An Ethnographic Study on Teachers’ Acceptance and Resistance Attitudes to Adopting Learning Analytics

Abstract

Learning analytics (LA) is an emerging field that has undergone extensive development in K-12 education, focusing on students’ learning processes and tracking their learning trajectories. However, numerous developments remain in the pilot phase without holistic adoption. Despite the growing use of teacher-facing analytics in K-12 education, little is known about the processes through which instructors interpret and negotiate analytics within everyday school practice. This paper presents an interpretive ethnographic study of teacher sensemaking and resistance regarding LA adoption in K-12 schools. Informed by the unified theory of acceptance and use of technology (UTAUT), we implemented a small-N ethnographic study with school teachers to explore their perceptions and experiences regarding LA adoption. The research question was “How do teachers interpret, negotiate, and make sense of learning analytics adoption in K-12 school settings?” The findings indicate that teachers’ responses to LA adoption were shaped through complex sensemaking processes involving perceived usefulness, institutional culture, emotional ambivalence, ethical issues, and human-centered understandings of data-driven educational practices.

1. Introduction

Learning analytics (LA) is an intersectional field of computer science, design, and pedagogy that focuses on educational data mining (Gasevic et al., 2019). LA may be considered a process in which data are collected from learners, an analysis is performed, and an intervention is determined (van Leeuwen, 2019). Such data include attendance logs, assignment submissions, interaction with educational resources, video engagement metrics, assessment performance, and collaboration activity within learning management systems (LMS) (D. Tzimas & Demetriadis, 2023).

Introducing LA in schools is considered a necessity for social and educational reasons (Tsai et al., 2018). Studies have reported varying levels of skepticism among teachers regarding the pedagogical value and practical implementation of LA-based interventions (Amida et al., 2022; D. E. Tzimas & Demetriadis, 2026). Literature has provided partial views on LA adoption. For instance, Ali et al. (2013) evaluated the adoption process based on users’ perceptions of ease of use and usefulness, without considering other aspects such as pedagogical and institutional characteristics. To the best of our knowledge, only a few studies have provided evidence of the pedagogical challenges for LA adoption from the perspective of school teachers, considering the unified theory of acceptance and use of technology (UTAUT) (Dwivedi et al., 2019). Informed by the UTAUT (Williams et al., 2015), we conducted ethnographic research with five school teachers to investigate their expectations and perceptions of LA adoption. The study adopted an interpretivist ethnographic perspective, focusing on teachers’ situated experiences and interactions regarding LA adoption within the everyday school culture.

2. Background

2.1. Adoption of Learning Analytics

LA interprets and visualizes big data to help teachers optimize learning conditions (Guzmán-Valenzuela et al., 2021). There has been growing interest in selecting evidence-based practices to support instructors in determining effective teaching strategies (Banihashem et al., 2022). Although K-12 education provides a rigorous human-centered learning analytics (HCLA) design context, most of the related research has been conducted in tertiary educational settings (Du et al., 2019; Ifenthaler, 2017; Wiley et al., 2023). Schools are moving to implement teacher-facing analytics, aiming for instructors to extract deep insights into student learning. Recent moves in the field towards HCLA emphasize analytics as part of a socio-technical system in which human factors play as critical a role as technical ones. The question of how instructors make sense of analytics in authentic contexts is receiving increasing attention.

There is evidence of active resistance to LA use in schools. One reason is the tendency of research in the LA area to focus primarily on technological aspects. Consequently, much less attention has been paid to the importance of human agency (Nazaretsky et al., 2022). Misrepresentation of teachers’ needs and challenges in LA design can affect adoption. The paradigm shift proposed by participatory design and co-design (Dimitriadis et al., 2021) involves moving toward design with people as equal collaborators (Buckingham Shum et al., 2019). Dollinger et al. (2019) found that the transparency of researchers’ perspectives and values in underlying decision-making is one of the current gaps in LA research. Finally, the lack of ethnographic studies documenting teacher involvement in the adoption process emphasizes how far LA must go to catch up with HCLA (Dwivedi et al., 2019). The current ethnographic study addresses these gaps by extracting the experiences and sensemaking of five instructors (Li et al., 2021; Viberg et al., 2018).

