ease of use (PEOU) are the two main predictors
for behavioural attitude in both the TAM as the UTAUT, both predictors have been measured often in
previous research and a scale can be constructed from these previous researches. A scale for both PU
and PEOU for both MT and AI was developed, however the differences between scales for MT and AI
were solely semantic. A six-item scale was constructed, on a 5 point Likert scale (1 = strongly
disagree, 5 = strongly agree) to measure one’s PU and PEOU for both MT and AI in education. The
scales, which were mostly adopted from the frameworks embedded into this research and previous
research related to the effect of PU and PEOU (Venkatesh et al,2003; Cimperman, Brenčič & Trkman,
2016; Šumak & Sorgo, 2016; Hogue & Sorwar, 2017; Khalilzadeh, Ozturk & Bilighan, 2017), but was
slightly adjusted to fit the topics of this research more clearly and effectively. The Cronbach’s alpha
for the scale that measured the PU of MT was .939, where this was .919 for the scale that measured
PU of AI. The scales that measure PEOU for MT and AI, had an Cronbach’s alpha of .870 and .816
respectively.

3.4.5) Attitude towards media technologies and AI

Considering that the attitude towards MT and AI is the most essential part of this research, it
was vital that a reliable and trustworthy scale was implemented to measure one’s attitude towards
MT and AI in education. While scales measuring attitudes towards (media) technologies have
historically been quite sparse, despite attitude being seen as a major factor in innovation adaptation
(Edison & Geissler, 2003), there is one major scale which is widely used in research, namely the

24
media and technology usage and attitudes scale (MTUAS) (Rosen, Whaling, Carrier, Cheeever &
Rokkum, 2013). The MTUAS, which consists out of 60 items divided into 15 subscales, has 4 subscales
for attitude towards media (technologies). From these subscales for attitude,6 items were taken,
slightly adjusted and implemented into this research on a 7-point Likert scale (1=Strongly disagree, 7
= Strongly agree). The items were slightly adjusted to fit the topic of this research, and play into the
biggest positives and negatives about both MT and AI in education. For example, two of the items
correlated with a negative attitude in the MTUAS attitudes subscales mention topics such as isolation
and wasting time. Considering the biggest criticism towards MT and AI in education are that they are
too distracting and dehumanize education, the items are slightly adjusted to fit this exact topic to
make more sense. The Cronbach’s alpha for attitude towards MT reported at .844, and the
Cronbach’s alpha for attitude towards AI reported at .904.

3.5) Survey design

Some conscious considerations were made in order to make the survey valid,
methodologically solid, and in order to have a high finish rate on the survey. Firstly, the survey starts
with explaining the context of the research, a small description of what it is about without actually
giving away what the underlying goal of the research is. This is done, as when context is created early
and effectively, ‘the less likely people will be to dismiss the questionnaire before they even start
responding’ (Thayer-Hart, Dykema, Elver, Schaeffer & Stevenson, 2010). Context is repeatedly
established in the survey, especially at the parts before questions specifically about MT and AI in
education. The difference between MT and AI is established clearly and through examples which are
as common as possible, considering that a high amount of people do not exactly know what AI is
(Mozilla, 2019; Pega, 2020). Attention was also given to the order of the questions. The survey starts,
as almost all surveys, with the socio-demographic factors, but after this the first scale that is brought
up is the one for Schwartz’ human values. This is done on purpose, because surveys should unfold in
a logical order, preferably starting with easy to answer, engaging questions that respondents will be
interested in (Pew Research Center, 2021; Qualtrics, 2021). Questions about one’s personality are
easy to answer, and people like answering questions about their personality (Dahl, 2017). Leading
from this, the other predictors come first, followed by ultimately the questions about MT and AI in
education, considering these take the most thought, and are therefore arguably the hardest. In order
to avoid that respondents figured out the pattern of scales, knowing which answers related to high
values and which to low values, reverse coding was applied for at least one, but mostly for two or
three items per scale. This is done as it increases validity of the scales and it negates response style
bias (Álvarez, Pedrosa, Lozano, Cuteo, Izquierdo & Fernandez, 2018). The questionnaire was carried
out between the 17th of May and the 28th of May. The questionnaire had a response rate of about

25
70%, after most people did not finish the questionnaire at either the questions about MT and AI, or
right at the beginning.

3.6) Tools, reliability and validity

For constructing the questionnaire, the platform Qualtrics was used, who suggest certain
adjustments in order to maximize response and engagement. To supplement the sample and make it
more representative, the crowd-sourcing platform prolific was used, where respondents receive a
small amount of payment for a valid response. To analyse the gathered data, the program SPSS was
used. The descriptive, such as the means and standard deviations, of all scales were reported.

The reliability of the different scales was assessed and reported through the Cronbach’s
alpha. The content validity of these scales is established through using the Likert scale verified, and
through mostly implementing (elements of) scales from previous research and literature. The validity
of the respondents was assessed carefully, and about 25 responses were deleted, because they
showed clear signs of response style bias or completed the questionnaire in a time that was deemed
unrealistic. Considering that respondents through prolific have an economic incentive to fill in the
survey, these respondents were assessed most carefully before administering them to the date. No
attention was given to the responses, but when for example a respondent of 65+ year old, or any age
for that matter, completed the survey in under two minutes and responded the highest or lowest
answer for every question, the data was deemed invalid and deleted.

4) Results

4.1 Descriptive statistics

4.1.1) Sample descriptives

This research administered 174 complete and valid responses. From these 174 respondents,
9 originated from countries other than the countries that were specified in the research unit, and
therefore these 9 respondents were excluded from the analyses, leaving 165 respondents available
for analyses (N=165). Gender was evenly distributed, with 83 of the respondents were male (50.3%),
81 were female (49.1%), and 1 being Non-binary (0.6%). The age group of 18-24 was represented
most with 61 respondents (37%), there were 31 respondents aged 25-39 (18.8%), 35 between 40-64
(21.2%), and 38 aged 65 and above (23%). The nationality of respondents was also quite evenly
distributed, with 36 being Dutch (21.8%), 28 being American (17%), 26 being Canadian (15.8%), and
25 from each Australia, Ireland and the UK (15.2% each). Lastly, when looking at the highest

26
education completed, most respondents finished high school with 54 respondents (32.7%), followed
by 46 who completed a bachelor of science (27.9%), and 35 who completed a practical education or
community college (21.2%). There were 12 respondents who had completed a master degree (7.3%),
9 with a bachelor of applied science (5.5%), 8 who finished no education (4.8%) and 1 who finished a
doctorate (0.6%).

4.1.2) Scale descriptives

The descriptive statistics for some of the measures scales resulted in somewhat surprising
results. The scale used to measure openness to change had a mean of 4.33, with a standard deviation
of 1.20. The average score for conservatism was 4.61 with a standard deviation of 1.26. Mobile self-
efficacy was quite high at a mean of 3.60, with a standard deviation of 1.38. The reported experience
with MT and AI were both quite low, with a mean of 2.37 for MT with a standard deviation of 1.38,
and a mean of 1.85 with a standard deviation of .95. This shows that not a lot of people have
experience with AI in education, and that the results for experience with MT in education are quite
mixed, considering the rather high standard deviation. The minimums and maximums, and the other
descriptive statistics, for all scales are found in table 1.

This research hypothesised that people would be predominantly positive towards both MT
and AI in education, based off of the proven positive influences it can bring to education. However,
the descriptive results for attitude towards MT and AI in education may have been the most
surprising of all scales. The mean for attitude towards MT in education laid at 4.15, with a standard
deviation of 1.28. Considering 4 corresponds a neutral score, and this thesis hypothesised that
people would be predominantly positive towards MT, H1 is rejected. The mean for attitude towards
AI in education reported somewhat lower, at 3.76 with a standard deviation of 1.55. This shows that
the respondents were slightly more negative towards AI in education than they were towards MT in
education, and also slightly negative in their view towards AI in education overall. As this thesis
hypothesised that people would be predominantly positive towards AI, H2 is also rejected. This
research however also hypothesised that people would be more positive towards MT in education
than they would be towards AI in education, given to which extent they are both implemented in
various education systems at the moment. The mean for attitudes towards MT in education was
higher than the mean for attitudes towards AI in education, and the standard deviation lower.
Besides this, a paired t-test was conducted to test whether the difference in attitude towards MT and
AI was significant. The test proved significant (t(130)=7,19, p<.001), and therefore H3 is accepted.

27
 Minimum Maximum Mean Std. Deviation
Openness to change 2.00 6.63 4.33 1.20
Conservatism 1.88 6.75 4.61 1.26
Mobile self-efficacy 1.00 5.00 3.60 1.38
MT experience 1.00 5.00 2.37 1.38
AI experience 1.00 4.67 1.85 .95
MT attitude 1.67 6.33 4.15 1.28
AI attitude 1.33 6.5 3.76 1.55
Table 1: Characteristics of used scales

In the next few sections, different analyses are conducted to test the significance of the hypothesised
predictors, but also to test the strength of the predictors and to find out the key influences on
attitudes. How this is done, is by first going through all predictors singularly, testing if the predictors
are associated with attitude towards MT and AI on their own (section 4.2, 4.3, 4.4 and 4.5). After this,
predictors are grouped together to see which predictors are key in influencing attitude, and if
perhaps predictors that are significant on their own are no longer significant when grouped with
others (section 4.6).

4.2) Effect of perceived usefulness and ease of use on attitude

Based on the various frameworks and theories embedded into this research, the hypotheses
were made that perceived usefulness (PU) and perceived ease of use (PEOU) would have a positive
associations with one’s attitude towards MT and AI in education. To test this, two simple linear
regressions were run. One with attitude towards MT as the dependent variable, and perceived
usefulness of MT as the independent variable. The other with attitude towards AI as the dependent
variable, and perceived usefulness AI separately as the independent variable. For both attitude
towards MT as AI, the normality of errors and the constant error variance in the two different
analyses was mostly held (except for a few deviations at the tails of the distribution for attitude
towards MT), and in both analyses the constant error variance (homoscedasticity) was not violated.
For attitude towards MT in education, the effect of PU proved significant (F(1,163)=753,07, p<.001).
The adjusted R2 reported at .821, and the standardized coefficients reported at .907, with a
significance of p<.001. This means that the increase of PU lead to an significant and strong increase
(b*=.907) of attitude towards MT in this sample. For the effect of PU on attitude towards AI, a similar
significant association was found (F(1,163)=739,08, p<.001). The adjusted R2 reported at .818, and
the standardized coefficients laid at a significant .905, p<.001. This means that the increase of PU

28
also lead to a strong and significant increase (b*=.905) of attitude towards AI in education. Because
PU was a strong and significant predictor for both attitudes towards MT and AI in education, and a
large amount in the variance could be explained by PU, H4 is supported.

