Behind the Scenes of Adaptive Learning: A Scoping Review of Teachers’ Perspectives on the Use of Adaptive Learning Technologies

Abstract

Adaptive learning technologies are emerging educational tools that personalize instruction and tasks to meet students’ current ability levels. Despite their increasing prevalence, the present discourse on adaptive learning software focuses on the technical aspects of adaptive learning systems without accounting for other factors that impact their use in classrooms. In acknowledgment of the importance of teachers’ roles in the implementation of educational tools, this review examined studies about teachers’ perspectives on the use of adaptive learning technologies in their courses. A systematic search was conducted in four databases (ERIC, Academic Search Complete, Google Scholar, and Taylor Francis Online). Twenty-nine full-text articles were retained for analysis. A hybrid of deductive and inductive approaches was applied to analyze the data, using the factors described in the Technology Acceptance Model and its extensions as a guide. The findings showed that teachers who saw the long-term benefits of using adaptive learning technologies were more willing to invest time and effort to learn and use the tools in their classes. Barriers identified were the scant professional development opportunities for teachers, the programs’ complexities, and the demands on teachers’ time. To make adaptive learning programs work, teachers must be on board as co-designers and content creators, and they require ongoing technical and institutional support and sufficient preparation in teacher education programs.

1. Introduction

Adaptive learning is a pedagogical method that emphasizes a learner-centered approach in education. In this approach, teachers deliver custom learning experiences for students in the form of personalized and well-timed feedback and differentiated content and instruction [1]. In principle, adaptive learning acknowledges the different abilities of individual learners and caters to these differences by providing appropriate scaffolding to enable them to progress at their own pace [2]. If properly executed, adaptive learning as an approach can enhance learning outcomes and help manage students’ cognitive load through adapting the level of complexity of presented learning content and tasks [3].

As education has always relied on the prevailing technologies of a particular era to enhance teaching and learning [4], adaptive learning platforms were developed to meet the learning demands of students with the above principles in mind. Adaptive learning platform is defined as “an emerging learning technology that dynamically adjusts instructional content to provide interactive and personalized learning paths to the individual to facilitate learning” [5]. Ideally, the learning activities found in the programs should match the students’ current capacities [6]. Thus, it is crucial for the system to be able to make a personalized assessment of a student’s readiness before they become exposed to more difficult content and tasks [7]. Adaptive learning platforms can produce customized learning paths, either on their own for some systems, or through teacher intervention [8]. Given these definitions, adaptive learning tools can include affective intelligent tutoring systems (ATS) that personalize learning using data from learners’ affective states [9].

1.1. Effectiveness of Adaptive Learning Programs

Adaptive learning programs cater to learners across various educational levels—primary education [10], secondary education [11], higher education [8], and medical education [12]. There have been reports on the effectiveness of adaptive learning systems in improving learning outcomes. For instance, a study among eighth-grade students in China revealed that the group who utilized an adaptive learning system (Squirrel AI Learning) had greater gains on a mathematics test compared with those who went through two common instructional approaches: large-group and small-group classroom instruction [13]. Adaptive learning technology also showed potential in closing outcome gaps among diverse learners in the U.S., thus helping students from minority backgrounds perform better [14]. Used in medical learning, adaptive learning technology also assisted in improving medical trainees’ skills in diagnosis [15]. Additionally, technology-supported personalized learning was revealed in a meta-analysis to have a moderately positive effect size on learning among youth from low-and middle-income countries [16].

1.2. Importance of Gathering Teachers’ Perspectives on the Use of Adaptive Learning Technologies

Despite emerging evidence of their effectiveness, wide-scale implementation [17] and the actual use of adaptive learning programs in courses [18] are reported to be low. This may be due to barriers and challenges with adaptive learning systems pointed out by researchers, such as the split between education and technology [19]. The split between education and technology is demonstrated by disparities between those who have access to technology and know how to use it and those who do not [20]. This phenomenon, also referred to as the digital divide in education, has been a longstanding issue that needs to be addressed through information campaigns and systematic research. Thus, to increase knowledge and understanding, it is important to lend a voice to teachers who are important stakeholders and implementers of teaching innovations. Obtaining teachers’ perspectives is a step towards raising awareness and identifying benefits and barriers to the successful implementation of adaptive learning technologies.

Adaptive learning technology has been gaining much attention, as evidenced by recent reviews [5,19,21]. Research that discusses the importance of developing digital competencies among teachers is abundant [22,23,24], but little attention has been paid to other factors that impact teachers’ technology use from a teacher’s point of view. Specifically, no attempt has been made to synthesize teachers’ views on adaptive learning programs. As teachers do not only possess content knowledge but also pedagogical knowledge, their help is often solicited in the conceptualization and design of adaptive learning programs [25]. Their views about adaptive systems, specifically on how these programs fit into current teaching practices, are essential. Thus, the current review gathers teachers’ perspectives on the use of adaptive learning platforms. In the following section, we discuss the framework that guided this review.

1.3. Guiding Framework: Technology Acceptance Model and Its Extensions

The Technology Acceptance Model or TAM [26,27] is a framework used to study the factors involved in the use and adoption of technology. The model points to the perceived ease of use of the system and its perceived usefulness as the main factors that affect a person’s intention to use new technology. However, studies that adopted this framework pointed out that the TAM was not sufficient in explaining how people decide whether to adopt a particular technology or not. For instance, while many advocate for more teacher training to increase people’s perceived ease of use of technology [28,29], a meta-analysis found perceived ease of use as the weakest factor in determining both the intention toward e-learning and actual usage [30].