The limitations of LA and its uptake by educators are under critical scrutiny (Lim et al., 2020). This is in part because of the lack of involvement of teachers in LA tool design (Prestigiacomo et al., 2020). In this paper, we propose a human-centered participatory approach for developing an understanding of teachers’ data needs through the lens of the UTAUT model. We outline teachers’ opinions on the evidence they should use to make better-informed decisions and address the implications of our approach for the future design of human-centered LA.

Social context. Greek secondary public schools have begun to use digital platforms to monitor student participation, assessment trajectories, and early indicators of dropout risk, but the region lacks a community for sharing experiences (D. Tzimas & Demetriadis, 2025). Thus, this ethnographic study focuses on the current perceptions of LA adoption by teachers in the context of Greek school education. Participants represented different levels of teaching experience and employment conditions toward educational technology adoption. The instructional technology settings used by Greek instructors are online and accessible via the Greek Ministry of Education. Specifically, MySchool is an online platform designed to support schools with administrative tasks, such as monitoring early school dropout, while the Electronic Classroom service allows teachers to create interactive online courses. The dashboard provides visualizations of data from the school’s LMS regarding student activities and performance. Recently, the online service eParents was added to inform parents and guardians about their children’s progress at school.

2.2. Acceptance Models for Technology and Learning Analytics

In this study, resistance is considered a multidimensional construct that may include passive reluctance, strategic skepticism, ethical issues, emotional anxiety, workload-related hesitation, and critical professional reflection regarding pedagogical autonomy and human agency. Technologically deterministic approaches might be contrasted with those focusing on sensemaking. Herodotou et al. (2019) stated that teachers’ technology acceptance, as conceptualized by the technology acceptance model (TAM), has an impact on LA adoption. TAM states that a user’s intention to use technology is influenced by perceived usefulness and perceived ease of use. In parallel, the academic resistance model (ARM) and learning analytics acceptance model (LAAM) (Ali et al., 2013) explain adaptation to LA in terms of teachers’ cognitive, emotional, and intentional attitudes. Finally, Rogers (2003) defined the innovation-decision process as the process by which individuals choose to adopt or reject innovation.

In conclusion, we draw from the UTAUT (Dawson et al., 2019; Williams et al., 2015), which is based on TAM and ARM, as a framework providing insights into how teachers perceive LA adoption. Therefore, UTAUT constructs were utilized to inform the design of the ethnographic research (e.g., interview protocol). In the UTAUT, five constructs predict intention to use the innovative technology: performance expectancy, effort expectancy, social influence, self-efficacy, and feelings. Despite the significance of UTAUT constructs in adopting technology, no ethnographic studies have used this framework to understand K-12 teachers’ LA adoption. We view adoption as occurring at the level of individual teachers, who may use LA daily. Specifically, through ethnographic observations and interviews, we examined whether the aforementioned factors could explain teachers’ motivations to adopt LA. Rather than viewing UTAUT constructs as isolated predictors of technology adoption, this study approaches them as socio-culturally mediated axes embedded in teachers’ professional sensemaking processes. Therefore, UTAUT is employed in this study as an interpretive framework rather than a deterministic explanatory model.

2.3. Research Question

We adopt a stakeholder-oriented approach, focusing on teachers’ perceptions of LA adoption in Greek schools. We explored which underlying factors are conducive or obstructive to future use. We started with a theoretically derived (UTAUT) research question about a specific cultural process and found a site (an ethnographic study in K-12 settings) where the following research question might be applicable.

“How do teachers interpret, negotiate, and make sense of learning analytics adoption in K-12 school settings?”