For the effect of PEOU on attitude towards MT and AI in education, once again two simple
linear regressions were run. One with attitude towards MT as the dependent variable, and PEOU of
MT as the independent variable. The other with attitude towards AI as the dependent variable, and
PEOU of AI as the independent variable. Homoscedasticity was established for both models, however
the normality of errors was not held for attitude towards AI, with some deviations, mostly at the
middle of the distribution, violating the normality. Because of this, the coefficient estimates are
possibly biased and the results for attitude towards AI are somewhat tentative. The effect of PEOU
proved significant on both attitude towards MT (F(1,163)=438,79, p<.001) as on attitude towards AI
(F(1,163)=264,82, p <.001). The adjusted R2 for attitude towards MT was .727, where this was .617
for attitude towards AI. For both attitude towards MT (b*=.854) as attitude towards AI (b*=.787),
PEOU was a strong and significant predictor, however for attitude towards AI the results are
somewhat tentative. Because of the strong and significant effect, H5 is, tentatively, supported.

While no specific hypotheses existed regarding the strengths of certain predictors, this
research did aim to find out the key predictors and influences in predicting attitude towards
innovations. To test the strengths of PU and PEOU combined on attitude towards both MT and AI,
two multiple regression analysis were run. The first with attitude towards MT as the dependent
variable, and PU and PEOU of MT as the independent variables. The second with attitude towards AI
as the dependent variable, and PU and PEOU of AI as the independent variables. For attitude towards
MT, the model with both PU and PEOU proved significant (F(2,162)=453,85,p<.001). The adjusted R2
laid at .847, meaning that this model including both PU and PEOU as independent variables explained
more variance in the data for attitude towards than the models including PU and PEOU separately
did, albeit only slightly more than the variance that was explained by PU of MT on its own (R2 =.821).
In this model, PU of MT was still strongly and positive associated with attitude towards MT (b*=.648,
p<.001), but PEOU of MT was now moderately and positively associated with attitude towards MT
(b*=.305, p<.001), where this was strongly and positively associated when analysed as a single
predictor (b*=.854). This means that for attitude towards MT, PU seems a significantly stronger
predictor than PEOU is. For attitude towards AI, the model including both PU and PEOU of AI as
independent variables also proved significant (F(2,162)=435,69, p<.001). The adjusted R2 reported at
.841, which is once again higher than the variance explained by either PU or PEOU of AI as single
predictors. In this model, PU of AI once again was strongly and positively associated with attitude
towards AI (b*=.726, p<.001), but PEOU of AI was now only weakly positively associated with attitude

29
towards AI (b*=.237, p<.001). This means that for attitude towards AI, PU is once again a stronger
predictor than PEOU, despite both of these predictor being strongly and positively associated with
both attitude towards MT as AI as single predictors.

4.3 Effect of mobile self-efficacy and experience

This research hypothesised that mobile self-efficacy would have a positive association with
attitude towards MT and AI in education. To test this, simple linear regressions were run with mobile
self-efficacy as the independent variable, and attitude towards MT and AI in education as the
dependent variable. Homoskedasticity was established in both cases, however the normality of
errors was not held for attitude towards MT, with deviations especially in the middle, leading to
possibly biased and tentative results, probably as the consequence of outliers. While there were
some deviations at the tails for attitude towards AI, the normality of errors was mostly held here.
The effect of mobile self-efficacy proved significant on both attitude towards MT in education
(F(1,163)=108,99, p<.001) as on attitude towards AI in education (F(1,163)=118,02, p<.001). The
variance for attitude towards AI was explained slightly more (R2 of .416) than the variance for
attitude towards MT (R2 of .397) by the effect of mobile self-efficacy, but mobile self-efficacy proved
a significant, strong influence of both attitude towards MT (b*=.633) as attitude towards AI
(b*=.648), with the notion that the results for attitude towards MT are somewhat tentative. Because
of these clear and decisive results, H10 is tentatively supported.

Furthermore, this research also hypothesised that direct experience with MT and AI in
education would also influence one’s attitude towards MT and AI in education. Simple linear
regression were run with attitudes towards MT and AI as the dependent variables, and experience
with MT and AI in education as independent variables. Homoskedasticity was established and the
normality of errors was not violated in either case. Once again, the effect of experience proved
significant on both attitude towards MT (F(1,163)=260,54, p <.001) as on attitude towards AI
(F(1,163)=166,32, p <.001) in education. For attitudes towards MT, the adjusted R2 was .613, where
this was .502 for attitudes towards AI in education. Experience had a strong and significant positive
effect on attitude towards MT (b*=.784), and also a strong and significant positive effect on attitude
towards AI (b*=.711), therefore H6 is supported.

4.4 Effect of age on attitude

This research hypothesized that there would be a significant difference in terms of attitudes
towards MT and AI in education when it comes to different age groups, namely that older people
would be more negative towards both. To test this, ANOVAs were run to compare the means of

30
attitude towards both topic, a separate ANOVA for each DV, between the different age groups. The
difference between age groups overall was highly significant (F(3)=123,31, p<.001), and when looking
at Scheffe’s test, almost every mean difference between the different age groups was significant. The
difference between 18-24 year olds (M=5.12, S.D= .689) and 25-39 year olds (M=4.98, S.D=.698) was
too small to be significant (p=.867), but the difference between 18-24 year olds and both 40-64 year
olds (p<.001) as 65+ year olds (p<.001) was highly significant , as was the difference between 25-39
year olds and both these age groups (p<.001 for both). The difference between 40-64 year olds
(M=3.34, S.D=.948) and 65+ year olds (M=2.65, S.D=.457) was also significant (p=.001). The drop-off
in attitude towards AI as age increases is both significant and interesting, as especially between the
age groups of 25-39, where the average is still predominantly positive about MT in education, and
40-64 year olds, where this has shifted to a predominantly negative view, this is especially apparent.
Furthermore, the standard deviation for the age group of 65+ is the lowest, meaning they are
without too many outliers considerably negative about MT in education.

Age groups Mean for MT Mdifference with Mdifference with Mdifference with Mdifference with
Attitude 18-24 25-39 40-64 65+
18-24 year 5.12 N/A .134 (p=.867) 1.779 2.464
(p<.001) (p<.001)
25-39 4.98 -.134 (p=.867) N/A 1.646 2.33 (p<.001)
(p<.001)
40-64 3.34 -1.779 -1.646 N/A .685 (p=.001)
(p<.001) (p<.001)
65+ 2.65 -2.464 -2.33 (p<.001) -.685 (p=.001) N/A
(p<.001)
Table 2: Mean and mean differences of attitude towards MT in education for age
groups

When looking at the ANOVA regarding the mean differences for age groups for attitude
towards AI in education, the overall results prove just as significant (F(3)=119,69, p<.001). The most
positive group towards AI in education is once again 18-24 year olds (M=5, S.D=.976), followed by 24-
39 year olds (M=4.63, S.D=.872), followed by 40-64 year olds (M=2.71, S.D=.927), followed by lastly
65+ year olds (M=2.02, S.D=.58). When looking at scheffe’s test, once again the only difference
between groups that is insignificant is between 18-24 year olds and 25-39 year olds (p=.294). Besides
this, all other differences between groups are significant. What is noticeable about the attitudes

31
towards AI between the different age groups is that the attitude for every age group is lower than
their same age group’s attitude towards MT in education, but that the differences get bigger as the
age gets older. For 18-24 year olds, the mean difference is .12, where this is .35 for 25-39 year olds,
.62 for 40-64 year olds and .63 for 65+ year olds. However, considering the strong differences in
attitudes, and also the significant and strong differences between even close age groups, H7 is
strongly supported.

Age groups Mean for AI Mdifference with Mdifference with Mdifference with Mdifference with
Attitude 18-24 25-39 40-64 65+
18-24 year 5 N/A .371 (p=.294) 2.29 (p<.001) 2.982
(p<.001)
25-39 4.63 -.371 (p=.294) N/A 1.646 2.33 (p<.001)
(p<.001)
40-64 2.71 -2.29 (p<.001) -1.646 N/A .692 (p=.011)
(p<.001)
65+ 2.02 -2.982 -2.33 (p<.001) -.692 (p=.011) N/A
(p<.001)
Table 3: Mean and mean differences of attitude towards AI in education for age
groups

4.5 effect of Schwartz’ human values

This research hypothesised that both Schwartz’ human universal values of openness to
change and conservatism would affect both one’s attitude towards MT and AI in education. To test
this, a simple linear regression was run with openness to change as the independent variable, and
attitude towards MT and AI as the dependent variable. In both cases, homoscedasticity was
established and the constant error variance was not violated (despite some minor deviations in the
middle for both). For attitude towards MT in education, openness to change proved a significant
factor (F(1,163)=279,70, p<.001), having an adjusted R2 of .63, and being a strong, positive effect
(b*=.795, p<.001). For attitude towards AI, openness to change also proved to be a significant factor
(F(1,163)=234,02, p<.001), having a adjusted R2 of .59, and also being a strong and positive effect
(b*=.768, p<.001). Considering openness to change has a strong positive effect on both attitude
towards MT as AI in education, H8 is supported.

32
` A simple linear regression was run to test whether conservatism was a significant predictor
for attitude towards MT and AI in education. Conservatism was the independent variable and
attitude towards MT and AI in education once again as the dependent variable. For both cases,
homoskedasticity was once again established, and the normal error variance was mostly held,
despite some small deviation at the tail for attitude towards MT, and some small deviation at the
middle for attitude towards AI. Conservatism too, was a significant factor for both attitude towards
MT (F(1,163)=188,55, p<.001) as attitude towards AI (F(1,163)=244,36, p<.001). For attitude towards
MT in education, conservatism explained 53.3% of the variance in the date (R2=.533), where this was
59.7% for attitude towards AI (R2=.597). Conservatism proved to be, for both attitude towards MT in
education (b*=-.732, p<.001) as attitude towards AI in education (b*=-.775, p<.001), a strong
negative, significant, effect. Because of this clear data, H9 is supported.