An extended TAM was proposed that included technology self-efficacy, subjective norms, and facilitating conditions as factors in technology acceptance [31]. Technology self-efficacy is defined as a teacher’s belief in their abilities to make use of educational technologies in teaching their classes. Subjective norms are defined as the tendency for students and teachers to accept or reject a technology based on how significant others view it as. Facilitating conditions capture elements needed for successful program implementation and increased acceptance and positive perceptions of educational technologies among teachers. Perceived ease of use and perceived usefulness, factors that influence technology acceptance in the original TAM, can be considered as part of the overarching theme of facilitating conditions, as recommended by researchers [31].

Another extension of the TAM suggested to include not only a technology acceptance perspective but also a value perspective [32]. The value-based adoption model (VAM) considers the user’s perceived value in adopting new technology beyond perceived ease of use and perceived usefulness. In the VAM, perceived ease of use and perceived usefulness are combined into one factor called perceived benefits. Beyond perceived benefits, the VAM proposed the inclusion of a factor called perceived sacrifices. Applications of new technology entail investments and incur both monetary and non-monetary costs such as time and effort, and these were often not accounted for in discussions of the original TAM. Accounting for teachers’ perceived sacrifices is important so that these can be weighed against the benefits and value that teachers perceive in using educational technologies.

The models described above—the original TAM and its proposed extensions [31,32]—suggest that teachers’ perspectives on the use of adaptive learning systems are likely to be influenced by technology self-efficacy, subjective norms, facilitating conditions, and perceived sacrifices. Therefore, these factors formed the analytical framework to guide this review.

1.4. The Present Review

The current discourse on adaptive learning, particularly in the computer science discipline, focuses on how to enhance the capacities of adaptive learning systems to teach students. Research focused on the technical aspects of educational technologies overlooks the fact that teachers have important roles to play, as they are the ones who make instructional decisions. As classroom managers, teachers are also tasked with facilitating the academic and socio-emotional learning of students [33]. In addition, teachers are involved in course design and development and its implementation and evaluation. A large-scale survey of teachers’ plans to use an adaptive learning program the following school year found that 86% of the teachers indicated an intent to use it more often [34]. However, the rationale behind their reasons for wanting to adopt the technology was not provided. It has been noted that individual, environmental, and technological factors could result in the underutilization of technology for teaching [31]. Therefore, examining teachers’ perspectives on adaptive learning technology might offer insights into the probable reasons for resistance towards its use. Such an investigation could also reveal the perceived benefits of those who have already implemented this technology in their classes.

Adaptive learning articles from computer science, engineering, and the sciences are biased toward the technical aspects of adaptive learning [5]. Also, adaptive learning aims to be learner-centered, but this focus has failed to consider teachers’ perspectives about the technology. This observation is echoed in a systematic review of research on artificial intelligence applications in higher education [35]. Conducting a scoping review of how educators perceive adaptive learning is a step towards narrowing the current perceived divide between educational science and computer science [19].

Martin et al. [5] pointed out that nearly half of all adaptive learning studies written in English are from Taiwan or the US and involve higher education. Xie et al. [21] likewise highlighted the dominance of higher education contexts in adaptive learning technology research. Therefore, it is necessary to have a more global approach and to be more inclusive of other instructional settings (e.g., K-12 classrooms) that are understudied in adaptive learning literature.

The analytical framework discussed earlier in the introduction based on the TAM [26,27] and its extensions [31,32] was applied to this review. A previous study found that teachers’ characteristics, such as their pedagogical beliefs, their self-efficacy, and their attitudes, influence their use of educational technology [36]. The current study likewise adopts a more person-centered approach and an emic viewpoint [37] by examining teachers’ perspectives on the use of adaptive learning programs.

2. Method

2.1. Research Design

A scoping review was conducted on teachers’ perspectives on adaptive learning technologies, following the PRISMA guidelines for scoping reviews [38]. The scoping review also follows a more flexible search procedure since the goal is to identify all relevant literature. It is useful in identifying gaps and potential areas for further research.

2.2. Procedure

2.2.1. Identifying the Research Question

This scoping review aimed to collate data on teachers’ perspectives on adaptive learning technologies. Through the review, we aimed to find answers to the following research questions using TAM and its proposed extensions [31,32] as the guiding framework:

(1) What are teachers’ perspectives regarding their self-efficacy in using adaptive learning technologies?

(2) What are teachers’ perspectives about the impact of their use of the adaptive learning technologies on students’ and colleagues’ normative beliefs and adoption behavior (subjective norms)?

(3) What are teachers’ perspectives about facilitating conditions in their acceptance of adaptive learning technologies (e.g., perceived benefits), corresponding technology barriers and related challenges encountered in using the tools?

(4) What are teachers’ perceived sacrifices (i.e., the cost of using adaptive learning technologies)?

