Investigation of users’ opinions, perceptions and attitudes towards the use of artificial intelligence in online learning environments
Categoria dell'articolo: Research Article
Pubblicato online: 19 nov 2024
Pagine: 70 - 92
DOI: https://doi.org/10.2478/eurodl-2024-0007
Parole chiave
© 2024 Hamza Aydemir et al., published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.
We are entering the age of artificial intelligence (AI), as AI and its applications permeate people’s work, education and daily lives. Like other technological developments, AI will affect how we learn as well as how we teach. Therefore, the education community must deal with the opportunities and challenges of AI technologies that can change radically the structure, functioning and governance of educational institutions (Chan, 2023; Popenici & Kerr, 2017). Accordingly, countries have started to implement next-generation talent education strategic planning and reap the advantages of being a pioneer in national AI development (Luo et al., 2019; Zhang & Cui, 2020).
Although societies have come to recognise the importance of AI, there is no clarity and consensus on what innovations AI will bring. Although the use of AI in education and its applications continue to develop rapidly (Duan & Gong, 2019), some challenges and uncertainties still remain. Therefore, it is necessary to collect users’ opinions and perceptions about AI and examine how they change over time.
Although AI technology encompasses a vast array of content, a comprehensive AI curriculum remains lacking, while AI applications in education are still in the early stages of development (Zhang & Cui, 2020). Furthermore, teachers and students have not yet acquired adequate knowledge on the subject of using AI in education, and understanding the algorithms that underpin AI applications remains a challenge for both. Given these circumstances, it is essential to consider the perspectives and attitudes of teachers, students and parents involved in AI and expect to be even more engaged in the future.
AI refers to the ability of a digital machine to execute tasks usually associated with intelligent beings, such as humans; it is divided into several branches such as computer vision, speech recognition, machine learning, big data analysis and natural language processing (Chiu et al., 2022; Xia et al., 2022). AI in education holds substantial potential to improve learning, teaching, assessment and educational management by providing students with more personalised and adaptive learning experiences, promoting teachers’ understanding of students’ learning process and providing immediate feedback (Chiu et al., 2023).
AI applications in education are utilised not only in traditional classroom settings, but also in online education environments where instruction and materials are transmitted via the Internet. Online education refers to education in which instruction and content are delivered over the Internet (Watson & Kalmon, 2005). Online education is commonly defined as realisation of the learning experience through technology (Conrad, 2002). However, there are also researchers who define online education by the technology used or by the characteristics of distance education with an emphasis on the word online. As defined by Keegan (1980), distance education is characterised by the physical separation of teachers and students, eliminating face-to-face interaction. Another point emphasised in the definition of distance education is the dissemination of educational content over computer networks, and the two-way communication. Khan (1998) focuses on the distribution of resources in distance education and refers to their synchronous and asynchronous use. Piña (2010) defined distance education as a learning style that replaces traditional face-to-face teaching or supports face-to-face teaching. From this point of view, it can be inferred that face-to-face teaching emphasises the real world whereas online teaching emphasises the virtual world. Lee et al. (2002) used the concept of the virtual learning environment. As can be seen, different concepts such as distance education, e-learning, virtual learning and cyber learning are used in online learning (Paulsen, 2002). Moore et al. (2011) emphasised the difficulty of providing a single valid definition of online learning, and stated that some researchers refer to the technology environment and context when discussing online education.
By increasing the teacher–student interaction in online teaching and learning environments, it is anticipated that student involvement will escalate and, consequently, academic attainment will be enhanced via efficient and effective communication (Gecer, 2013; Harper, 2018). Due to the global impact of the pandemic and the recent calamities affecting Türkiye, schools have been forced to shift to online teaching, albeit temporarily. However, teachers have encountered various challenges in monitoring students’ progress during and after online lessons due to the absence of the physical presence of students. They also experience difficulties in keeping students engaged in online classes. One of the reasons for this is that teachers often have limited resources and support in creating an effective environment to motivate student learning (Upton, 2006). Teaching in an online environment requires pedagogical strategies aimed at maximising student engagement (Bomia et al., 1997; Muir et al., 2022; Vermeulen & Volman, 2024). As a result, it is worth exploring the use of AI and its applications in such educational settings.