4.6 Strength of predictors

Considering all hypothesised predictors proved significant in affecting attitude towards MT

and AI, additional tests were run to check which predictors were the strongest and explained for
most variance in the data. To test the strongest predictor for attitudes towards MT in education, a
multiple regression analysis was conducted with all continuous variables, being openness to change,
conservatism, mobile self-efficacy, experience, PU and PEOU as independent variables and attitude
towards MT as the dependent variable. The effect of all predictors overall was significant
(F(6,158)=171,87, p<.001), and this model explained more variance in the data than any single
predictor when run in analysis (R2=.862, where the highest for a single predictor was R2=.821 for PU).
This model also explained more of the variance in the data than the model containing only PU and
PEOU of MT as independent variables (R2=.847) However, when looking at the individual effect of the
predictors in this overall model, only 3 predictors still proved significant. PU still had a moderate
positive effect (b*=.573, p<.001), PEOU was still a weak positive effect (b*=.235, p<.001), and
conservatism was still a weak negative effect (b*=-.121, p=.026). The effect of openness to change,
mobile self-efficacy and experience proved non-significant in this model, and PU proved to be the
strongest predictor for attitude towards MT, which is also was on its own when run as a simple linear
regression. The comparison of the strengths of the predictors, between being used as a single
predictor versus in this wider model, can be found in table 4. Different multiple regression analyses
were run with combinations of the variables in order to try and find a model that would explain more
variance in the data, but no other model explained more variance in the data than the model
containing all these predictors. The loss of significance for some of the predictors that were
associated with attitude towards MT on their own might lead to think they are not significant at all.
However, considering that all predictors had significant and strong associations with attitude towards

33
 MT as single predictors, and the fact that the deletion of these certain predictors in the bigger model
did not lead to more variance explained in the data, no hypothesises are adjusted.

Independent Variance Variance Variance b* as single b* as predictor b* difference

variables explained explained explained predictor in wider model between
as single in the difference single and
predictor wider between single wider model
model and wider model
PU of MT .821 .862 .041 .907, p<.001 .573, p<.001 .334
PEOU of MT .727 .862 .135 .854, p<.001 .235, p<.001 .619
Mobile self- .397 .862 .465 .633, p<.001 .074, p=.065 .559
efficacy
MT experience .613 .862 .249 .784, p<.001 .076, p=.193 .708

Openness to .630 .862 .232 .795, p<.001 -.062, p=.316 .857

change
Conservatism .533 .867 .329 -.732, p<.001 -.121, p=.026 .611
Table 4: Variance explained and effect of independent variables on attitude towards
MT as single predictors compared to the wider model

The same multiple regression analysis was run for attitude towards AI in education, now with
attitude towards AI as the dependent variable and the same predictors (except PU, PEOU and
experience of/with MT switched for PU, PEOU and experience of/witch AI) as independent variables.
Once again, the results overall proved significant (F(6,158)=177,58, p<.001) and the variance
explained for this model was once again higher than any single linear regression (R2=.866 compared
to R2=.821 for PU). Precisely similar to attitude towards MT, the same three predictors proved
significant in this wider model. The strongest was once again PU (b*=.548, p<.001), followed by
PEOU (b*=.220, p<.001) and lastly conservatism (b*=-.197, p<.001). Mobile self-efficacy was still
weakly significant (b*=.074, p=.061), but experience and openness to change proved to be non-
significant. PU of AI had a moderate positive association with attitude towards AI, while PEOU of AI
had a weak positive association with attitude towards AI, and conservatism had a weak negative
association with attitude towards AI. Once again, PU proved to be the strongest predictor, albeit

34
 slightly less strong than it was for MT. The comparison of the strengths of the predictors, between
being used as a single predictor versus in this wider model, can be found in table 5.

Independent Variance Variance Variance b* as single b* as predictor b* difference

variables explained explained explained predictor in wider model between
as single in the difference single and
predictor wider between single wider model
model and wider model
PU of AI .818 .866 .048 .905, p<.001 .548, p<.001 .357
PEOU of AI .617 .866 .249 .787, p<.001 .220, p<.001 .567
Mobile self- .416 .866 .450 .648, p<.001 .074, p=.061 .574
efficacy
AI experience .502 .866 .364 .711, p<.001 .070, p=.123 .641

Openness to .590 .866 .276 .768, p<.001 -.062, p=.253 .830

change
Conservatism .597 .866 .269 -.775, p<.001 -.197, p=.026 .578
Table 4: Variance explained and effect of independent variables on attitude towards
AI as single predictors compared to the wider model

While PU seemed to be the strongest of the continuous variable, the most important
predictor for attitude towards MT and AI, and also for the other predictors in this research, is
possibly age. Chapter 4.4 already showcased how age was of high significance for someone’s attitude
towards MT and AI, but age had an effect on literally all predictors too. The effect of age was
significant for openness to change (F(3,161)=110,75, p<.001), conservatism (F(3,161)=79,19, p<.001),
mobile self-efficacy (F(3,161)=39,43, p<.001), experience with MT (F(3,161)=164,96, p<.001),
experience with AI (F(3,161)=117,32, p<.001), PU of MT (F(3,161)=136,94, p<.001), PU of AI
(F(3,161)=130,21, p<.001), PEOU of MT (F(3,161)=92,65, p<.001) and PEOU of AI (F(3,161)=45,08,
p<.001). There is a consistent significant drop-off in means between 25-39 year olds and 40-64 year
olds, and the mean differences between 18-24 year olds and 65+ year olds are high in every category.
When compared to 65+ year olds, 18-24 year olds have higher means for every predictor, most
noticeably experience with MT (Mdifference of 2.58 over a scale of 5, p<.001), PU of MT (Mdifference of
2.27 over a scale of 5, p<.001) and openness to change (Mdifference of 2.35 over a scale of 7, p<.001).
Only for conservatism do 65+ year olds score (significantly) higher compared to 18-24 year olds

35
Mdifference of 2.07 over a scale of 7, p<.001), but considering conservatism was a negative effect on
attitude towards MT and AI in education, this follows the trend seen in this research. The effect of
age is seen throughout the entire research, and might be the most important predictor for attitude
towards MT and AI overall.

5) Conclusion

5.1 Summary of findings

This research tried to answer the overall question of to what extent people are positive about the
integration of media technologies (MT) and artificial intelligence (AI) into education, what elements
would be in predicting this attitude and what the most important factors in predicting these attitudes
would be, in order to establish an extended framework on innovation adaptation and the
encompassing challenges that brings. This research hypothesised, based on a balanced overview of
the positives and negatives that MT and AI bring to education, that people would be predominantly
positive towards both MT and AI in education. Tests proved that people overall were neither very
positive nor negative towards MT and AI in education, although the standard deviation for both
attitudes were quite high. Younger people were predominantly positive towards both MT and AI in
education, but older adults proved to be quite negative towards both. This research furthermore
hypothesises, based on the diffusion of innovation theory, that people would be more positive
towards MT in education than they would be towards AI in education, considering MT are already
integrated into education to a considerable amount, where AI is not yet. Respondents were slightly
more positive about MT in education than they were about AI in education, however the difference
was not monumental.

This research also hypothesised, based on the technology acceptance model (TAM) the
unified theory of technology acceptance and usage (UTAUT), and previous related research, that a
higher amount perceived usefulness (PU) and perceived ease of use (PEOU) would both lead to a
more positive attitude towards both MT and AI in education. Perceived usefulness had a significant,
strong and positive association with both attitude towards MT as attitude towards AI, meaning that a
higher amount of perceived usefulness lead to a more positive attitude towards MT and AI in
education. Perceived ease of use also had significant, strong, positive association with both attitude
towards MT and AI in education. The association between perceived ease of use and attitude
towards AI in education was tentative, as the constant error variance was not held. However, the

36
strong, positive association and the high explained amount of variance in the model lead to the
assumption that the result is still valid.

This research furthermore hypothesised that the amount of mobile self-efficacy one has, and
the amount of experience one has had with MT and AI in education respectively, would also have a
positive association with attitude towards MT and AI in education, based on previous research, and
experience being an external factor in the TAM and UTAUT. Mobile self-efficacy proved to have a
strong, positive, significant association with both attitude towards MT and AI in education, meaning
that an increase in mobile self-efficacy would lead to a more positive attitude towards MT and AI in
education. The association mobile self-efficacy had with attitude towards AI in education was slightly
higher than the association it had with attitude towards MT in education, but it was not far apart.
Experience with MT and AI in education also proved a significant factor in explaining attitude towards
both MT as AI. Experience once again had a strong and positive association with both attitude
towards MT and AI, meaning an increase in experience lead to a more positive attitude towards MT
and AI in education. The association between experience and attitude towards MT was higher, but
the overall amount of experience with AI amongst the respondents was a lot lower.

This research also hypothesised that amongst demographic factors, age would be an
important factor in attitude towards MT and AI in education, based on previous research, the fact
that older adults have less experience and competence with MT and are more conservative. Age
proved to be a highly significant factor in explaining attitude towards MT and AI, as there was a
sizeable drop off in attitude between the age groups of 25-39 and 40-64. Also, the differences in
attitude between the youngest age group (18-24) and the oldest age group (65+) was surprisingly
high for both attitude towards MT and AI, leading to believe that age is a very important factor in
attitude towards new technologies and innovations. Age was not only a factor in attitudes towards
MT and AI however, as age influenced nearly all other predictors as well. As age increased,
conservatism increased, openness to change decreased, experience decreased, mobile self-efficacy
decreased, PU decreased and PEOU decreased, meaning that age influenced all significant predictors
for attitude towards MT and AI.

Ultimately, this thesis hypothesised that two of Schwartz’ human values, namely openness to
change and conservatism, would affect one’s attitude towards MT and AI in education, considering
the relation between both openness to change and conservatism, and innovation adaptation has
been discovered in previous research. The association that openness to change had with both
attitude towards MT and AI was significant. The association was strong and positive, meaning that
more openness to change led to a more positive attitude towards MT and AI in education. The

37
association that conservatism had with both attitudes towards MT and AI in education also
significant. However, conservatism had a strong negative association, meaning that the more
conservative a person was, the more negative that person’s attitude was towards MT and AI in
education.

Lastly, while no hypothesis was specifically made for this, some extra tests were run to test
which predictor proved the most influential, considering nearly all predictors which were
hypothesised to, actually had a strong association with both attitude towards MT and AI. When run
in a bigger model, results were slightly different. For attitude towards MT, PU was still the strongest
predictor, followed by PEOU and after that conservatism. Experience, mobile self-efficacy and
openness to change proved an insignificant association with attitude towards MT in education, in the
wider model. The significant predictors in the wider model however were moderate and weak
compared to strong when they were measured separately, however they did account for more
variance in the data than any predictor did on its own for attitude towards AI in education, the same
model was run, and the same predictors proved significant and insignificant as in the model with
attitude towards MT. The strongest predictor in the model for attitude towards AI was once again
PU, followed by PEOU and lastly conservatism. Experience, mobile self-efficacy and openness to
change were once again non-significant. The significant predictors in the wider model for attitude
towards AI also changed from strong to moderate and weak associations however, yet once again
the predictors combined accounted for more variance in the data than any one of the predictors
measured separately. While the insignificance of some predictors in the wider model could lead to
believe that they are overall insignificant, and that the hypothesis for these predictors should be
adjusted, these predictors were still significant on their own, and including them in the wider model
did lead to a higher amount in variance explained. This means that while they are not they key
predictors, they still have associations with attitude. All the independent variables are interrelated,
and while the wider model might make it look like some of them become insignificant, they still do
have an effect, but the strength of the key predictors that were discovered might masque this.
Therefore, the hypothesises related to these predictors that were accepted, remain accepted.