2.2.2. Identifying Relevant Studies and the Inclusion Criteria

A systematic search was initially conducted on the empirical studies using the following search strings: “adaptive learning” OR “personalized learning” OR “individualized learning” AND “teacher perspective” OR “teacher perception”. The following databases were used in the search: ERIC, Academic Search Complete (EBSCOhost Integrated Research), Google Scholar, and Taylor Francis Online. The following inclusion criteria were applied in the search:

• Empirical studies with full text only;
• On any educational level;
• Published between 1970 and 2022;
• Journal articles published in English or theses or dissertations or conference proceedings in English;
• Must be about teachers’ perspectives on the use of adaptive learning technologies

The year 1970 was set as a starting point in our initial search since this is the year when work on adaptive learning technologies began [39]. However, the first 10 articles that resulted from this initial search and that met the inclusion criteria were all from the year 2018 and up, demonstrating the scarcity of studies on adaptive learning that were teacher focused. Thus, the team decided that 2018 will be the starting point of the search.

Adhering to the PRISMA guidelines for scoping reviews [38], we present the electronic search strategy for four databases (ERIC, EBSCOhost Academic Search Complete, Google Scholar, and Taylor and Francis Online).

A sequential three-stage process (identification, screening, and inclusion) was implemented. After the initial random search, search terms were further refined to include other terms and the following terms were used: “adaptive learning system” OR “adaptive learning program” OR “adaptive learning technology” AND “teacher perspective” OR “teacher perception”. A keyword search for “adaptive” within the full-text articles was performed. A similar process was performed using the keyword “personalized”. If the articles with the “personalized” keyword had adaptive elements, the articles were retained for in-depth, full-text review. While personalized learning and adaptive learning are often used interchangeably, in the sense that both involve customization of a student’s learning experience, the latter term is mainly used to refer to an approach that relies heavily on technology (algorithms and data analytics) to personalize learning in real time [40].

The identified articles were organized in a shared Google Drive folder. The team conducted a screening process using our inclusion criteria to select the articles relevant to the review. After this, full texts for screened articles were downloaded and further assessed for eligibility based on the inclusion criteria.

Despite having passed initial screening through keyword search, full-text articles were removed from the review due to the following reasons: (1) the study did not have teachers as participants (e.g., AI subject matter experts, system designers); (2) the study had multiple stakeholders as respondents and answers from teachers could not be distinguished from the rest; (3) the study had a different focus and not on adaptive learning specifically (e.g., online language learning, educational technologies, open educational resources); (4) the study’s participants talked about personalized learning as a teaching philosophy but did not mention any particular adaptive learning technology (e.g., teachers in the study expressed general opinions on the role of educational technologies); (5) the study’s participants were applying traditional methods of teaching and were merely evaluating a design model; (6) the study was a duplicated record retrieved from another database.

2.2.3. Data Analysis

Thematic analysis [41] was used to analyze the articles that met the inclusion criteria. The coding process followed a hybrid of deductive and inductive approaches, where open coding (inductive) was first applied to 10 articles. The codes derived from the 10 articles were sent to other researchers in the team for checking. The initial codes were condensed into themes that fit well with factors described in the TAM and its extensions. Once themes had been finalized, a deductive approach was then applied to the analysis of the rest of the articles using the analytical framework summarized above.

In addition to the analysis described, the descriptive information of each article was coded including (1) year of publication, (2) publication source (journal/thesis/dissertation/conference proceeding), (3) research design (quantitative, qualitative, mixed), (4) number of participating teachers per study, (5) regions where the studies are conducted, (6) subjects/courses where the adaptive learning technologies were used, (7) educational levels of students catered for by the adaptive software evaluated.

3. Results

3.1. Description of the Articles Identified

In total, 6434 records were identified using a systematic search with refined keywords. These articles were screened further if the full text contained the keywords “adaptive” and “personalized” and if they appeared to meet the inclusion criteria set. Forty-nine records were assessed for eligibility through in-depth reading of the full text. Twenty records were excluded in this step due to the reasons mentioned above. Twenty-nine articles met the inclusion criteria for this scoping review. The screening and selection procedure is illustrated in the PRISMA flow diagram (see Figure 1).

The year with the most publications on teachers’ perceptions of adaptive learning platforms was 2020 with 12 studies (41.38%), followed by the years 2019 and 2021 which had 6 articles (20.69%) each. Twenty-two studies (75.86%) were published in peer-reviewed journals. There were 15 (51.72%) qualitative studies, 6 (20.69%) quantitative studies, and 8 (27.59%) mixed-method studies. Among the latter eight, three (10.34%) employed participatory design.

Most of the studies were conducted in Europe (44.83%) and North America (41.38%), and some studies were from multiple locations (6.90%). There were also platforms discussed in the articles that were used for more than one subject, but the most common was for Mathematics (37.93%). The total number of participating teachers for all of the included studies was 1790. See

Table 1. Characteristics of each study included in the review.