The effects of AI on education are yet to be fully understood (Holmes et al., 2021). Therefore, more research is needed to understand how emerging AI technology benefits education (Chiu et al., 2023). A similar gap exists in the review of studies that address stakeholders’ views on the use of AI in online learning (Raffaghelli et al., 2022). This study aims to examine the opinions, perceptions and attitudes of teachers and students towards the use of AI in online education through a systematic review of the relevant literature. We hypothesise that AI offers an opportunity to increase student engagement and hence motivation in online learning, which will provide some insights to improve the quality of online learning. Understanding the perspectives of teachers, students and parents on AI can offer valuable insights for those involved in the development and implementation of AI in education. Furthermore, this study presents important insights for researchers conducting studies in the AI education domain. These findings are important for the following reasons. First, the rise in the number of publications offers an understanding of the growth of the global scientific community in this field (Aksnes et al., 2016). Second, scrutinising journals aids researchers and students who are either currently or potentially working within the field to recognise sources of publications and to contribute to the field with the greatest potential for research development and publication (Mair & Reischauer, 2017; Song et al., 2019). Third, the examination of research subjects allows for recognition of the key themes in research publications within a specific field, spanning past, present and future (Snyder, 2011).
To attain the objective of the research, the study attempted to answer the following questions:
1. What information does the literature (2020–2023) provide to researchers who will conduct studies in the field of AI in education? 1.1. How is the distribution of studies according to years? 1.2. In which journals were the studies published? 1.3. In which disciplines were the studies conducted? 1.4. What is the sample group and size of the studies? 1.5. Which data collection tools were used in the studies? 2. For what purposes are AI used in an online learning environment, and how do they compare with traditional educational methods? 2.1. Especially how are chatbots used in education? 3. What are teachers, students and parents’ opinions, perceptions and attitudes towards using AI in an online learning environment? 3.1. What are the perceptions, opinions and towards especially for chatbots?
Systematic reviews of the literature on AI in education can provide insights for further research by identifying the current trends, gaps and challenges in the field. In this context, in this stage, current studies were analysed by considering their research aims, methods and findings. The studies focused on the use of chatbots in education, the advantages and disadvantages of chatbots and the role of AI in education.
Chen et al. (2020) aimed to provide multiple perspectives on the development of AI in education in terms of relevant conferences, journals, citations, categories, sources, countries, impacts, tools and keywords. In this context, the researchers reviewed and reported the literature in various aspects such as the number of AI in education publications, citations, categories, sources, countries, impacts, tools and keywords.
Okonkwo and Ade-Ibijola (2021) presented a systematic review of previous studies on the application of chatbots in education. In this context, the researchers used the protocol guidelines for systematic reviews in software engineering introduced to the literature by Kitchenham and Charters (2007). Accordingly, they presented the benefits and challenges related to the application of chatbots in education. The study suggested prospective areas for future research concerning the application of chatbot technology in education.
Zhan et al. (2022) aimed to review empirical studies by adopting a game-based learning approach with AI, and to reveal future research perspectives. The researchers identified a total sample of 125 empirical articles through a systematic keyword search and snowballing approach in the online database. The results showed that games are used in AI education in three main ways: as a teaching tool, as student work and as a competition mechanism. However, in the studies, researchers generally examined the impact of gamebased learning on students’ opinions and attitudes, and learning achievement.
Chiu et al. (2023) focused on the problem of neglecting the relationships between AI and learning outcomes for students and teachers in the literature on AI applications in education. In this context, they aimed to reveal the opportunities and challenges of using AI in education. The authors utilised Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA), matrix coding and content analysis to analyse the literature from the past. The results revealed that the 13 roles of AI technologies in basic education areas support 7 learning outcomes of AI in education. In addition, 10 major challenges in the literature were identified. The study also presents a schematic representation of the roles and outcomes of AI applications within education.