5.2 Theoretical implications

A number of findings from this research can be put in a wider theoretical perspective. The
overall answer to the main question in this research, namely to what extent people would be positive
towards the integration of MT and AI in education, was that people overall are somewhat neutral
towards it, so neither very positive or negative. However, this is where the data might mislead us, as
when looking at different age groups, none of the groups are neutral towards either MT or AI.

38
Because older groups are predominantly negative towards MT and AI, and younger groups
predominantly positive, the overall score somewhat evens out in the middle, leading to believe the
overall population is neutral, but this would not be exactly correct. So what would be theoretically
interesting is to figure out why older adults are more negatively opinionated towards MT and AI in
education, besides the effects found in this research. What is interesting is that the positives about
MT and AI in education are very factual and objective, for example the increased efficiency and
accuracy that both MT and AI bring to education (Murphy, 2016; Schacter & Jo, 2017; Kuma, 2019;
Southworth, 2019), yet older groups still perceived both as not very useful. The negatives that
surround MT and AI in education are all more subjective and personal, such as the distracting factor
they bring and the possible dehumanization of education (Haslam, 2006; Alhumaid, 2019), yet
younger groups agreed with this more than that older groups agreed with MT and AI being effective
and accurate. The reasoning behind these polarizing views should be explored further, perhaps with
qualitative research to fully try and grasp exactly why younger groups think positively about this, and
why older groups do not.

The strong association both perceived usefulness and perceived ease of use had with both
attitudes towards MT and AI was hypothesised, and possibly should be internalised in attitudes
themselves. PU being the strongest predictor, both as a single predictor as the most influential one in
the wider model, backs up what the UTAUT already stated, namely that PU is the most consistently
heavy influencer of attitude (Venkatesh et al., 2003). The same goes for PEOU, which was the second
strongest predictor both on its own as in the model. However, to truly advance a present and future-
proof framework of innovation adaptation, it might prove fruitful to invest other important
factors/predictors for attitude, considering it is arguable that PU and PEOU are part of one’s attitude.
Since attitude is defined as the beliefs, perceptions and emotions someone has about a
thing/person/object/etc., it is arguable that to what extent someone thinks an innovation is useful
and easy to use, is already part of their overall attitude towards that innovation. Following that logic,
it is arguable, not that it is useless to investigate PU and PEOU, but that other factors which might be
more easily traceable and influenceable are looked for, in order to help adaptation innovation and
establish a more comprehensible framework. In any case, attention given to PU and PEOU in order to
promote innovation adaptation is clear, considering that the group that was the most negative and
had the least amount of PU and PEOU, is the group that has the least experience, competence, and
digital literacy skills related to MT and AI. Furthermore, older people are overall less knowledgeable
about MT and AI, and fairly often do not know what it is (Mozilla, 2019; Pega, 2020), where young
people are now growing up around it and get this knowledge in early on in their life (Axente, 2018;
Hao, 2019) .Education and training in dealing with these topics should be favourable to the whole of

39
society, as previous research does show that this works, and it seems unavoidable that MT and AI are
losing their place in our society anytime soon.

Lastly, the influence age had in this research can not be dismissed, and should be further
explored theoretically or taken action upon in real life. This research not only showed that age is an
important factor in attitudes towards MT and AI in education, but also a factor in conservatism,
openness to change, mobile self-efficacy, experience, where previous research has already shown
age to be a factor in digital literacy and new media competence. Age might be the most important
factor in a innovation adaptation framework, while current frameworks solely see it as an external
factor. There are already examples out there where age has shown to be a negative factor in taking
action or adapting innovation. Take for example the problem of climate change, which is arguably the
result of human action, despite the general scientific consensus that it is. Older generations believe
less in climate change (Reinhart, 2018), they are less concerned about it (Ballew, et al., 2019), they
believe it will not affect them as much (Haq, Brown, & Hards, 2010) and they are less willing to take
action against it (Pew Research Center, 2020), despite claims that it is these generations that are
most to blame for climate change, as they have for example emitted the most carbon emissions
(Nagourney, 2013; Loria, 2018). This trend of older generations that are somewhat negative about
things they perhaps do not entirely understand, both seen in this research, many other research and
the climate change example, is worrying considering that not only the average age in the world is
increasing (Ritchie & Roser, 2019), but that most politicians and political leaders, especially those in
the G20, are relatively old, often significantly older than the average age of the population (The
Learning Network, 2020; Asrar, 2021). If a wider comprehensive model for innovation adaptation is
to be devised, age might be the single most important factor in this model, and it would likely take
education and experience, or a bigger influence of younger generations, to minimise the effect age
has on the adaptation of innovation.

5.3 Limitations and future research

This research had some considerable limitations, which keep a part of the findings from being
as generally applicable and generalizable as they could have been, and as the goal was. Firstly, due to
the costs and time constraints, and also due to the exceptional circumstances regarding the covid-19
pandemic, the sample was not as representative as it could have been. Non-probability snowball
sampling, as was the case in the first stage of the sampling for this research, is not ideal, considering
that the respondents will be somewhat close to the researcher, meaning they are often coming from
the same sphere of education, culture and preferences. The second stage of the sampling also could
have been subjected to some improvements. Due to the personal influence on the first stage of the

40
sampling, predominantly younger people and certain nationalities were represented more than
others, meaning that the second stage of the sampling existed more to add to an otherwise
unrepresentative sample. However, considering age was a big factor in this research, having older
respondents who are probably more competent with MT than the average of that age, considering
they would have had to be proficient in navigating an account and website where surveys need to be
filled in, might have skewed this part of the sample to more experienced MT users, and therefore
more positive about MT and AI. Perhaps the effect of age would have been bigger if older
respondents were garnered naturally as opposed to paid respondents. Another issue this research
has is the problem of causality. While never explicitly stated in this research, it is implicit that most
predictors have a causal relationship with attitude towards MT and AI, meaning that if for example
openness to change would increase, then an increase in attitude towards MT and AI would follow.
While the hypothesises were guided by research and existing models, it is possible the hypothesises
would go in the opposite direction and still have an effect. For example, someone’s lower amount of
mobile self-efficacy and experience with MT and AI might be explained by one’s attitude towards MT
and AI being low, thus feeling no need to use it. Lastly, this research is hindered possibly by its
oversimplification of predictors. While purposefully done to establish some key predictors for
attitude towards MT and AI, and in extension for a framework for innovation adaptation, in practice
it probably is not as simple. In reality, a lot of these predictors are interrelated, moderate each other
and all have effect on each other and ultimately attitude. While PU and age seemed the biggest
predictors in this research, they might be moderated by, or moderate themselves, other predictors
who then could be seen as the most important. While simplicity should be key, reality often is not
that simple, and pointing to a few predictors in what is in essence a highly complicated societal
process might be underestimating a lot of other predictors.

More research on this topic and its predictors is highly desirable. While some predictors, such
as PU and PEOU, are already firmly established as predictors for attitude in previous research,
research on less researched predictors and perhaps predominantly easily checkable socio-
demographic factors would further a comprehensible model for innovation adaptation on both a
personal level as a societal level. Ideally, research should continue for, for example, the UTAUT, in
identifying more complex latent predictors, as to fully understand which predictors all influence
attitude and if possible also see which predictors moderate which. However, research and
development of a easily comprehensible model is also desirable, mostly with easily checkable socio-
demographic factors as to quickly be able to assess a country’s/city’s/region’s willingness for
innovation adaptation. A model like this would make it possible for decision/policy makers to quickly
assess where energy and resources need to go in accepting innovation. If for example a country then

41
has a high median age, more energy and resources could be allocated towards schooling and
education of the public in order to make sure everyone is informed, experienced and competent with
the innovation, thus easing the adaptation process and preventing a situation now where
predominantly older people call for, for example, MT to be banned from classrooms. Lastly,
considering that AI was influenced easier by certain predictors, further research on knowledge and
attitude seems desirable. It would be fruitful to research how knowledge first of all can be increased
quickly and efficiently, and also what predictors and how much attitude are affected by knowledge.
Innovation is inherent to mankind, and understanding how we deal with is as not only an individual
but as a society is extremely relevant, and should be of high importance.

References

Aaron, L. S., & Lipton, T. (2018). Digital distraction: Shedding light on the 21st-century college
classroom. Journal of Educational Technology Systems, 46(3), 363-378.

Al-Emran, M., & Granic, A. (2021). Is it still valid or outdated? A Bibliometric analysis of the
technology acceptance model and its applications from 2010 to 2020. Recent Advances in
Technology Acceptance Models and Theories, 1-12.

Álvarez, J. S, Pedrosa, I., Lozano, L. M., Cueto, E. G., Izquierdo, M. C., & Fernández, J. M. (2018). Using
reversed items in Likert scales: A questionable practice. Psicothema, 30.

Anderson, M., & Jiang, J. (2018, May 31). Teens, Social Media & Technology 2018. Retrieved from
Pew Research Center: https://www.pewinternet.org/2018/05/31/teens-social-media-
technology-2018/

Asrar, S. (2021, March 1). Can you guess the age of a leader based on average age of country's
population. Retrieved from The Times of India:
https://timesofindia.indiatimes.com/world/rest-of-world/can-you-guess-the-age-of-a-leader-
based-on-the-average-age-of-countrys-people/articleshow/81275841.cms

Axente, M. (2018, May 11). Who decides the future of artificial intelligence? Young people (if we
support them). Retrieved from ITU News: https://news.itu.int/who-decides-the-future-of-
artificial-intelligence/

Aydeniz, M., & Southerland, S. A. (2012). A national survey of middle and high school science
teachers’ Responses to Standardized Testing: Is Science Being Devalued in Schools? Journal
of Science Teacher Education, 23(3), 233-257.

42
Bagozzi, R. P. (2007). The Legacy of the Technology Acceptance Model and a Proposal for a Paradigm
Shift. Journal of the Association for Information Systems, 8(4), 244-254.

Ballew, M., Marlon, J., Rosenthal, S., Gustafson, A., Kotcher, J., Maibach, E., & Leiserowitz, A. (2019,
June 11). Do younger generations care more about global warming? Retrieved from Yale
Climate Change Communication: https://climatecommunication.yale.edu/publications/do-
younger-generations-care-more-about-global-warming/

Bandura, A (1977). Self-efficacy: Toward a Unifying Theory of Behavioral Change. Psychological

Review, 84(2): 191–215. doi:10.1037/0033-295x.84.2.191

Bandura, A. (1986). Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood
Cliffs, NJ: Prentice-Hall.