3. Method

3.1. Participants and Context

Using the ethnography (Emerson et al., 2011; Marcus, 1995) method of data collection, we followed the LA topic in Greek secondary schools. The participants included five teaching staff members (three part-time) teaching math, literature, computer science, and economics (content-independent research). Convenience sampling was adopted, i.e., participants were mainly recruited from the school where the researchers were based or known through their networks. We selected various teacher profiles to eliminate bias, ensure representative participation, and gain views from teachers with diverse teaching experiences. The ethnographic study lasted six months. On average, the participants had 7.2 years of teaching experience. All participants provided informed consent and agreed that their data might be used in the study.

Teacher 1 is interested in educational technology and believes that training is essential. Teacher 2 explained that her perspectives on student learning have evolved throughout more than 20 years of teaching experience. However, unlike Teacher 2, Teacher 4, who had only six months of teaching experience in secondary education, viewed technology adoption as a necessity in education. Teacher 3 believed that her actions as a teacher were more effective when she focused on inspiring students rather than providing them with technological artifacts. Finally, Teacher 5 is a computer science teacher with more than 20 years of teaching experience.

3.2. Research Design

We used the ethnographic approach (Emerson et al., 2011; Pink et al., 2016) for internal validity and consistency to analyze the data and interrelate the results (Table 1) (Twining et al., 2017). Specifically, this process consisted of contrasting evidence from stakeholders involved in the study (early adopters and skeptics), different data sources (observations, field notes, and interviews), participant checking (allowing participants to comment on transcripts), and theoretical triangulation (interpreting the data using multiple theoretical frameworks).

Table 1. Summary of key terminology related to the study (based on Twining et al., 2017).

	Level
	Theoretical stance: Epistemology	Meaning is culturally defined
Approach	Methodology	Qualitative (Interpretivist)
	Ethnography Design	Emphasizes inductive reasoning
	Techniques for data collection	Interview; fieldnotes; observation
	Instruments (Specific data collection tools)	Interview schedule; Observation sheets
	Analysis	Phenomenography (How the data is processed to answer the RQ)

3.3. Small-N Design

There is a history of educational research employing idiographic designs known as small-N designs, which consider the individual participant as the replication unit (Barker et al., 2002). We argue that robust and valuable findings could be obtained not by using statistical inference on large samples but by using small-N designs in which a significant number of observations are made on a relatively small population of participants in authentic settings. We argue that, if LA is to become a mature research topic, its fundamental theoretical aim should be to investigate functional relationships as they arise at the individual participant level (Smith & Little, 2018). Finally, small-N designs are a potentially appealing way of blending science and practice because they enable researchers to integrate formal research methods into their daily work.

Following our methodological pluralism stance, we argue that a rigorous investigation of any topic area should combine both large-N and small-N approaches (D. Tzimas & Demetriadis, 2024a). Our research follows an ethnographic approach (Hammersley & Atkinson, 2019; Wolcott, 2008), which is a qualitative data collection method for gathering data often used in the social and behavioral sciences. Data are collected through participant observations and personal, in-depth interviews, which are then used to determine how societies and individuals function (Eriksson & Kovalainen, 2008). Furthermore, the ethnographic study of education combines participatory research methodologies, theoretical engagement, and a richly descriptive genre of writing to depict the everyday complexities of learning in all its forms. Finally, ethnography is useful because it can provide better insight into a specific culture without directly influencing that group (Hammersley & Atkinson, 2019). In this way, information is learned in an unbiased and distinct manner, avoiding the Hawthorne effect (Eriksson & Kovalainen, 2008).

3.4. Ethnographic Study

Ethnographic qualitative research adopts a cultural lens to investigate teachers’ beliefs, values, and practices within their communities. This means that we attempt to examine the culture under study from the inside, through the meanings that the members of that culture live with. Some ethnographers argue that it is just as important to study in more familiar settings (Spradley, 1980). Indeed, you should not assume that you already know other people’s perspectives. This is especially true in large, complex societies; therefore, it is necessary to understand the culture of the group that you are studying before you can explain the actions of its members. Therefore, participant observation (involving direct contact with participants) and open interviews are central to ethnographic research (Spradley, 1980). Conducting ethnographic research means getting to know people (e.g., managers, teachers, students) and acquiring their trust.