Lin and Yu (2023) attempt to address the research gap by offering an educational outlook to scrutinise the present AI chatbot literature. In this context, they combined bibliometric and citation network analyses. Consequently, they revealed that there is advancement in the academic use of chatbots in the course of post-COVID-19 learning environments, and this expansion persists. Popular research areas on this topic include students’ perception of chatbots and their effectiveness in different educational contexts. The researchers presented an extended framework to facilitate the expansion of AI chatbot applications in education.
The field of AI in education aims to provide comprehensive and in-depth knowledge of AI and its implications for teaching and learning. Previous studies have investigated the effects of AI on various educational outcomes and contexts, but most of them have focused on the application and evaluation of specific AI tools or systems. However, there is a lack of systematic reviews that synthesise the perspectives and experiences of teachers and students who interact with AI in education. This study addresses this gap by conducting a systematic review of the literature on AI tools, chatbots and machine learning tools in education. Unlike previous reviews, this study takes a holistic approach to examine the use of AI in education.
A systematic literature review was conducted to explore studies on increasing students’ engagement in an online learning environment using AI. A systematic review is a comprehensive process used to collect and synthesise findings from previous studies to gather evidence and answer research questions (Xiao & Watson, 2019). The PRISMA guidelines were followed to improve the reporting of the systematic review which was conducted within the scope of the study (Moher et al., 2015). Figure 1 shows the PRISMA flowchart of the systematic review process.
PRISMA flowchart of the systematic review conducted within the scope of the study. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.
As seen in Figure 1, 207 articles accessed through the database by applying PRISMA guidelines were reduced to 22 articles at the end of the process; These 22 articles were analysed for qualitative analysis.
The search strategy comprises three stages: identification, review and eligibility. The stages of the search strategy are explained in the following section.
For this study, we searched the Web of Science, ScienceDirect, Taylor and Francis and Wiley databases as they cover numerous educational journals (Guan et al., 2020). Web of Science is a research platform that provides access to a wide array of academic literature. ScienceDirect is a database containing a comprehensive collection of scientific and technical journals. Taylor and Francis is a publisher of scholarly books and journals. Wiley is a publisher of print and electronic books, journals, encyclopaedias and educational materials.
Keywords were used at the beginning of the search process in the databases. This technique is preferred to select articles from a pool of words tagged by their authors. The search was conducted in April 2023 using English keywords. The selection of keywords was based on Lin and Yu’s (2023) bibliometric analysis, which reviews the existing literature on AI chatbots from an educational perspective. The selected keywords were finalised by a group of four researchers, who are experts in the field of computer and instructional technologies in two rounds of discussion, with a 1-week interval. The search terms were divided into three categories:
1. Online Learning (‘online learning’ 2. Opinion (‘opinion’ 1. AI (‘artificial intelligence’
The document type was specified as article and the research field as education. Furthermore, considering the rapid development of AI, the focus was on current studies. In this context, the study considered the past 4 years, covering the post-pandemic period. In the database search conducted at this stage, 77 articles were accessed in Web of Science, 114 in ScienceDirect, 11 in Taylor and Francis, and 5 in Wiley.
In this stage, we reviewed the 207 articles identified in the first stage. Initially, we examined the article titles and obtained 124 potential articles. We then identified and removed 10 duplicate articles that appeared in multiple databases, leaving us with 114 articles for analysis. We reviewed the abstracts of these remaining articles using the inclusion and exclusion criteria outlined in Table 1. The criteria were developed by examining review studies in the literature.
Inclusion and exclusion criteria of the articles
| Inclusion criteria | Exclusion criteria |
|---|---|
| Must be written in English. | Written in a language other than English. |
| Full text should be accessible. | The full text is not available. |
| It should be experimental to demonstrate the effects of AI on student engagement. | Non-experimental, using self-report only, such as interviews or questionnaires. |
| It is a conference paper or book chapter. |
AI, artificial intelligence.