Bandura, A. (2010). Self‐efficacy. The Corsini encyclopedia of psychology, 1-3.

doi:10.1002/9780470479216.corpsy0836

Benbasat, I., & Barki, H. (2007). Quo Vadis TAM? Journal of the Association for Information Systems,
8(4), 211-218. Available at http://aisel.aisnet.org/jais/vol8/iss4/12

Berry, J. W., Poortinga, Y. H., Pandey, J., Segall, M. H., & Kâğıtçıbaşı, Ç. (1997). Handbook of cross-
cultural psychology: Theory and method (Vol. 1). John Berry.

Bhardwaj, A. (2016). Importance of education in human life: A holistic approach. International

Journal of Science and Consciousness, 2(2), 23-28.

Bilsky, W., Janik, M., & Schwartz, S. H. (2011). The structural organization of human values-evidence
from three rounds of the European Social Survey (ESS). Journal of Cross-Cultural Psychology,
42(5), 759-776.

Broady, T., Chan, A., & Caputi, P. (2010). Comparison of older and younger adults' attitudes towards
and abilities with computers: Implications for training and learning. British Journal of
Educational Technology, 41(3), 473-485

Bryant, J., Heitz, C., Sanghvi, S., & Wagle, D. (2020, January 14). How artificial intelligence will impact
K-12 teachers. Retrieved from McKinsey & Company:
https://www.mckinsey.com/industries/public-and-social-sector/our-insights/how-artificial-
intelligence-will-impact-k-12-teachers

Bucholz, K. (2019, January 15). Americans Fear the AI Apocalypse. Retrieved from Statista:
https://www.statista.com/chart/16623/attitudes-of-americans-towards-ai/

43
Buckingham, S. J., & Luckin, R. (2019). Learning analytics and AI: Politics, pedagogy and practices.
British Journal of Educational Technoloy 50(6), 2785-2793. doi:10.1111/bjet.12880

Busteed, B. (2020, September 25). Wake Up Higher Education. The Degree Is On The Decline.
Retrieved from Forbes: https://www.forbes.com/sites/brandonbusteed/2020/09/25/wake-
up-higher-education-the-degree-is-on-the-decline/?sh=30e530987ecb

Cavanagh, S. (2015, September 24). Students’ Tablet, Smartphone Usage Climbs, With Strong
Appetite for Apps. Retrieved from Market Brief:
https://marketbrief.edweek.org/marketplace-k-
12/student_tablet_smartphone_usage_climbs_with_strong_appetite_for_apps/#:~:text=Acc
ording%20to%20the%20poll%2C%2078,58%20percent%20the%20year%20before.

Chade, M. (2021, March 18). The evolution of interactive whiteboards – from the wall into your hands
(or both!). Retrieved from Explain Everything: https://explaineverything.com/the-evolution-
of-interactive-whiteboards-from-the-wall-into-your-hands-2/

Chakraborty, P., Mittal, P., Gupta, M. S., Yadav, S., & Arora, A. Opinion of students on online
education during the COVID‐19 pandemic. Human Behavior and Emerging Technologies.
DOI: 10.1002/hbe2.240

Chao, C. M. (2019). Factors determining the behavioral intention to use mobile learning: An
application and extension of the UTAUT model. Frontiers in psychology, 10, 1652.

Cheung, A. (2021). Language teaching during a pandemic: A case study of zoom use by a secondary
ESL teacher in Hong Kong. RELC Journal, 1–16. doi: 10.1177/0033688220981784

Chuttur, M. Y. (2009). Overview of the Technology Acceptance Model: Origins, Developments and
Future Directions. Sprouts, 9(37). Available at http://sprouts.aisnet.org/9-37

Cimperman, M., Brenčič, M. M., and Trkman, P. (2016). Analyzing older users’ home telehealth
services acceptance behavior—applying an extended UTAUT model. Int. J. Med. Inform. 90,
22–31. doi: 10.1016/j.ijmedinf.2016.03.002

Clark, R. E. (1983). Reconsidering research on learning from media. Review of educational research,
53(4), 445-459.

Clark, R. E. (1994). Media will never influence learning. Educational Technology Research and
Development, 42(2), 21-29.

Clark, R. E. (Ed.). (2001). Learning from media: Arguments, analysis, and evidence. IAP.

44
Clark, D. (2019, August 27). Share of population with knowledge of English in non-native European
countries as of March 2019. Retrieved from Statista:
https://www.statista.com/statistics/990547/countries-in-europe-for-english/

Criddle, C. (2021, April 7). Mobile phones should be banned in schools - Gavin Williamson. Retrieved
from BBC : https://www.bbc.com/news/technology-56663010

Dahl, M. (2017, June 14). People Will Like You More If You Ask Them Questions. Retrieved from the
Cut: https://www.thecut.com/2017/06/people-will-like-you-more-if-you-ask-them-
questions.html

Davis, F. D. (1986). A Technology Acceptance Model for Empirically Testing New End-User Information
Systems: Theory and Results. Cambridge, MA: MIT Sloan School of Management

Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information
technology. MIS quarterly, 319-340.

De Blij, H. J., Muller, P. O., & Nijman, J. (2017). Geography: Realms, Regions, and Concepts.
(17th edition). Amsterdam: John Wiley & Sons, Incorporated.

Delgado, P. (2018, November 9). The problem of standardized tests. Retrieved from Tecnológico de
Monterrey: https://observatory.tec.mx/edu-news/the-problem-of-standardized-tests

Dukadinovska, M. (2020, August 12). 7 Ways AI Is Changing The Education Industry. Retrieved from
IdeaMotive: https://www.ideamotive.co/blog/ways-ai-is-changing-the-education-
industry#:~:text=According%20to%20Market%20Research%20Engine,annual%20growth%20r
ate%20of%2045%25.

Domine, V. (2009). A social history of media, technology and schooling. Journal of Media Literacy
Education, 1(1), 4.

Douglas, D., Angel, H., & Bethany, W. (2012). Digital devices, distraction, and student performance:
Does in-class cell phone use reduce learning? Astronomy education review. doi:
10.3847/AER2012011

Edison, S. W., & Geissler, G. L. (2003). Measuring attitudes towards general technology: Antecedents,
hypotheses and scale development. Journal of targeting, Measurement and Analysis for
Marketing, 12(2), 137-156.

45
Ellis, P. (2019, March 19). Changing the world through education – how Nelson Mandela created the
conditions for success. Retrieved from Cambridge International:
https://blog.cambridgeinternational.org/nelson-mandela/

Faggella, D. (2019, November 21). Examples of Artificial Intelligence in Education. Retrieved from
Emerj: https://emerj.com/ai-sector-overviews/examples-of-artificial-intelligence-in-
education/

Faggella, D. (2020, April 11). Everyday Examples of Artificial Intelligence and Machine Learning.
Retrieved from Emerj: https://emerj.com/ai-sector-overviews/everyday-examples-of-ai/

Gadanidis, G. (2017). Artificial intelligence, computational thinking, and mathematics education. The
International Journal of Information and Learning Technology.Goundar, S. (2014). The
distraction of technology in the classroom. Journal of Education & Human Development, 3(1),
211-229.

Gall, R. (2018, March 15). 5 polarizing Quotes from Professor Stephen Hawking on artificial
intelligence. Retrieved from Packt: https://hub.packtpub.com/stephen-hawking-artificial-
intelligence-quotes/

Ghafourifar, A. (2017, March 12). Why parents might not be ready for AI in the classroom. Retrieved
from VentureBeat: https://venturebeat.com/2017/03/12/why-parents-might-not-be-ready-
for-ai-in-the-classroom/

Gupta, J. (2020, September 21). 7 Real-Life Examples of AI in Education. Retrieved from Wire19:
https://wire19.com/real-life-examples-of-ai-in-education/

Hamidi, F., Ghorbandordinejad, F., Rezaee, M., & Jafari, M. (2011). A comparison of the use of
educational technology in the developed/developing countries. Procedia Computer Science,
3, 374-377.

Hao, K. (2019, September 13). Kids are surrounded by AI. They should know how it works. Retrieved
from MIT technology Review: https://www.technologyreview.com/2019/09/13/133056/kids-
are-surrounded-by-ai-they-should-know-how-it-works/Kids are surrounded by AI. They
should know how it works.

Haselhuhn, M. P., Pope, D. G., Schweitzer, M. E., & Fishman, P. (2012). The impact of personal
experience on behavior: Evidence from video-rental fines. Management Science, 58(1), 52-
61.

46
Haßler, B., Major, L., & Hennessy, S. (2016). Tablet use in schools: A critical review of the evidence for
learning outcomes. Journal of Computer Assisted Learning, 32(2), 139-156.

Hart, S. A., & Laher, S. (2015). Perceived usefulness and culture as predictors of teachers attitudes
towards educational technology in South Africa. South African Journal of Education, 35(4).

Hartmann, T., Wirth, W., Schramm, H., Klimmt, C., Vorderer, P., Gysbers, A., Böcking, S., Ravaja, N.,
Laarni, J., Saari, T., Gouveia, F. & Sacau, A. M. (2015). The spatial presence experience scale
(SPES). Journal of Media Psychology. Vol. 28(1):1–15. DOI: 10.1027/1864-1105/a000137

Haq, G., Brown, D., & Hards, S. (2010). Older People and Climate Change: the Case for Better
Engagement. Stockholm: Stockholm Environment Institute.

Hazelrigg, N. (2019, July 10). Survey: Nearly Half of Students Distracted by Technology. Retrieved
from Inside Higher Ed: https://www.insidehighered.com/digital-
learning/article/2019/07/10/survey-shows-nearly-half-students-distracted-
technology#:~:text=A%20recent%20survey%20found%20the,in%20the%20classroom%20is%
20unavoidable.&text=Most%20students%20surveyed%20said%20they,use

Hess, A. J. (2019, January 19). Research continually shows how distracting cell phones are—so some
schools want to ban them. Retrieved from Cnbc:
https://www.cnbc.com/2019/01/18/research-shows-that-cell-phones-distract-students--so-
france-banned-them-in-school--.html

Holden, J. T., & Westfall, P. J. L. (2007). An instructional media selection guide for distance learning.
Online Submission.

Hoque, R., and Sorwar, G. (2017). Understanding factors influencing the adoption of mHealth by the
elderly: an extension of the UTAUT model. Int. J. Med. Inform. 101, 75–84. doi:
10.1016/j.ijmedinf.2017.02.002

Hughes, J. (2021, January 13). What is the Future of Education Technology? Retrieved from KeyStone
Online Studies: https://www.onlinestudies.com/article/what-is-the-future-of-education-
technology/

Ibáñez, M. B., Di Serio, Á., Villarán, D., & Kloos, C. D. (2014). Experimenting with electromagnetism
using augmented reality: Impact on flow student experience and educational effectiveness.
Computers & Education, 71, 1-13

IDRC. (2020, September 20). AI Readiness Index 2020. Retrieved from Oxford Insights:
https://www.oxfordinsights.com/government-ai-readiness-index-2020

47
Inspectie van het Onderwijs (2018). De staat van het onderwijs. Onderwijsverslag 2016/2017.
Utrecht: Inspectie van het Onderwijs

IPRAN. (2020 , February 17). 43% of the world’s EdTech companies are based in the US, but Sweden
leads the way on VC Funding. Retrieved from Uk Tech News:
https://uktechnews.co.uk/2020/02/17/43-of-the-worlds-edtech-companies-are-based-in-
the-us-but-sweden-leads-the-way-on-vc-funding/

Ireland, S. (2020, May 10). Ranked: World’s Best Countries For Education System, 2020. Retrieved
from CeoWorld Magazine: https://ceoworld.biz/2020/05/10/ranked-worlds-best-countries-
for-education-system-2020/

Isomursu, M., Ervasti, M., Kinnula, M., & Isomursu, P. (2011). Understanding human values in
adopting new technology—A case study and methodological discussion. International journal
of human-computer studies, 69(4), 183-200.