Author/s	Year Published	Article Type	Research Design	No. of Participating Teachers Per Study	Locations Where the Studies Were Conducted	Subjects the Tools Were Used in	Educational Level of Students Catered by the Program Evaluated
Adiguzel et al. [42]	2019	Peer-reviewed journal	Mixed	4	Turkey	Calculus I and II, Physics I and II, and General Chemistry I	Higher education
2. Amro and Borup [43]	2019	Peer-reviewed journal	Qualitative	11	Not specified	English	Elementary
3. An and Mindrila [44]	2020	Peer-reviewed journal	Mixed	125	United States of America	Not specified	Not specified
4. Daruwala et al. [45]	2021	Peer-reviewed journal	Qualitative	Not specified	United States of America	Mathematics	K-8 schools (elementary-middle school)
5. Dollinger et al. [46]	2019	Peer-reviewed journal	Qualitative	20	Australia	Various subjects	Higher education
6. Duren et al. [47]	2021	Peer-reviewed journal	Mixed	1	Not specified	Not specified	Higher education
7. Germeroth et al. [48]	2018	Peer-reviewed journal	Quantitative	21	United States of America	English	Kindergarten to third grade
8. Guven and Cakir [49]	2020	Peer-reviewed journal	Qualitative	6	Not specified	Not specified	Primary school
9. Haas [50]	2020	Peer-reviewed journal	Qualitative	5	United States of America	Social studies, English, Arts	Not specified
10. Holstein et al. [51]	2019	Peer-reviewed journal	Mixed	40	United States of America	Mathematics	Middle school
11. Holstein et al. [52]	2019	Conference proceeding	Qualitative	10	United States of America	Not specified	K-12
12. Istiyono et al. [53]	2020	Peer-reviewed journal	Quantitative	10	Indonesia	Physics	Grade 10
13. Johnson and Zone [54]	2018	Peer-reviewed journal	Qualitative	Not specified	United States of America	Mathematics and English	Not specified
14. Knorr and Berkling [55]	2020	Conference proceeding	Qualitative	8	Germany	Language	Grades 1–3
15. Krumm et al. [56]	2021	Peer-reviewed journal	Qualitative	14	United States of America	Mathematics	Grade 3
16. Liu [57]	2020	Thesis	Qualitative	27	China	Various subjects	Secondary Education
17. Liu [58]	2022	Peer-reviewed journal	Mixed	2	Taiwan	Mathematics	Elementary to Middle School Students
18. Martin et al. [5]	2020	Peer-reviewed journal	Quantitative	247	United States of America	Various subjects	Not specified
19. Moltudal et al. [10]	2022	Peer-reviewed journal	Qualitative	3	Norway	Mathematics	Grades 5–7
20. Nagro [59]	2021	Peer-reviewed journal	Quantitative	406	Saudi Arabia	Various subjects	Not specified
21. Niaki et al. [60]	2019	Conference proceeding	Quantitative	Not specified	United States of America	Mathematics	Middle School and High School
22. Perumal et al. [61]	2020	Peer-reviewed journal	Quantitative	253	Malaysia	Not specified	Not specified
23. Pfeiffer et al. [62]	2021	Conference proceeding	Mixed	64	Not specified	Not specified	Not specified
24. Shahbazi [63]	2020	Peer-reviewed journal	Qualitative	3	Canada	English	Secondary Education
25. Sheromova et al. [64]	2020	Peer-reviewed journal	Mixed	Not specified	Russia	Mathematics	Secondary Education
26. Smith [65]	2018	Peer-reviewed journal	Qualitative	17	United States of America	Mathematics	Not specified
27. Standen et al. [66]	2021	Conference proceeding	Qualitative	21	United Kingdom, Italy, and Spain	Mathematics	Special Education
28. Tenório et al. [67]	2020	Conference proceeding	Qualitative	15	Brazil	Not specified	Secondary and Post-Secondary Education
29. Whitelock et al. [68]	2020	Peer-reviewed journal	Mixed	457	Spain, Netherlands, United Kingdom, Turkey, Finland, and Bulgaria	Not specified	Higher Education

for more detailed characteristics of each study included in the review.

The findings from the studies reviewed were organized under themes based on the study’s framework described in an earlier section of the paper. The findings below are divided into four main sections, with the section on facilitating conditions in technology acceptance, technology barriers, and related challenges having multiple sub-sections under it.

3.2. Technology Self-Efficacy

Teachers expressed difficulty in using the adaptive learning programs and admitted to having low technology self-efficacy [62]. The reasons for this include the lack of a solid technological background and how the concept of adaptivity was not a part of either their pre- or in-service teacher training. Teachers emphasized the continued support faculty members needed throughout the stages of technology adoption [54,63,66,68]. A common source of skepticism for teachers was their unfamiliarity with the software. Thus, after getting to know the software’s affordances, teachers gained a better appreciation for the program and became more confident in using it [43]. Familiarity led to increased use [47], as teachers were found to be increasingly likely to adjust their instruction over time as they became more familiar with an online platform [60]. In one study, teachers who integrated the adaptive system into their teaching practice reported high levels of self-efficacy with the technology but still wanted training for how to maximize its use [54].

3.3. Subjective Norms

Students are likely to be more engaged with the programs and to use them more often if their teachers invest in the use of the technologies themselves. This perception was reflected in a study involving three interactive digital tools in first-year foundational courses at the Faculty of Engineering. Teachers interviewed thought that “if the use of the product was modeled in the lectures on a continuous basis demonstrating the value of various product features and learning resources, students might have been more willing to follow that model after the lectures” [42]. In another study, teachers perceived it as their responsibility to orient students to the adaptive learning software, troubleshoot technical issues, motivate students to fully engage with the software, monitor students’ behavior and personalized learning in the system, and provide students with additional instruction [43]. All of these were expected to improve students’ capacity to effectively use the software.

The English language teachers interviewed in Amro and Borup’s [43] study also lamented the mixed responses of their fellow teachers to the adaptive learning software. They attributed the skepticism of some of the teachers to these teachers’ lack of awareness of the software’s potential to provide students with personalized learning experiences. One of the teachers reported that the previously skeptical teachers had a better appreciation for the program once the software’s affordances had been demonstrated to them.