In the review process, after applying the inclusion and exclusion criteria outlined in Table 1, 58 articles were excluded. Consequently, 56 articles underwent full-text analysis.
In this stage, researchers thoroughly assessed the complete texts of 56 identified articles from the second stage for relevance. Upon completion of the reading process, we concluded that 34 articles were irrelevant. The articles were disregarded as they did not utilise AI for stakeholders, were conceptual articles or were opinion or definition papers that did not mention stakeholder perspectives. Thus, at the end of the selection process, 22 articles remained for systematic analysis. The reference list of the studies included in the review is presented in Appendix A. The PRISMA flowchart in Figure 1 shows the identification, review and eligibility processes used.
The data were obtained by collecting and coding information extracted from 22 selected studies. In the coding process, the data extraction approach proposed by Petticrew and Roberts (2006) was adopted. The data were organised according to the title, year of publication, author(s), discipline(s) of the study, sample group, sample size, data collection method, the purpose of the study, the outcome of the study, the role of AI in the study and the results obtained about the stakeholders’ views on AI. In this study, the articles were reviewed thoroughly to identify relevant sections that discussed the use of AI and the perspectives of stakeholders. These sections were then analysed to categorise stakeholders’ views on AI as either positive or negative, and determine the role of AI in the research. The results compiled in this context are presented in Appendix B. The coding categories regarding stakeholders’ perspectives on AI and their role in the study were established via a mixed methodology that blends inductive and deductive methods, including thorough readings and re-readings of the information (Fereday & Muir-Cochrane, 2006).
The literature review was carried out using an exploratory approach. The studies were downloaded onto the researcher’s computer and the titles of the studies were reported in a spreadsheet file. The same researcher reviewed the titles and removed duplicate studies, and the abstract was reviewed. The same researcher also conducted the process in the eligibility stage where the full text of the studies was read. In all these processes, the validation checks approach proposed by Pettticrew and Roberts (2006) was adopted. In this context, an expert from the field of instructional technology was involved in the process as an independent reviewer and performed the validation checks. To ensure the reliability of the study, databases containing peer-reviewed journals and the main reference point for systematic reviews were used in the database search phase. Furthermore, the results of the analyses were co-established by the researchers with continuous and repeated revisions via discussion, doubt elimination and consensus-building (Boesch et al., 2013). Thus, interrater reliability was fully ensured.
The articles analysed within the scope of the study are presented in Appendix B. Two studies were undertaken in 2020, with a further four in the succeeding year, 11 in 2022 and five in 2023. The studies were mostly conducted with secondary school and undergraduate students and almost all of them used questionnaires as data collection tools.
When the studies selected within the scope of the first research question were examined, the number of articles by year and the journals in which the studies were published were revealed. The distribution of the disciplines in which AI studies were conducted, the sample group studied and the data collection tools used in the study were found to be useful for researchers in future studies. The selected studies’ distribution by year is demonstrated in Figure 2. Two studies were conducted in 2020, four in 2021, 11 in 2022 and five in 2023.
Number and distribution of selected studies by year of publication.
As seen in Figure 2, there is a noticeable upward trend in the number of studies conducted each year, with the highest number occurring in 2022. In 2023, fewer studies were conducted, which can be attributed to the fact that this study covers the data from the first 4 months of 2023. Researchers anticipate that the total number of publications in 2023 will likely surpass those in 2022, given the ongoing growth in the field.
It is important for researchers to reveal the journals preferred by the AI community in education. In this regard, Table 2 lists the journals in which the selected studies were published.