Jackson, L. D. (2012). Is Mobile Technology in the Classroom a Helpful Tool or a Distraction?: A Report
of University Students' Attitudes, Usage Practices, and Suggestions for Policies. International
Journal of Technology, Knowledge & Society, 8(5).

Jonassen, D. H. (1996). Learning with technology: Using computers as cognitive tools. Handbook of
research for educational communications and technology.

Jones, T. L., Baxter, M., & Khanduja, V. (2013). A quick guide to survey research. The Annals of The
Royal College of Surgeons of England, 95(1), 5-7.

Joo, Y. J., Park, S., & Lim, E. (2018). Factors influencing preservice teachers’ intention to use
technology: TPACK, teacher self-efficacy, and technology acceptance model. Journal of
Educational Technology & Society, 21(3), 48-59.

Keith, M. J., Babb Jr, J. S., Furner, C. P., & Abdullat, A. (2011). The role of mobile self-efficacy in the
adoption of location-based applications: An iPhone experiment. Hawaii International
Conference on System Sciences (44), 1-10

Kelly, J. (2020, October 27). U.S. Lost Over 60 Million Jobs—Now Robots, Tech And Artificial
Intelligence Will Take Millions More. Retrieved from Forbes:
https://www.forbes.com/sites/jackkelly/2020/10/27/us-lost-over-60-million-jobs-now-
robots-tech-and-artificial-intelligence-will-take-millions-more/?sh=8a4be481a525

Kemp, B. R., & Montez, J. K. (2020). Why does the importance of education for health differ across
the United States?. Socius, 6, 2378023119899545.

48
Kemp, J. (2008, February 26). Driving Children from Distraction. Retrieved from The Guardian:
https://www.theguardian.com/education/2008/feb/26/teaching.schools1

Khalilzadeh, J., Ozturk, A. B., and Bilgihan, A. (2017). Security-related factors in extended UTAUT
model for NFC based mobile payment in the restaurant industry. Comput. Hum. Behav. 70,
460–474. doi: 10.1016/j.chb.2017.01.001

Khoshnevisan, B. (2019). To integrate media and technology into language education: For and
against. Advances in Global Education and Research: Volume 3, 85-92

Kim, S., & Choudhury, A. (2020). Comparison of Older and Younger Adults’ Attitudes Toward the
Adoption and Use of Activity Trackers. JMIR mHealth and uHealth, 8(10), e18312.

Kozma, R. B. (1991). Learning with media. Review of educational research, 61(2), 179-211.

Kulkarni, A. (2019, September 6). AI in Education: Where is It Now and What is the Future? Retrieved
from Lexalytics: https://www.lexalytics.com/lexablog/ai-in-education-present-future-
ethics#:~:text=AI%20in%20education%20generally%20focuses,on%20each%20student's%20s
pecific%20needs.&text=Many%20EdTech%20companies%2C%20including%20Squirrel,Knowl
edge%20Space%20Theory%20

Kumar, S. (2019, November 25). Advantages and Disadvantages of Artificial Intelligence. Retrieved
from Towards Data Science: https://towardsdatascience.com/advantages-and-
disadvantages-of-artificial-intelligence-182a5ef6588c

Kurz, L. (2019, March 19). Devices in the Classroom: To Ban or Not to Ban? Retrieved from Indiana
Bloomington University: https://blogs.iu.edu/citl/2019/03/19/ban-devices/#.YLkGDahKhPY

Kwon, K., Ottenbreit-Leftwich, A. T., Sari, A. R., Khlaif, Z., Zhu, M., Nadir, H., & Gok, F. (2019).
Teachers’ self-efficacy matters: Exploring the integration of mobile computing device in middle
schools. TechTrends, 63(6), 682-692.

Lee, C. C., Czaja, S. J., Moxley, J. H., Sharit, J., Boot, W. R., Charness, N., & Rogers, W. A. (2019).
Attitudes toward computers across adulthood from 1994 to 2013. The Gerontologist, 59(1),
22-33.

Letchumanan, M., & Muniandy, B. (2013). Migrating to e‐book: a study on perceived usefulness and
ease of use. Library Hi Tech News.

Li, J. (2020), Blockchain technology adoption: Examining the Fundamental Drivers. ACM Publication,
pp. 253–260. doi:10.1145/3396743.3396750

49
Lieberman, M. (2017, November 29). Enough With the Laptop Ban Debate! Retrieved from Inside
Higher Ed: https://www.insidehighered.com/digital-learning/article/2017/11/29/debate-
over-banning-laptops-resurfaces-academics-seek-more

Lim, W. M. (2018). Dialectic antidotes to critics of the technology acceptance model: Conceptual,
methodological, and replication treatments for behavioural modelling in technology-
mediated environments. Australasian Journal of Information Systems, 22.

Lindeman, M., & Verkasalo, M. (2005). "Measuring Values With the Short Schwartz's Value Survey".
Journal of Personality Assessment. 85 (2): 170–178.

Loria, K. (2018, June 23). Past generations created a climate crisis for Millennials and Generation Z.
Today marks 30 years of inaction. Retrieved from Business insiders:
https://www.businessinsider.nl/baby-boomers-millennials-climate-change-generations-
2017-7?international=true&r=US

Luckin, R., Holmes, W., Griffiths, M., & Forcier, L. B. (2016). Intelligence unleashed: An argument for
AI in education. Pearson Education, London.

Lundblad, J. P. (2003). A review and critique of Rogers' diffusion of innovation theory as it applies to
organizations. Organization Development Journal, 21(4), 50.

Lyytinen, K., & Damsgaard, J. (2001, April). What’s wrong with the diffusion of innovation theory?. In
Working conference on diffusing software product and process innovations (pp. 173-190).
Springer, Boston, MA.

Mandela, N., (1990). Nelson Mandela : the struggle is my life : his speeches and writings brought
together with historical documents and accounts of Mandela in prison byfellow-prisoners.
London :IDAF Publications,

Marr, B. (2018, July 25). How Is AI Used In Education -- Real World Examples Of Today And A Peek
Into The Future. Retrieved from Bernard Marr & Co:
https://bernardmarr.com/default.asp?contentID=1541

Marr, B. (2018, Novemer 19). Is Artificial Intelligence Dangerous? 6 AI Risks Everyone Should Know
About. Retrieved from Forbes: https://www.forbes.com/sites/bernardmarr/2018/11/19/is-
artificial-intelligence-dangerous-6-ai-risks-everyone-should-know-about/?sh=6d177ad52404

Medlin, B.D. (2001). The factors that may influence a faculty member's decision to adopt electronic
technologies in instruction (Doctoral dissertation, Virginia Polytechnic Institute and State
University, 2001). ProQuest DigitalDissertations. (UMI No. AAT 3095210).

50
Milano, M. (2019, March 12). The digital skills gap is widening fast. Here’s how to bridge it. Retrieved
from WeForum: https://www.weforum.org/agenda/2019/03/the-digital-skills-gap-is-
widening-fast-heres-how-to-bridge-it/

Mlekus, L., Bentler, D., Paruzel, A., Kato-Beiderwieden, A. L., & Maier, G. W. (2020). How to raise
technology acceptance: user experience characteristics as technology-inherent determinants.
Zeitschrift für Angewandte Organisationspsychologie (GIO), 51(3), 273-283.

Mozilla. (2019, November 7). We Asked People Around the World How They Feel About Artificial
Intelligence. Here’s What We Learned. Retrieved from Mozilla:
https://foundation.mozilla.org/en/blog/we-asked-people-around-the-world-how-they-feel-
about-artificial-intelligence-heres-what-we-learned/

Mueller, P. A., & Oppenheimer, D. M. (2014). The pen is mightier than the keyboard: Advantages of
longhand over laptop note taking. Psychological science, 25(6), 1159-1168.

Mühlig-Versen, A., Bowen, C. E., & Staudinger, U. M. (2012). Personality plasticity in later adulthood:
Contextual and personal resources are needed to increase openness to new experiences.
Psychology and Aging, 27(4), 855.

Murphy, D. (2016). A Literature Review: The Effect of Implementing Technology in a High School
Mathematics Classroom. International Journal of Research in Education and Science, 2(2),
295-299.

Nagourney, E. (2013, February 13). Why is Gaia Angry With Me? Retrieved from The New York Times:
https://www.nytimes.com/2013/02/15/booming/baby-boomers-produce-highest-carbon-
emissions.html

Nikou, S. A., and Economides, A. A. (2017). Mobile-based assessment: investigating the factors that
influence behavioral intention to use. Comput. Educ. 109, 56–73.
doi:10.1016/j.compedu.2017.02.005

Ozturk, A. B., Bilgihan, A., Nusair, K., and Okumus, F. (2016). What keeps the mobile hotel booking
users loyal? Investigating the roles of self-efficacy, compatibility, perceived ease of use, and
perceived convenience. Int. J. Inf. Manag. 36, 1350–1359. doi:
10.1016/j.ijinfomgt.2016.04.005

Parisot, A.H. (1995). Technology and teaching: The adoption and diffusion of technological
innovations by a community college faculty. (Doctoral dissertation, Montana State University,
1995). ProQuest DigitalDissertations. (UMI No. AAT 9542260).