The responses of teachers in the study by Dollinger et al. [46] were more positive. In this article, usage encounters were described as key to having more teachers adopt the adaptive learning program in their classes. Two steps in particular, co-maintenance (which involved teachers providing ad hoc support and training to colleagues) and co-promotion (which involved teachers sharing SES practices and success stories with colleagues), were identified as the main reasons for the increased uptake of the program across universities.

What differed between the studies of Amro and Borup [43] and Dollinger et al. [46] is that in the latter teachers were involved in the conception and design of the software being evaluated. To quote one teacher’s comment, “for me, the biggest selling point about SRES is that it’s made by a teacher. […] I can trust that it will do what I want it to do because a teacher built it and a teacher would think the way that I think” [46].

3.4. Facilitating Conditions in Technology Acceptance, Technology Barriers, and Related Challenges

3.4.1. The Need for Adequate Institutional Structures and Infrastructure Support

Maximizing institutional structures and infrastructure was noted as a prerequisite for the successful implementation of adaptive learning programs based on teachers’ responses [54]. This includes good leadership, investment in faculty, commitment to continuously test the efficacy of the use of adaptive learning in courses and an institution-wide recognition of the benefits of adaptive learning for students. Schools must be willing to invest in the initial set-up costs and maintenance of such a complex system and the equipment required to implement it, as teachers reported having worked under precarious conditions and expressed a desire for more infrastructure support [49,66]. As teachers needed to regularly troubleshoot technical issues, having a technical support team with the sole function of training and assisting teachers in fixing the glitches in the system was also imperative [43,46].

The need for adequate infrastructure is highlighted in the hardware issues and network-related issues reported by teachers in the application of adaptive learning programs in their courses [43,49,57,63,66]. A lack of internet access made it challenging for teachers and students to access the program. Particularly in China, K-12 education is still primarily offline, and students are not allowed to use electronic devices in schools [57]. This is problematic as AI-based education platforms rely on students’ interaction with the platform to collect data that would help provide a personalized learning experience for students. Due to the mentioned limitations, teachers had to take on the additional task of uploading students’ data after exams, and this becomes the sole source of learning behavior data.

3.4.2. Teachers’ Perceived Role as Co-Designers and Content Developers and the Importance of Curriculum Alignment

Faculty tend to be more accepting of adaptive learning technology when they participate in the development and revision of course content, as they can gain a sense of ownership in the process [54]. Control over instructional and curricular activities increases the likelihood of faculty adopting the technology. Teachers are assigned to work with instructional designers and often tapped as subject matter experts [54,58]. The importance of teachers being closely involved in co-designing was evident in the successful implementation of one learning analytics program adopted at five Australian universities [46]. The formation of a centralized team (involving teachers) led to many improvements in the functionality of the system, and teacher involvement in the participatory design of the program also resulted in increased buy-in and support for the platform.

The teachers in the studies reviewed often emphasized the importance of aligning the activities in the adaptive learning programs with the course’s objectives. Structured and guided assignments for students that match the content covered in lectures facilitated technology acceptance among teachers [42,57]. A program for high school English language learners was perceived by teachers to be appropriate for students’ learning needs as it had a library of appropriate readings, gave immediate learner-centered feedback, and included assessment tasks that targeted students’ strengths and weaknesses [63].

While programs were treated as a supplement to what was being learned by the students in the classroom [68], teachers often reported a misalignment between the content covered in class and the practice problems found in the adaptive learning software. Teachers noted how some activities in AI-based education platforms were not suitable for all levels of students [49,57]. Platforms recommending learning materials with content beyond students’ comprehension and course syllabus results in student frustration [57]. Programs giving out personalized content suited to students’ abilities became problematic in a case study of a three-year implementation program [45]. State standardized assessments naturally test for grade-level content, but unfamiliarity with grade-level material deflated students’ confidence, increased their anxiety, and had a negative impact on their test scores. This further brings to light the importance of aligning the activities in the adaptive learning programs with the course’s objectives and intended learning outcomes.

3.4.3. Teachers’ Roles Beyond Designing and Developing Content and Their Reported Difficulties

In addition to co-designing and developing content, teachers perceive orienting the students to the functions of adaptive technology as one of their roles. However, as one study has shown, it was particularly challenging for teachers to address each student’s technological needs due to them coming from diverse educational and cultural backgrounds [43]. One teacher shared that some of her students were so inexperienced with technology she had to show them how to do something as simple as clicking the mouse. Other roles reported by teachers are monitoring students’ learning and behavior in the program, motivating the students to engage and providing students with additional instruction individually or in groups depending on need. Without teacher supervision, students tend to guess and develop poor work habits when using adaptive learning software. For instance, the students are often unable to demonstrate their thinking process even if they get the correct answers. Thus, teachers should be able to intervene based on the real-time feedback provided by monitoring tools built into the program [51]. Visualization of the percentage of students interacting with the software is perceived by the teachers to be important, as they believe this informs their decision of whether to proceed to the next topic or to provide intervention to help students reach their goals [67]. If the goal is to increase student engagement with the platform, reports noted that structured guidance from teachers on the benefits of using the digital product is imperative [42]. While most of the teachers received initial training prior to implementation, they did not perceive the training to be sufficient for adoption [43,49]. The teachers perceived adaptive learning programs as hard to use, harder to accommodate than others, cumbersome, and impractical.