The journals in which the selected studies were published and the number of studies in the relevant journal
| Journals | Number of studies |
|---|---|
| 1 | |
| 1 | |
| 1 | |
| 4 | |
| 4 | |
| 2 | |
| 1 | |
| 1 | |
| 1 | |
| 1 | |
| 1 | |
| 1 | |
| 1 | |
| 1 | |
| 1 |
Table 2 shows that academic researchers studying AI in education have a higher propensity for utilising journals such as
When analysing the selected studies according to their respective disciplines, five studies did not specify their scope and discipline. The remaining studies’ disciplines are represented in Figure 3.
The disciplines in which the selected studies are conducted.
Figure 3 shows that the majority of studies on AI in education took place in the science discipline, with a total of three studies. Additionally, two studies were conducted in Business Administration, Mathematics, Medicine and Foreign Language disciplines, respectively, while one study was carried out in other fields. Furthermore, a few of the selected studies encompassed more than one discipline.
The studies were analysed with regard to their sample characteristics. Figure 4 presents details of the sample group, sample size and number of included studies. The findings presented in this study are based on data from a total of 2,042 participants across 22 studies.
Sample groups and the size of the groups in the selected studies.
As can be seen in Figure 4, most studies on AI in education primarily targeted undergraduate students. Accordingly, out of 22 studies, 11 were conducted with undergraduate students, with a cumulative participant count of 836 students. It was found that the second largest study group was secondary school students with 426 students. Accordingly, six studies focused on secondary school students. In two studies with undergraduate and graduate mixed groups, 134 students took part. In summary, AI studies in education have largely focused on secondary school, undergraduate and graduate students. There are also two studies conducted with teachers. A total of 386 secondary school teachers took part in these studies. In one of the studies, the participant group consisted of 74 students taking courses in MOOCs.
Analysis of the tools used for data collection among teachers and students in the studies showed that questionnaires were the predominant tool used, featuring in 21 of the studies. In the remaining 1 study, the interview method was preferred as the data collection tool. One of the studies that utilised questionnaires also incorporated interviews as the secondary data collection tool, while in another study, group discussion was used. The research findings indicated a greater preference among teachers and students for using questionnaires to collect data. However, there is a need to diversify the data collection tools in order to delve more deeply into the opinions, perceptions and attitudes towards AI applications in education.
Within the scope of the second research question, the role of AI in educational environments was examined when the selected studies were analysed. The studies were categorised according to their purposes and the way AI is used (Table 3).
The role of AI in selected studies
| Categories | Study no. | Number of studies |
|---|---|---|
| Chatbot | [5], [6], [10], [12], [13], [18], [20] | 7 |
| Content | [3], [7], [11], [17], [21] | 5 |
| Learning assistant | [9], [14], [19], [22] | 4 |
| Encouragement to learn | [2], [4], [8] | 3 |
| Measurement and evaluation | [1], [16] | 2 |
| Early warning system | [15] | 1 |
AI, artificial intelligence.
As seen in Table 3, AI in education has found the most application area as a chatbot. It is also widely used as a course content and learning assistant. It was found that AI was used in online education environments for various purposes such as encouraging participants to learn and increasing their motivation, experiencing an AI application and using it as an early warning system based on student data.
In the selected studies, it was seen that AI was most commonly used as a chatbot. In most of the studies on AI in education, AI chatbots and interactive books that allow students to engage in conversation with machines about their learning experiences were used (Chiu et al., 2023). Chatbots have proven to be highly beneficial in online education for teaching specific subjects, enhancing students’ language skills, answering common questions, correcting students’ misunderstandings, alleviating stress and anxiety, improving teachers’ pedagogical abilities and enhancing students’ communication skills (Figure 5).
The use of Chatbot in online educational environments.
In the selected studies, it was seen that one of the common purposes of AI was to be used as course content or to produce course materials. AI is used in educational environments as a content or content generation tool for purposes such as understanding and applying deepfake technology by students, testing AI models in the real world, teaching AI for medical imaging, automatic creation of course materials according to the student and training in AI applications in medicine.