51
Paudyal, P., Lee, J., Kamzin, A., Soudki, M., Banerjee, A., & Gupta, S. K. (2019). Learn2Sign:
Explainable AI for Sign Language Learning. In IUI Workshops 2019, Arizona State University

Paykamina, B. (2021, March 29). What’s Driving Growth in the Ed Tech Market? Retrieved from
Government Technology: https://www.govtech.com/education/k-12/whats-driving-growth-
in-the-ed-tech-market.html

Pega. (2020). What Consumers Really Think About AI: A Global Study. Retrieved from Pega:
https://hyken.com/wp-content/uploads/what-consumers-really-think-about-ai.pdf

Pew Research Center. (2020, June 24). Younger Republicans prioritize alternative energy sources
more than older Republicans. Retrieved from Pew Reseach Center:
https://www.pewresearch.org/fact-tank/2020/06/24/millennial-and-gen-z-republicans-
stand-out-from-their-elders-on-climate-and-energy-issues/ft_2020-06-
24_climateupdate_01/

Piletic, P. (2018, September 10). Examining The Positive And Negative Impacts of AI On Education.
Retrieved from Smart data collective: https://www.smartdatacollective.com/examining-the-
positive-and-negative-impacts-of-ai-on-education/

Purcell, K., Buchanan, J., & Friedrich, L. (2013, February 28). How Teachers Are Using Technology at
Home and in Their Classrooms. Retrieved from Pew Research Center:
https://www.pewresearch.org/internet/2013/02/28/how-teachers-are-using-technology-at-
home-and-in-their-classrooms/

Ramayah, T., & Ignatius, J. (2005). Impact of perceived usefulness, perceived ease of use
and perceived enjoyment on intention to shop online. ICFAI Journal of Systems Management
(IJSM), 3(3), 36-51.

Raza, S. A., Umer, A., & Shah, N. (2017). New determinants of ease of use and perceived usefulness
for mobile banking adoption. International Journal of Electronic Customer Relationship
Management, 11(1), 44-65.

Reed, L. (2016, January 15). Digital distraction in class is on the rise, study says. Retrieved from Phys
Org: https://phys.org/news/2016-01-digital-distraction-class.html

Reeves, T. C. (1998). The impact of media and technology in schools. Journal of The Journal of Art and
Design Education, 2, 58-63.

52
Reinhart, R. (2018, May 11). Global Warming Age Gap: Younger Americans Most Worried. Retrieved
from Gallup News: https://news.gallup.com/poll/234314/global-warming-age-gap-younger-
americans-worried.aspx

Rimer, S. (2020, November 9). BU Students: Zoom vs In-Person Classes? It’s Complicated. Retrieved
from Boston University: http://www.bu.edu/articles/2020/students-learn-from-anywhere/

Ritchie, H., & Roser, M. (2019, September). Age Structure. Retrieved from OurWorldinData:
https://ourworldindata.org/age-
structure#:~:text=The%20global%20median%20age%20has,bracket%20between%2025%20a
nd%2065.

Rockmore, D. (2014). The case for banning laptops in the classroom. The New Yorker, 6

Rogers, E.M. (1962) Diffusion of Innovations. Free Press, New York.

Rogers, E. (2003). Diffusion of Innovations. Fifth edition. Free Press: New York.

Rouhiainen, L. (2019, October 14). How AI and Data Could Personalize Higher Education. Retrieved
from Harvard Business Review: https://hbr.org/2019/10/how-ai-and-data-could-
personalize-higher-education

Sahin, I. (2006). Detailed review of Rogers' diffusion of innovations theory and educational
technology-related studies based on Rogers' theory. Turkish Online Journal of Educational
Technology-TOJET, 5(2), 14-23.

Schacter, J., & Jo, B. (2017). Improving preschoolers’ mathematics achievement with tablets: a
randomized controlled trial. Mathematics Education Research Journal, 29(3), 313-327.

Schwartz, S. H. (1992), "Universals in the Content and Structure of Values: Theoretical Advances and
Empirical Tests in 20 Countries". Advances in Experimental Social Psychology, 25, Elsevier,
pp. 1–65

Schwartz, S. H. (2006). Basic human values: An overview.

Schwartz, S. H. (2012). An overview of the Schwartz theory of basic values. Online readings in
Psychology and Culture, 2(1), 2307-0919.

Schwartz, S. (2020, November 16). Survey: Teachers and Students Are Struggling With Online
Learning. Retrieved from Education Week: https://www.edweek.org/teaching-
learning/survey-teachers-and-students-are-struggling-with-online-learning/2020/11

53
Seels, B., Berry, L. H., Fullerton, K., & Horn, L. J. (1996). Research on learning from television. In D. H.
Jonassen (Ed.), Handbook of research for educational communications and technology (pp.
299-377). New York: Macmillan

Sekar, A. (2018, May 2). 5 Industries Which Rely Heavily on Artificial Intelligence and Machine
Learning. Retrieved from Jigsaw Academy: https://www.jigsawacademy.com/5-industries-
which-rely-heavily-on-artificial-intelligence-and-machine-learning/

Selwyn, N., & Aagaard, J. (2021). Banning mobile phones from classrooms—An opportunity to
advance understandings of technology addiction, distraction and cyberbullying. British
Journal of Educational Technology, 52(1), 8-19.

Shalamanov, J. (2021, February 2). Why AI Will Replace Some Jobs and Others Will Stick Around.
Retrieved from Udacity: https://www.udacity.com/blog/2021/02/why-ai-will-replace-some-
jobs-and-others-will-stick-
around.html#:~:text=According%20to%20a%20report%20from,by%202025%20due%20to%2
0AI.

Shelton, S. A., & Brooks, T. (2019). " We Need to Get These Scores Up": A Narrative Examination of
the Challenges of Teaching Literature in the Age of Standardized Testing. Journal of Language
and Literacy Education, 15(2), n2.

Singhal, S., Bagga, S., Goyal, P., & Saxena, V. (2012). Augmented chemistry: Interactive education
system. International Journal of Computer Applications, 49(15), 1-5.

Smith, A., & Anderson, M. (2018, March 1). Social Media Use in 2018. Retrieved from Pew Research
Center: https://www.pewresearch.org/internet/2018/03/01/social-media-use-in-2018/

Southworth, P. (2019, July 22). Teachers spend more time marking and planning than in the
classroom, Ofsted survey reveals. Retrieved from The Telegraph:
https://www.telegraph.co.uk/news/2019/07/21/teachers-spend-time-marking-planning-
classroom-ofsted-survey/

Spacey, J. (2019, February 15). Types of Media Technology. Retrieved from Simplicable:
https://simplicable.com/new/media-
technology#:~:text=Media%20technology%20is%20any%20hardware,reality%20and%20aug
mented%20reality%20environments.

Spitzer, M. (2014). Information technology in education: Risks and side effects. Trends in
Neuroscience and Education, 3(3-4), 81-85.

54
Suki, N. M., & Suki, N. M. (2011). Exploring the relationship between perceived usefulness, perceived
ease of use, perceived enjoyment, attitude and subscribers’ intention towards using 3G
mobile services. Journal of Information Technology Management, 22(1), 1-7.

Šumak, B., and Šorgo, A. (2016). The acceptance and use of interactive whiteboards among teachers:
differences in UTAUT determinants between pre-and post-adopters. Comput. Hum. Behav.
64, 602–620. doi: 10.1016/j.chb.2016.07.037

Taneja, A., Fiore, V., & Fischer, B. (2015). Cyber-slacking in the classroom: Potential for digital
distraction in the new age. Computers & Education, 82, 141-151.

Tempelaar, D., Rienties, B., & Nguyen, Q. (2021). The contribution of dispositional learning analytics
to precision education. Educational Technology & Society, 24(1), 109-122.

Thayer-Hart, N., Dykema, J., Elver, K., Schaeffer, N. C., & Stevenson, J. (2010). Survey fundamentals:
A guide to designing and implementing surveys. University of Wisconsin

The Learning Network. (2020, October 22). What’s Going On in This Graph? | World Leader Ages.
Retrieved from The New York Times: https://www.nytimes.com/2020/10/22/learning/whats-
going-on-in-this-graph-world-leader-ages.html

Thompson, R. L., Higgins, C. A., & Howell, J. M. (1994). Influence of experience on personal computer
utilization: Testing a conceptual model. Journal of management information systems, 11(1)
,167-187.

Tilley, J. (2015, November 3). Do we really become more conservative with age? Retrieved from the
Guardian: https://www.theguardian.com/commentisfree/2015/nov/03/do-we-become-
more-conservative-with-age-young-old-politics

Torani, S., Majd, P. M., Maroufi, S. S., Dowlati, M., & Sheikhi, R. A. (2019). The importance of
education on disasters and emergencies: A review article. Journal of education and health
promotion, 8.

Trucano, M. (2014, December 23). Collecting data about educational technology use in *all*
countries in the world. Retrieved from Edu Tech:
https://blogs.worldbank.org/edutech/collecting-data-about-educational-technology-use-all-
countries-world

55
Truong, D. (2020, February 2). More students are learning on laptops and tablets in class. Some
parents want to hit the off switch. Retrieved from The Washington Post:
https://www.washingtonpost.com/local/education/more-students-are-learning-on-laptops-
and-tablets-in-class-some-parents-want-to-hit-the-off-switch/2020/02/01/d53134d0-db1e-
11e9-a688-303693fb4b0b_story.html

Tsuboya-Newell, I. (2019, January 31). Why education must keep pace with technology to stay
relevant. Retrieved from The Japan Times:
https://www.japantimes.co.jp/opinion/2019/01/31/commentary/japan-
commentary/education-must-keep-pace-technology-stay-relevant/

Udeze, S. E., & Oko, T. (2013) The New Media and Digital Divide: Knowledge Gap Exacerbated.
Review of Public Administration and Management 2(3)

Uria-Recio, P. (2019, August 8). Artificial Intelligence will Make the Workplace More Human, not Less.
Retrieved from TedX: https://towardsdatascience.com/artificial-intelligence-will-make-the-
workplace-more-human-not-less-49af1ce6cd0d

van den Heuvel, S. (2020, March 26). COVID-19 beats current technology acceptance theories.
Retrieved from LinkedIn: https://www.linkedin.com/pulse/covid-19-beats-current-
technology-acceptance-theories-van-den-heuvel

van Lier, L. (2003). ‘A tale of two computer classrooms: The ecology of project-based language
learning’. In J. Leather and J. van Dam (Eds.), The Ecology of Language Acquisition (pp. 49-64).
Dordrecht: Kluwer Academic

Van Raaij, E. M.; Schepers, J. J. L. (2008). The acceptance and use of a virtual learning environment in
China. Computers & Education. 50(3), 838–852. doi:10.1016/j.compedu.2006.09.001

Venkatesh, V., & Davis, F. D. (1996). A Model of the Antecedents of Perceived Ease of Use:
Development and Test. Decision Sciences, 27(3), 451-481. doi:
https://doi.org/10.1111/j.1540-5915.1996.tb01822.x

Venkatesh, V., Morris, M. G., Davis,G. B., Davis, F. D. (2003). "User Acceptance of Information
Technology: Toward a Unified View". MIS Quarterly. 27(3): 425–478

Venkatesh, V., Thong, J. Y., & Xu, X. (2012). Consumer acceptance and use of information technology:
extending the unified theory of acceptance and use of technology. MIS quarterly, 157-178.

Vesselinov, R., & Grego, J. (2012). Duolingo effectiveness study. City University of New York, USA,
28(1-25).