3.5. Other Facilitating Conditions

One of the ways to help students who were encountering problems in navigating the platform is by recruiting successful and knowledgeable students to assist them [43]. These students were called the “computer people”, “learning monitors” [43], or “digital leader” [42]. Even in classes for students with learning disabilities and autism, some children who were more advanced than others were designated as “helpers” (Standen et al. 2021, p. 446).

Teachers perceived gamification within the program as one strategy that is useful in supporting student engagement [10,43,55]. A system’s ability to give hands-on experience and its usefulness in helping students understand the concepts that they need to apply in their homework were likewise perceived as facilitators of student engagement [54]. Teachers shared that built-in reward tools from within the software motivated the students by providing them with reward points and achievement certificates that they could print and hand out to the students [43]. However, others cautioned against the use of reward systems (e.g., digital trophies and points), since they perceive that the true reward for students should be the learning that they gain from the program.

3.5.1. Different Impacts of Class Sizes from Teachers’ Perspectives

There were different views from teachers in terms of whether having small classes or large classes facilitate the use of adaptive learning programs. For a truly personalized learning experience, visualization of each student’s interaction with the software through learning analytics was claimed by teachers to be only ideal for small classes, if data from the program are going to be useful [67]. To further demonstrate how class size impacts program adoption, Engineering Math and Science teachers from one study reported that due to increasing enrollment numbers, they were overwhelmed with the number of students in their classes. Thus, they focused their efforts on delivering the lecture and grading the exams and did not include digital products in their lectures due to a lack of time [42]. This reflects teachers’ reservations about class-level implementation of adaptive learning programs, especially in large classes [66].

In contrast, adaptive learning technologies are perceived by other teachers as able to augment teacher abilities in large classes [46]. There is evidence of how teachers found automatic grading important, especially with high student enrollment [42]. Furthermore, the summary report in some of the programs clarified the performance of individuals and groups of students and aided teachers in making decisions about their classes [56].

3.5.2. Perceived Benefits for Teachers

Teachers see several benefits in using adaptive learning platforms. One study that explored a collaborative approach to sharing learner event data from both instruction- and assessment-focused technologies found that reports from the program were easily understood by teachers and were easy to use for assigning student groups [56]. The skills descriptions provided by the program made the process of developing instructional interventions easy. Likewise, teachers who used a new technology application for reading that used music to teach fluency shared that they found the platform relevant to their work and easy to use [48]. This made them inclined to recommend its use to a colleague.

Studies also reported how teachers appreciated getting real-time feedback and being able to identify students in need of teacher help [43,51,52,54,58,59]. Teachers found the option provided to students for invisible hand raising especially useful for shy students who would like to ask questions anonymously. Monitoring tools were utilized by teachers to check on failed skills more than mastered skills since they help the teachers plan their small group or one-on-one interventions. There were also mechanisms in place that alerted teachers to student frustration, misbehavior, or idleness [43]. Additionally, teachers appreciated receiving positive information on students, since it served as motivation for them and helped them identify students who were not being challenged enough by the software [51,55].

The automatic grading feature of platforms and summary reports produced were also lauded by teachers, as they allowed students extended practice and saved them time [56,57]. Students’ descriptive data being presented through graphs rather than text was preferred by teachers, since graphs were found to be more enjoyable, useful, and relevant [67]. Graphs were seen as a convenient way to present students’ progress to parents and caregivers [66].

Adaptive learning programs also promote equality and inclusion through their use in teaching children with special needs. An adaptive learning system based on multimodal affect recognition was developed for students with learning disabilities and autism [66]. One of the major benefits of using such a system with special needs children is its capacity to unearth previously undetected abilities. Despite this benefit, researchers found that technologies cannot provide the emotional closeness offered by teachers. Even in a system designed to utilize machine learning to make predictions and inferences, teachers were not convinced that an algorithm can take the place of a teacher’s ability to feel the child’s emotions and the sensitivity of human eyes [66,68]. Eye gaze trackers for those with autism, for instance, were not perceived to be the most accurate way to assess the affective state of a child with autism.

3.5.3. Perceived Benefits for Students

Teachers also reported several benefits for students who use adaptive learning tools. There are programs that provide students with the option to pick an alternative path through the content, to attempt new content, or to practice previous concepts [54]. As reported by primary education teachers, their students enjoy not having to write and they simply wait for tasks to appear [10]. Teachers’ positive perceptions of adaptive learning programs in the reports include an improvement in student motivation and work ethic while using the program [63]. Furthermore, teachers perceive as a benefit the development of students’ digital literacy skills and sense of achievement and enjoyment for language from their interaction with adaptive learning programs [55]. Among children on the autism spectrum, a teacher noted that the repetition of learning material afforded by the program is very useful [66]. They also used the program to control disruptive behavior, by giving students more tasks to do in the system when they misbehave, with the added benefit for students to learn more in the process.