The ethnographic study began by choosing a field site. Thereafter, we generated a guiding research question appropriate for that site. Specifically, we tested a pre-defined theoretical model (UTAUT) in K-12 settings. The theory was essential in generating guiding questions for ethnographic investigations, allowing us to address larger questions about how culture works among diverse groups of people. Furthermore, by applying theoretical tools to what we learn in the field, we can better understand social lives in specific research settings. Finally, we needed to find some key participants, that is, knowledgeable agents central in explaining their thoughts, intentions, and actions (Gioia et al., 2013). Our ethnographic fieldwork began with participant observation, which was complemented by interviews. Keeping field notes was a key activity, and everyday events were recorded along with the participants’ viewpoints and interpretations. Overall, we attempted to immerse ourselves in the educational culture by collecting information from different perspectives and capturing participants’ views of their experiences in their own words.

3.5. Measures and Research Instruments

Observations. We shared the everyday lives of the teachers in the chosen settings, working at the study site. This helped us develop an insider’s perspective, meaning that we could experience and feel what it is like to be a member of a group being studied (Hammersley & Atkinson, 2019; Spradley, 1980; Wolcott, 2008). Simultaneously, there was always an observer side to this process. The challenge was to combine participation and observation in a way that enabled an understanding of the site as an insider while describing it to outsiders. Being a participating member of the group gained knowledge that enabled an understanding of the culture under study. We were both outsiders and insiders, remaining on the margins of the group.

Interviews. In addition to observation, an ethnographic study can be based on an open-ended narrative, also known as an ethnographic interview (Hammersley & Atkinson, 2019; Pink et al., 2016). Overall, interviews can be invaluable when choosing a site, after participant observation is completed, or when the participants are experiencing changes that interest the researcher (Spradley, 1980). While participant observation provided information on actions and behaviors, interviews provided a chance to learn how people reflect on their behaviors, circumstances, and events. An important part of the interviews was to establish rapport with the participants. Finally, we developed an interview protocol that included 10 open-ended questions (see Table 2) inspired by the UTAUT. Teachers were asked to share their thoughts and feelings about adopting LA in approximately 45 min (see Table 3 for extracts).

Table 2. Interview questions and their relationships with the UTAUT.

Code	Question
[Q1]	How do you think LA impacts your teaching practice? (Performance expectancy & educational purpose)
[Q2]	How much time are you willing to invest in LA? (Effort expectancy)
[Q3]	Do you need any help in understanding the features of LA? (Self-efficacy)
[Q4]	What feelings does LA provoke in you? Why? (Anxiety)
[Q5]	Would you have any issues regarding its future use in the teaching process? (Concerns)
[Q6]	Do you intend to use LA in the future? Explain. (Behavioral intention)
[Q7]	How should teaching staff conduct students’ data analysis? (Intervention)
[Q8]	What kind of data could help you improve your practice and, consequently, students’ educational experiences? (Teaching needs)
[Q9]	Do you have any ethical or legal concerns related to the use of student data? (Ethics)
[Q10]	Would you have any further information that would be important?

Table 3. Interview results: Qualitative emerging themes and evidence extracts.