56
Villalta, F. (2019, August 26). Importance of Computers in Our Daily Life. Retrieved from Find4Me:
https://www.find4me.com.au/computer/importance-of-computers-in-our-daily-life.htm

Walsh, P. (2020, July 2020). Innovative Technology Is The Future Of Education. Retrieved from Forbes:
https://www.forbes.com/sites/forbestechcouncil/2020/07/23/innovative-technology-is-the-
future-of-education/?sh=6c65685b70e3

Weale, S. (2020, December 8). Decline in science attainment among England's year 9 pupils.
Retrieved from The Guardian:
https://www.theguardian.com/education/2020/dec/08/decline-in-science-attainment-
among-england-year-9-pupils

Weng, F., Yang, R. J., Ho, H. J., & Su, H. M. (2018). A TAM-based study of the attitude towards use
intention of multimedia among school teachers. Applied system innovation, 1(3), 36.

Westera, W. (2015). Reframing the Role of Educational Media Technologies. Quarterly Review of
Distance Education, 16(2), 19-32.

Wiliam, D. (2010). Standardized testing and school accountability. Educational Psychologist, 45(2),
107-122.

Wong, J. (2021, March 15). Online school will still be around post-pandemic, so what have we
learned? Retrieved from CBC: https://www.cbc.ca/news/canada/ontario-online-virtual-
school-1.5944407

Wong, J. (2021, January 5). Students struggle as school goes online-only in 4 provinces after Christmas
break. Retrieved from CBC: https://www.cbc.ca/news/canada/school-new-year-virtual-
restart-1.5860964

Yamamoto, K. (2007). Banning laptops in the classroom: Is it worth the hassles?. Journal of Legal
Education, 57(4), 477-520.

57
Appendix A: Survey

Master Thesis - Attitudes towards Media technology/AI

Start of Block: Introduction

Q1 Welcome, my name is Timo, and thank you for taking the time to participate in this master thesis
survey.

Over the last few decades, media technologies such as laptops and tablets have become increasingly
important in education. Additionally, Artificial Intelligence, in for example learning applications, is
also starting to become very important in education. This survey aims to discover attitudes towards
the usage of media technologies and Artifical Intelligence in education, and what the reasons for
these attitudes could be.

The survey consists out of three stages. In the first stage questions are asked about your media usage
and personality. The second stage asks about your opinion of media technology usage in education.
The Third stage asks the same questions as the second, but now about your opinion on AI usage in
education.

Your responses are completely anonymous and will be held confidentially. They will not be shared
with any third party. Your participiation is voluntary and you may stop the survey at any point if you
do not wish to continue. The survey should take no longer than 5 to 6 minutes. Thank you for your
participation.

If you have any questions, please send them to 447060tl@student.eur.nl

Q2 By clicking 'I agree', you indicate that you take this survey voluntarily and that you are 18 years or
older.

o I agree (1)

End of Block: Introduction

58
Start of Block: Control variables/Indirect predictors

Q3 What is your gender?

o Male (1)

o Female (2)

o Non-binary / third gender (3)

o Prefer not to say (4)

Q4 What is your age?

o 18-24 years old (1)

o 25-39 years old (2)

o 40-64 years old (3)

o 65 years or older (4)

Q5 What is your nationality?

________________________________________________________________

Q6 What is the highest education that you have completed?

o None (1)

59
o High school (2)

o Practical education/Community college (6)

o Bachelor of applied Science (7)

o Bachelor of Science (3)

o Master (4)

o Doctorate (5)

End of Block: Control variables/Indirect predictors

Start of Block: Schwarz Human Values

Q17 Please indicate to what extent you agree with the following statements

Strongly disagree (1) Disagree (2) Somewhat disagree (3) Neither agree nor disagree (4)
Somewhat agree (5) Agree (6) Strongly agree (7)

I like to do things in my own original way (1) o o o o

o o o

I like to be free to plan and choose activities for myself (2) o o o

o o o o

I like to be curious and to try and understand all sorts of things (3) o o
o o o o o

I do not like to rely on myself (4) o o o o

o o o

I am always looking for new things to try (5) o o o o

o o o

I do not like taking risks (6) o o o o o

o o

60
I like surprises (7) o o o o o
o o

It is important to me to see life as an adventure (8) o o o

o o o o

Q18 Please indicate to what extent you agree with the following statements

Strongly disagree (1) Disagree (2) Somewhat disagree (3) Neither agree nor disagree (4)
Somewhat agree (5) Agree (6) Strongly agree (7)

I tend to avoid things that might endanger my safety (1) o o o

o o o o

I believe that people should follow rules at all times, even when no one is watching (2) o
o o o o o o

I never try to disturb or irritate others (3) o o o o

o o o

I believe people should be satisfied with what they have (4) o o o

o o o o

I like drawing attention to myself (5) o o o o

o o o

I believe it is important to follow the customs I have learned (6) o o

o o o o o

I tend to do things even if people say they are wrong (7) o o o

o o o o

I find it important that the established order is protected (8) o o o

o o o o

61
End of Block: Schwarz Human Values

Start of Block: New media literacy/Mobile self-efficacy & Experience

Q11 Please indicate to what extent you agree with the following statements

Strongly disagree (1) Somewhat disagree (2) Neither agree nor disagree (3) Somewhat
agree (4) Strongly agree (5)

I stay up to date with changes in the media (1) o o o o

o

I can determine when media content has commercial messages in them (2) o o
o o o

I sometimes find it difficult finding the online information I want (3) o o

o o o

I can distinguish facts and opinions in media content (4) o o o

o o

I sometimes find it hard to create user profiles/accounts on new websites (5) o o

o o o

I can use laptops/tablets/phones to create media content quite well (6) o o

o o o

I am good at sharing digital media contents and messages on the internet (7) o o
o o o

62
Q14 Please indicate how much the following technologies were used (by students or teachers) during
your education, for educational purposes

Never (1) A small amount (2) Sometimes (3) A lot (4) Most of the time (5)

Laptops (1) o o o o o

Tablets (2) o o o o o

Digital whiteboards (3) o o o o o

Learning apps (such as duolingo) (4) o o o o

o

Automatic grading systems (5) o o o o o

Personal online quizzes or tests (6) o o o o

o

End of Block: New media literacy/Mobile self-efficacy & Experience

Start of Block: All MT related questions (Perceived usefulness/ease of use & Attitude)

Q8 The following sets of questions are about new media technologies in education.

You might be wondering what the difference between media technologies and AI in education is. The
main difference is that media technologies are hardware, so for example laptops, tablets or phones,
while AI is certain software run on those media technologies. Certain software that reacts to your
input, just as how Siri or Google reacts to your input.

63
Some examples of new media technologies that are used in education are laptops, tablets, phones,
digital whiteboards and other technologies of that kind.

Q12 Indicate to what extent you agree with the following statements

Strongly disagree (1) Somewhat disagree (2) Neither agree nor disagree (3) Somewhat
agree (4) Strongly agree (5)

Media technologies in education make students accomplish tasks quicker (1) o o

o o o

Media technologies in education increase productivity (2) o o o

o o

Media technologies in education are not effective (3) o o o

o o

Media technologies in education help out teachers (4) o o o

o o

Media technologies in education make it easier to do work (5) o o o

o o

Media technologies in education are useful (6) o o o o

o

Q13 Please indicate to what extent you agree with the following statements

Strongly disagree (1) Somewhat disagree (2) Neither agree nor disagree (3) Somewhat
agree (4) Strongly agree (5)

64
I think it is easy to learn how to use media technologies (1) o o o
o o

I think teachers find it easy to teach using media technologies (2) o o

o o o

I think students find it more difficult to learn using media technologies (3) o o
o o o

It is hard to become skillfull in using media technologies (4) o o o

o o

It is hard to understand how media technologies work (5) o o o

o o

Media technologies in education are easy to use (6) o o o

o o

Q7 Please indicate to which extent you agree with the following statements

Strongly disagree (1) Disagree (2) Somewhat disagree (3) Neither agree nor disagree (4)
Somewhat agree (5) Agree (6) Strongly agree (7)

Media technologies provide solutions to many problems in education (1) o o

o o o o o

Media technologies in education are too distracting (2) o o o

o o o o

It is important that media technologies are used in education (3) o o

o o o o o

Media technologies in education make students accomplish more (4) o o

o o o o o

65
Media technologies dehumanize education (5) o o o o
o o o

The positives of media technologies in education outweigh the negatives (6) o o

o o o o o

End of Block: All MT related questions (Perceived usefulness/ease of use & Attitude)

Start of Block: All AI related questions (Perceived usefulness/ease of use & Attitude)

Q10 The following questions are about the usage of artifical intelligence (or AI in short) in education.

Some examples of how artifical intelligence is used in education, are learning apps that adapt based
on your strenghts and weaknesses, automatic grading, personalized curriculums for students, and
data analytics to identify the needs of students.

The following questions are mostly similar to the ones you just filled in. However, these questions are
on a new topic, so please do not worry you are repeating yourself.

You might be wondering what the difference between media technologies and AI in education is. The
main difference is that media technologies are hardware, so for example laptops, tablets or phones,
while AI is certain software run on those media technologies. Certain software that reacts to your
input, just as how Siri or Google reacts to your input.

66
Q15 Indicate to what extent you agree with the following statements

Strongly disagree (1) Somewhat disagree (2) Neither agree nor disagree (3) Somewhat
agree (4) Strongly agree (5)

AI in education makes students accomplish tasks quicker (1) o o o

o o

AI in education increases productivity (2) o o o o

o

AI in education is not effective (3) o o o o

o

AI in education helps out teachers (4) o o o o

o

AI in education makes it easier for students to do work (5) o o o

o o

AI in education is useful (6) o o o o o

Q19 Please indicate to what extent you agree with the following statements

Strongly disagree (1) Somewhat disagree (2) Neither agree nor disagree (3) Somewhat
agree (4) Strongly agree (5)

67
I think it is easy to learn how to deal with AI (1) o o o
o o

I think teachers find it easy to teach using AI (2) o o o

o o

I think students find it more difficult to learn using AI (3) o o o

o o

It is hard to become skillfull in dealing with AI in education (4) o o o

o o

It is hard to understand how AI work (5) o o o o

o

AI in education is easy to use (6) o o o o

o

Q9 Please indicate to which extent you agree with the following statements

Strongly disagree (1) Disagree (2) Somewhat disagree (3) Neither agree nor disagree (4)
Somewhat agree (5) Agree (6) Strongly agree (7)

AI provides solutions to many problems in education (1) o o o

o o o o

AI in education is too distracting (2) o o o o

o o o

It is important that AI is used in education (3) o o o o

o o o

AI in education helps students to accomplish more (4) o o o

o o o o

68
AI dehumanizes education (5) o o o o o
o o

The positives of AI in education outweigh the negatives (6) o o o

o o o o

End of Block: All AI related questions (Perceived usefulness/ease of use & Attitude)

69