There are AI-based education platforms that are widely applied in a region/country, housing materials and resources collated from educators from various institutions [57]. Teachers perceive this kind of platform to be beneficial for students as it allows them to access more learning resources from different regions. Teachers believe that the programs are useful for giving extra work for students who need to be challenged and for struggling students who need some practice [43,57]. Since learning analytics is an inherent feature of most adaptive learning platforms, they can be useful in addressing the learning needs of students who are falling behind their classmates. A case study found evidence for one platform’s usefulness for low academic performing students, by allowing teachers to design personalized assessment tasks depending on students’ learning conditions [58]. Programs that make use of artificial intelligence also promote self-learning, provide detailed feedback, and guide students in areas they need to improve on [59]. One study found that the most influential motivator for teachers to integrate digital technology into their courses is perceiving the use of technology as beneficial to student learning [69].

3.5.4. Perceived Sacrifices

Teachers recognize that the benefits of using adaptive learning programs come with a cost. Teachers have busy schedules and heavy work demands, and they report that extensive use of adaptive learning programs in courses takes up a significant amount of their time. Teachers in the articles reviewed expressed needing time to reflect and plan, as using the program involved rethinking their lesson plans and allotting time to solve organizational issues [43,66]. While learning analytics was considered helpful and useful, many teachers admitted to not having been able to maximize the data provided by monitoring tools due to a lack of time and the complexity of the program’s features [43,47,68].

Special education teachers likewise saw the time investment required in developing learning materials as daunting, but the task became less overwhelming with the thought that the materials can be reused [66]. Teachers who believe in the benefits of adaptive learning technologies dedicate more time to navigating the software and in building connections between the curriculum and content within the adaptive learning tool [63]. Teachers are also aware of the limitations in using adaptive learning software, such as limiting the interactions among students and between the students and their teachers [47]. A potential downside of the increased reliance on the program for teaching is that less time is dedicated to building relationships with students and getting to know them. Teachers also shared that use of the program limits students’ opportunities for verbalizing their understanding of the subject matter, which is a valuable part of learning.

The following discussion summarizes the findings from the reviewed studies, focusing on practical implications and recommendations.

4. Discussion

This scoping review aimed to summarize findings from the literature on teachers’ perspectives on the use of adaptive learning technologies. We employed an extension of the Technology Acceptance Model or TAM [26] to organize the extracted data from the articles reviewed. The framework includes technology self-efficacy, subjective norms, and facilitating conditions in the implementation of adaptive learning technologies [31]. In addition, findings were also interpreted through the value-based adoption model that adopts a value perspective in weighing the perceived benefits and perceived sacrifices in technology adoption [32]. In this section, we focus on the practical implications and recommendations gleaned from the review.

4.1. Recommendations for Teacher Professional Development and for Improving Teachers’ Technology Self-Efficacy and Adoption

Despite emerging evidence on the effectiveness of adaptive learning programs in improving student outcomes [13,15,16] and the perceived benefits and value shared by the teachers in the reviewed studies, the wide-scale implementation [17] and actual use of adaptive learning programs in courses [18] are reported to be low. Teachers identified support structures to be prerequisites for them to feel more confident in the use of the tools. There were also several challenges shared by teachers, such as having to work under precarious conditions, being unfamiliar with the software, and experiencing a lack of technical support. Thus, to increase technology self-efficacy and integration in courses, teachers need sufficient training and support. Training teachers on how to best utilize adaptive learning technologies can be part of teacher education programs. The focus of teacher training should not only be on the technology’s functionality but also on how to optimize its use in teaching. These teacher training programs should target developing teachers’ appreciation of the technology before they get to implement it in their classes. Testimonies and demonstrations from colleagues who have successfully integrated the adaptive program in their courses would be helpful in this regard, especially since there is evidence that teachers were more likely to adopt the program after hearing success stories from fellow teachers [46]. Student engagement with the adaptive tools, likewise, was observed to be affected by teachers’ interest and engagement with the program. Thus, encouraging teachers to adopt a growth mindset, openness, and flexibility in using the programs in their courses could also motivate students to use these learning tools.

Pre-service training gives teachers a good baseline and helps with program familiarization. However, ongoing support from a technical team in the process of technology adoption would be most ideal, since issues encountered in the use of a new platform can easily discourage teachers from continuing to use them [44]. Emerging adaptive learning programs have unique features and constraints than more traditional educational technology tools [10], thus specialized training for teachers is needed for them to be adept at using them. This is important as familiarity with the program turned out to be a factor that increases teachers’ use of the program later [47,60].

4.2. Recommendations to Improve Adaptive Learning Technologies

Teachers also shared their insights on how to improve existing adaptive learning tools. Involving teachers in the design of the technology, in the creation of content, and in the evaluation of the program lead to many benefits and improvements in the system [46,54]. Evidence in this review suggests that teachers taking on the role of co-designers and co-creators of the programs had a positive effect on fellow teachers and led to increased uptake across universities [46].

While teachers often adapted their practice to a given software, it would also be good to be able to adapt a software to a teacher’s practice. Ideally, adaptive learning technologies should be able to make teaching easier and more effective. Thus, software designers must be aware of the various technological issues and hurdles shared by teachers in this review. Improving the ease of navigation, learning analytics, summary reports, feedback functions, and automatic grading features of the programs could unburden teachers and allow them to focus on engaging and building quality relationships with students. Especially since the latter was shown to be important in influencing teaching quality ratings and students’ learning experiences [70].