2nd-Order Themes	1st-Order Concepts	Evidence Quotes
Performance expectancy	Time and workload management Student evidence-based progress monitoring Differentiated teaching Participation, Visualization Decreased dropout rate Understanding learning Actionability, Decision-making Encouragement, Behavior change Reflection, Awareness Collaborative learning	Teacher 1 (T1): Time management for more sophisticated actions (e.g., monitoring student progress, attendance, and absences). LA is an instrument for differentiated teaching (e.g., “learning path” service in eClass). Alert for disinterested students. To increase participation through game-based learning. Comparing with classmates helps to create an overall picture of yourself and increases creative stress. T2: I am impressed by the use of LA in differentiated teaching. Clustering algorithms will place students in each category, and the workload will be removed from the teacher. I want LA to make decisions and not be content with the level of statistical data. The use of LA does not threaten the role of the teacher. I would like LA to work as analytics in medicine, e.g., embedded devices for people with diabetes. T3: I understand LA as a feedback method. T4: LA use helps understand student performance and learning paths. T5: Visualization helps teachers get an overall picture of the class, avoid missing any details, and identify hidden variables.
Effort expectancy	Training Heavy time & workload Data literacy skills Complexity Interpretation difficulties Early adopters vs. Laggards Guidance	T1: I attend a relevant seminar because it is the future; otherwise, I will have no future in the classroom. T2: Visualization helps create an overall picture (participation, etc.) at a glance to extract gaps for on-time intervention. It is not linked to gender or age. T3: This is a significant work that requires specialized knowledge. Collecting learning material is easy, but analyzing and interpreting it is challenging. T4: I am addicted to adopting new technologies in teaching. There is age differentiation. T5: I would like the data collection platform to be smart and save time for the teacher. I would use LA, as long as it is not time-consuming and laborious.
Feelings	Anxiety, fear Agony, skepticism Discouragement Expectations Satisfaction Enthusiasm	T1: Anxiety initially because it takes me out of my comfort zone, and if I do not manage to use it in the right way. I am concerned that the role of the teacher will be degraded. T2: I am not impressed by the simple recording of the data. I am interested in analyzing them to draw useful insights without human intervention, using only algorithms. Categorizing students can make them feel uncomfortable. I have an agony for understanding LA. T3: Comparisons can be discouraging for a low-performing student. Risk of placing a stamp on the student. Anxiety and fear of trying to replace the role of the teacher. T4: Satisfaction with being involved in teaching development. T5: Monitoring through LA evokes positive emotions in me. Comparison may cause anxiety and discourage students if it is not accompanied by discussion and interpretation with the teacher.
Human-centeredness	Critical thinking Sensemaking Well-being Co-design issues Teacher agency Explainable Learning theories Obligation to act Shared experience	T1: LA is a tool and cannot replace the teacher. The critical thinking abilities of students can only be triggered by the teacher. Will it help me, or is it meaningless? Linking student mood, behavior, and psychology to performance. Encourage students with low self-confidence and those who have been labelled as mediocre or low performers. Unleash students from their assigned profiles (e.g., last-desk students). Find correlations and causes of behavioral changes over time. T2: LA is a tool. The point is for LA to facilitate the learning process. It does not record student concerns unrelated to learning performance and does not detect higher-order critical thinking skills. Have a supporting role and do not substitute for in-class communication, immediacy, and interactivity. It has a greater impact on distance education. Seeing 100 connected students in eClass tells me nothing. Useful, but with concern for future use. T3: The role of the teacher is crucial. LA tools may identify low performance but not the underlying cause, and they do not measure psychological factors. It is a black box for teachers. Data-based decision-making is not always human-centered. T5: Emphasis on the internal motivations of the teacher.