It was also observed that one of the common roles of AI is serving as a learning assistant. AI is used in educational environments as an assistant to provide a personalised learning experience for the professional development of teachers, to provide students with selfdevelopment for the course outside of online courses, to encourage students to set their goals for the course before the course starts and to eliminate students’ pronunciation problems, expand their vocabulary and improve their listening–speaking skills in foreign language education.
Furthermore, AI applications play a pivotal role in analysing students’ cognitive performance, encouraging them to learn, and fostering students’ motivation through personalisation and gamification. It is also employed in assessing and evaluating student outputs, autonomously grading them, verifying online exam candidate identities via face recognition and detecting possible academic misconduct by analysing student conduct. Moreover, AI is used in educational environments to predict whether students are at risk of failing the course and to provide early intervention to the problem.
Students were positive about the use of AI in online learning environments. They regarded automated assessment processes as interesting, effective and reliable, especially when used in combination with traditional assessment methods. They reported that the use of AI, such as adaptive learning systems and chatbots, enhanced their learning experience and increased their motivation and engagement. Moreover, they expressed satisfaction with the assistance and companionship provided by chatbots, and the ease of utilising AI tools in online learning platforms, along with their accessibility. They expressed their belief that AI can improve their learning outcomes and deepen their understanding of the courses.
Students expressed that they perceived the AI applications they experienced within the scope of the study positively. They perceived chatbots as useful and fun tools that provide additional support, feedback and interaction in their learning process. Students emphasised that they value the conversational quality of chatbots, which increases their sense of interaction compared with traditional teaching methods. In addition, students were satisfied with the user-friendly and readily accessible nature of AI tools, acknowledging their role in facilitating learning both inside and outside the classroom. Students also perceived AI as a tool to generate new ideas, solve real problems and improve decision-making processes. In short, the perception of the positive impact of AI on learning and teaching processes was a common point for both students and teachers.
The majority of students showed a positive attitude towards AI and its applications, especially in health-related contexts. Their opinions underscored the potential for AI to enhance learning experiences and improve learning outcomes. Students enjoyed using AI tools such as chatbots and found them interesting, joyful and reliable. The personalisation of AI interactions, including features such as the robot addressing them by name during the chat and including emojis in the chat, is in line with students’ expectations. Students expressed a desire for more AI integration in their courses, highlighting the benefits of AI in clarifying learning objectives, providing extra information and promoting a deeper understanding of topics. However, a common finding across the different studies was that some students were concerned about limited responses, repetition of answers and privacy issues related to sharing personal information with chatbots.
In summary, the findings from the 22 selected studies show that students have positive views about the effectiveness and validity of AI applications and perceive AI applications as valuable tools that enhance learning experiences and increase engagement. It has been a common finding in many studies that AI is useful in online education environments for decision-making, problem-solving and generating new ideas. In general, students’ perceptions of chatbots are positive. They liked the support, friendship and easy use of chatbots. The overall attitude towards AI is positive, students are eager for more integration of AI into their courses, and teachers and students believe in the potential of AI to improve teaching–learning outcomes (Figure 6).
Students perception in relation to the use of AI in online education. AI, artificial intelligence.
The number of studies conducted with teachers on the use of AI in education is limited. However, these studies offer valuable insights into teachers’ views and perceptions about AI. In general, teachers favour the integration of chatbots into teaching activities and are eager to use them in education. In this way, they believe that the quality of teaching will increase. An intriguing discovery from the analysed studies is that there exists a differentiation between the views of teachers and students in regard to the language used by chatbots. Teachers stated that they prefer chatbots with a formal language to chatbots with a more social language. The fact that there is no study that addresses the opinions, perceptions or attitudes of parents in the analysed studies points to a gap in the literature.
This study aims to examine the opinions, perceptions and attitudes of teachers and students towards the use of AI in online education through a systematic review of the literature on AI applications in online learning environments. The PRISMA guidelines were used to enhance the reporting and transparency of the systematic review conducted for this purpose. The 207 articles accessed through the database by applying PRISMA guidelines were narrowed down to 22 articles at the end of the process. In this context, 22 articles were analysed to answer the research questions in the study.