4.3. Curriculum Alignment: Personalized vs. Grade-Level Content

Teachers pointed out that it was vital that the adaptive program’s contents aligned with the curriculum and course objectives. Among the many challenges, matching personalized content in the program with what the students are supposed to learn at a specific grade level (grade-level content) is a teacher concern worth highlighting. The adaptivity of a system is what makes it a learner-centered tool, as it adjusts recommended exercises to match the ability of a student and to meet them where they are. While this is consistent with a personalized learning philosophy, teachers pointed out that this raises problems during high stakes standardized assessments. Students tend to take exams unprepared for grade-level content. Thus, there is an important decision to be made as to whether to adapt the system to align it with curriculum requirements or to adjust grade-level expectations to respect each student’s unique pace in learning and their individuality.

4.4. Factors That Impact Adaptive Learning Technology Acceptance from Teachers’ Perspectives

The findings confirmed the propositions in extensions of the TAM [31,32]. The results implied that factors influencing the intention to use new technology go beyond the perceived ease of use and usefulness identified in the original TAM. Only a few teachers defined the adaptive learning programs as easy to use and discussed benefits more in terms of their perceived usefulness in aiding classroom instruction and in assisting students with achieving their goals. Technology self-efficacy was quite low among teachers but could be improved through consistent technical support provided to teachers in the process of technology adoption. Subjective norms also played a role, since teachers’ colleagues and their students’ engagement with the platform was found to be contingent upon their own appreciation of the value of the adaptive learning technology being introduced. Likewise, the promotion of the system through success stories from colleagues greatly contributed to the increased use of a particular program among teachers [46]. Consistent with the value perspective [32], teachers were willing to invest time to learn the program when they see its value. They are also more likely to participate in the design and creation of course content on the platform when they recognize its potential benefit for student learning and for the improvement of current classroom practices. The importance of collaboration among different stakeholders was highlighted, with teachers who used to be behind the scenes taking center stage through participatory design and the creation of course content within the adaptive tools that are aligned with learning objectives.

4.5. Synthesis

Taken together, evidence from the reviewed studies emphasizes the importance of advocating for more professional development opportunities for teachers in adaptive learning technologies (Evans, 2021; Kim et al., 2013). Innovative professional development programs (e.g., virtual coaching model; see [71]) should allow teachers to keep updated with evolving educational technologies and to increase their digital learning competencies. Evidence suggests that the development of teachers’ skills is crucial for enabling student learning activities involving digital technology in schools [72].

Overall, the findings of this review suggest that teachers are willing to make the required effort to learn and implement the technology despite the many challenges if they perceive the adaptive learning tools as valuable and beneficial in the long run. To maximize the learning feedback provided by built-in monitoring tools in the programs, teachers must invest valuable time in learning program features and navigating the platform. Investments include redesigning courses and curriculum content to match the software’s functions or participating in the design process and content creation so that the program is aligned with what students are supposed to learn. The willingness of teachers to embrace the use of an adaptive learning software depends on them perceiving the attached costs and difficulties to be worth what they and their students get in return. For administrators, their willingness to invest resources in the required structure and infrastructure support can encourage teachers to integrate adaptive learning programs into their classes.

4.6. Limitations and Future Directions

In this section, we describe some limitations of this study and suggest ways on how future research can address these. First, while the initial search included the search string “individualized learning”, the subsequent keyword search only involved searching for the keyword “personalized” within the full-text articles. Researchers can improve upon this by considering all possible iterations that would best capture the target articles in future adaptive learning reviews. Second, given that we only focused on teachers’ perspectives on adaptive learning technology, future reviews may explore the perceptions and experiences of students in using this innovative tool. Students are also end-users and are therefore credible sources on how to improve the current systems. Third, teachers acknowledged the importance of having institutional support for successful program implementation and sustainability of usage. Therefore, reviews examining administrators’ insights and opinions can inform planning for institution-wide implementation of the programs. Fourth, only articles written in English were scoped. While the finding that most of the literature reviewed was conducted in Europe and North America (mainly in the United States) is informative, there is not enough evidence to conclude that there are more Western than Eastern studies on adaptive learning. There could be studies in Asia, for instance, that were written in a different language (e.g., Chinese, Japanese, or Korean). Thus, it is recommended that future reviews be more encompassing to include empirical studies in languages other than English. Fifth, the studies included were not all from peer-reviewed journals and did not go through risk of bias assessment. In the future, when more studies are conducted on the topic, researchers can be more selective in the articles they include. Lastly, many of the studies included in this review had data that were collected from teachers at only one point in time. Therefore, we recommend that more longitudinal (e.g., effectiveness) studies be conducted to better capture and monitor the course of a specific adaptive program’s implementation. The findings from studies that had a participatory design imply that as adaptive learning technologies evolve, so do teachers’ perceptions. Including teachers as co-designers and co-creators in participatory research may then be one of the best options for building knowledge about evolving technologies such as adaptive learning programs.

5. Conclusions

Developing effective training programs for teachers requires an understanding of their perceptions about adaptive learning technologies. This study put forward useful suggestions from the perspective of educators, hoping to provide good value for the future large-scale introduction of AI-based education platforms in schools and universities. For successful implementation, teachers pointed out the need to have more professional training, more support, and more access to technology infrastructure. Given the highlighted challenges, it is important to set realistic expectations for usage sustainability. To create more value, it is crucial to involve teachers in a participatory design process in collaboration with various stakeholders. Teachers and administrators who see the potential long-term benefits of using adaptive learning technologies acknowledge that the return of investment could take time. The willingness to invest time and effort depends on teachers’ perceived value of the tool in improving their instructional practices and in benefiting their students’ learning.