Data culture	Comfort zone, resistance Established practices Parents’ involvement Laggards communities Self-efficacy, data literacy skills Redesign of educational practices Extent of change Positive attitude New roles for teachers	T1: The majority of teachers do not want changes. There are communities, such as laggards or innovators, that socialize with each other and self-affirm. There is age differentiation in the adoption of innovations such as LA. Older people resist. Millennial teachers will adopt more readily. I find no correlation between male and female teachers and full-time and part-time teachers. Seminars will be hosted by the Ministry of Education and external vendors. I insist on training teachers, students, and parents. How will parents adopt LA? I have a positive attitude if it is done in the right way. I will save time (e.g., in one graph, I will see various insights). I give a positive sign with worry. I am attending a relevant seminar. I am worried that I am not doing this correctly. I need training. T2: There is a culture of resistance in Greek schools. Nothing from the outside gets in. Teachers are not adaptable. We do not want changes. I avoid laggard communities. T3: I am optimistic about changes, but with careful steps. The age of the teacher plays a decisive role. It is hard to get out of the comfort zone and into the risk zone. Teachers lack data literacy skills and won’t take risks using a technique with uncertain results. Teachers possess ICT skills but must become familiar with data use and visualizations. They need to practice. T4: There is an established culture. Teachers don’t want to get out of their comfort zone. A multidisciplinary approach is needed. Adoption should be gradual. Otherwise, there will be reactions. T5: After the COVID era, Greek teachers acquired an ICT culture, but they need to be convinced that any educational innovation is worthwhile. Hands-on training is needed. ICT teachers play a key role. There is an age difference. Secondary school teachers would find it easier to adopt LA.
Ethics	Data privacy and ownership Surveillance Accuracy Labeling Algorithm centered Data misuse and consent Trust Hawthorne effect	T1: I am concerned about the use of personal data. T2: Not watching the student like a test subject under a microscope. Consent from parents. It scares me to have a company capture all the learning data. T3: LA tools provide only pictures of the moment. A student may be gaming on the system, e.g., logged in but not watching. Note that the algorithms do not have 100% accuracy. I don’t trust decisions from natural language processing algorithms (e.g., in forums), especially in the ChatGPT era. Parents need to be involved. It reminds me of Big Brother and maybe not authentic student behavior. I don’t want strict supervision and monitoring. T4: Training is needed. Avoid gaming the system. Who uses the data and how? Do we have a new Big Brother? We have privacy issues. Student labeling is problematic. T5: If LA is linked to individual assessment, many teachers will game the system. There is the issue of surveillance of the students’ learning progress.
Social context	Facilitating working conditions Centralized educational system Teacher evaluation and support Communication Professional development Parents Attitudes Change management Technological infrastructure Bottom-up process Comfort zone	T1: Teachers are not willing to invest time in learning LA. In particular, the older ones. They find it difficult. It takes them out of their comfort zone. Young teachers are more willing. I am concerned that LA tools are designed top-down for teachers without input from teachers. How can I use it? Will it help me, or is it useless? I mainly want in-school training from experts and colleagues. T2: LA is a lonely ranger process. It should be designed based on the requirements of teachers and students. T3: As a part-time teacher, I change schools every year. If I had received reports and visualizations in mid-September, it would have affected me negatively because of the labelling. Not to lose face-to-face teacher-parent communication. The metrics of LMSs don’t tell me anything. How we interpret LA plays an important role. T4: In Greece, we do not have a culture of adopting new technologies and innovations in education. We don’t want a top-down approach from the Ministry of Education. Training is required to avoid the misuse of LA. Maybe teacher evaluation will help, and it will also provide an external incentive for professional development. I used LA in the COVID era. Interdisciplinary experts (lawyers, psychologists) are needed. Parents need to be actively involved. Therefore, parents take responsibility. Training is also needed for parents to interpret the data. T5: A transition period and maturity are required. I believe in gradual in-school training in real-life conditions. This would provide evidence for teacher-parent discussions and student assessment.