Within the scope of the first research question, when the selected studies were examined, it was seen that the studies on AI in education have shown a consistent increase every year, indicating the growing interest in this field. The journals preferred by the AI in the education community for publication were
Based on the second research question, which was about revealing the purpose of AI applications used in an online learning environment, it was concluded that the role of AI in education was divided into six categories, namely chatbot, content, learning assistant, learning incentive, measurement and evaluation, and early warning system (Figure 7). It was observed that AI was most commonly used as a chatbot. Chatbots are used in online education environments for teaching a specific subject, improving students’ skills in language education, answering frequently asked questions, eliminating students’ misconceptions, reducing students’ stress and anxiety levels, improving teachers’ professional skills and improving students’ communication skills. AI is also widely used as a course content and learning assistant. It is used in online education environments for various purposes such as encouraging participants to learn and increase their motivation, experiencing an AI application and using it as an early warning system based on student data.
Six categories of the role of AI in education. AI, artificial intelligence.
Another conclusion reached in this study was about the opinions, perceptions and attitudes of teachers and students towards using AI in an online learning environment (Figure 8). Based on the analysis of the selected studies, students have generally positive views towards AI applications, especially in relation to automated assessment processes, adaptive learning systems and chatbots. Students perceive AI as a tool that enhances learning experiences, increases motivation and engagement, and improves learning outcomes. Students appreciated the ease of use, accessibility and conversational features of AI tools and emphasised the importance of combining human intelligence with AI. However, concerns about limited responses, repetition of responses and privacy issues need to be addressed to ensure a satisfactory AI experience for students. The positive perception of the impact of AI on teaching and learning processes is shared by both students and teachers, emphasising the potential of AI to improve education. The number of AI studies conducted with teachers is limited. However, the analysed studies, albeit limited, give an idea about teachers’ views and perceptions about AI. Most of the teachers think that chatbots should be included in teaching activities. Teachers expressed that they want to use chatbots in educational environments and believe that this will increase the quality of teaching. It has been observed that teachers think similarly as students in this context. However, unlike students, teachers stated that they prefer chatbots with a formal language to chatbots with a more social language.
Teacher and student views, perceptions and attitudes towards AI. AI, artificial intelligence.
Further research is needed to gain a better understanding of the impact of AI on educational processes and to enable the development of effective AI-integrated online learning environments. In this context, experimental studies that can reveal the impact of AI can be conducted. Education stakeholders may be included in the studies and their opinions, attitudes, perceptions and recommendations can be acquired utilising various data collection methods. Especially, addressing the opinions, perceptions and attitudes of parents as well as students and teachers is a necessity in considering the context when analysing the use of AI in education.
This study on the use of AI in education, which encompassed the review of 22 experimental studies and the compilation of findings from a total of 207 studies from reputable databases including Web of Science, ScienceDirect, Taylor and Francis, and Wiley, presents several limitations that warrant acknowledgement. First, although the relevant literature was reviewed from four different databases, it may not cover all existing studies on the subject. Although efforts were made to include recent articles covering the pandemic and post-pandemic period to capture the most recent research in this area, the scope of the study may be affected by the partial inclusion of relevant references. Additionally, this review was limited to studies published in the English language, potentially excluding valuable research conducted in other languages. Second, the study’s authors admit to a limitation in their technological expertise, which may have hindered a comprehensive interpretation of certain studies reporting technological advancements. While they possess knowledge of educational technology applications, their understanding of the intricate design, development and optimisation of AI programmes may be limited. Consequently, theoretical discussions and technical optimisations may remain a focus for future research carried out by experts with more specialised expertise in these areas. Despite these limitations, this study aimed to provide a representative overview of the research landscape, trends and primary application contexts of AI in education, setting the stage for further exploration by researchers with the requisite expertise.