Field notes. Field notes are a key ethnographic research tool because of the subjectivity of the ethnographer (Emerson et al., 2011; Hammersley & Atkinson, 2019). We wrote field notes during our stay in the field. Specifically, phrases were written down at the field site. Descriptions were recorded about what we could remember about a specific event. We also analyzed what we learned in the setting about our research question, focused our observations on the next visit, and reflected on what we had learned while making observations.

3.6. Data Collection and Analysis

In ethnographic research, analysis occurs throughout the research project and is tightly connected to interpretation. Interpretive approaches seek themes of meaning in the data, based not only on the number of times an idea is mentioned but also on its significance to participants (patterns, connections, or contrastive points). This means that, during our research process, we continuously analyzed and learned from our empirical data. The data collection period was between January and June 2024. Participation in the interviews was voluntary and anonymous, so the privacy of opinions was ensured. A 45 min interview was conducted with each instructor to obtain both retrospective and real-time accounts of the participating teachers experiencing the phenomenon. The interview initially focused on the instructors’ backgrounds and overall motivation to use the analytics. Then, the interview focused on how the participants learned to use LA and any institutional support during this process. Table 2 presents the guiding questions that were posed in order, along with more specific prompts.

The interviews were analyzed using the phenomenography research approach, which investigates qualitatively varied ways in which people experience certain things (Marton, 1981). Each newly identified issue was filed as a theme (pattern), and any further references to the same issue were added to the same pattern. The interviews were transcribed, helping us identify significant quotes using the interviewees’ exact words and producing 32 significant key phrases, 18 concepts, and 13 categories, resulting in seven major themes. During the analysis, a bottom-up and top-down approach was combined using raw data and a theoretical lens (UTAUT) to identify relevant codes and themes. To strengthen interpretive consistency, both authors discussed interpretations until conceptual agreement was reached. Emphasis was placed on credibility, triangulation, and interpretive depth rather than statistical inter-coder reliability measures. When we have a complete set of first-order concepts and second-order aggregate themes, we can build a data structure (see Table 3). Therefore, we focus on building an inductive model grounded in data that captures teachers’ needs.

4. Results

4.1. Interview Results

We selected meaningful extracts related to the research question and literature. Combining themes and refining their semantics resulted in the seven major themes presented in Table 3. We named the first-order concepts, which represented the participant-centric terms directly obtained from the interviews, and the second-order themes, which represented the theory-centric concepts covering various influencing factors for adopting LA in schools.

4.2. Observation and Field Notes Results

During this phase of the study, the researcher participated in the everyday school culture and recorded observations focusing on teachers’ interactions, practices, and informal discussions regarding LA. Field observations revealed that teachers did not approach LA as a purely technical innovation but rather as a socio-technical and pedagogical intervention associated with emotions, professional identity, and concerns regarding student support.

Observed discussions among teachers frequently focused on the contribution of LA to student attendance monitoring, participation, and the early identification of dropout risk. Several participants described analytics tools as potentially useful for supporting vulnerable students. Simultaneously, field observations revealed significant ambivalence toward the presence of data-driven practices in the school environment. During informal staffroom discussions following professional development activities, some teachers expressed enthusiasm regarding the future possibilities of analytics-supported interventions, whereas others responded with humor or skepticism, describing analytics initiatives as another reform that will disappear in a few years. These interactions reflected broader tensions between innovation-oriented and resistant professional subcultures.

Teacher practices regarding LA adoption also differed considerably. T2 appeared interested in using analytics for attendance monitoring and intervention planning related to dropout prevention. T3 demonstrated cautious and reflective engagement, frequently expressing concerns regarding the increasing administrative and monitoring dimensions of analytics use. Conversely, T4 actively demonstrated analytics dashboards to colleagues during informal conversations in the teachers’ office and appeared eager to experiment with attendance alerts and participation monitoring tools. T5 primarily used LA functionalities for communication and early-alert purposes related to absenteeism and student engagement.

Generational differences were also visible during everyday interactions among participants. Younger teachers generally approached analytics tools more pragmatically and with greater confidence, while more experienced educators expressed uncertainty regarding data interpretation. Overall, field observations suggested that teachers perceived LA simultaneously as a potentially valuable support mechanism and source of professional pressure.

5. Discussion

This ethnographic study contributes to our understanding of how teachers make sense of analytics. Teachers’ acceptance or resistance toward LA was not determined solely by technological usability or perceived usefulness but was dynamically shaped through socio-cultural sense-making processes embedded within the everyday school life. Based on Table 3, the following discussion integrates the findings and interprets their relevance to the current literature in terms of the following axes.

Effort expectancy and Feelings. The perceived complexity of an innovation indicates how difficult it seems to comprehend the innovation, which is negatively related to the innovation’s rate of adoption. Teachers simultaneously perceived LA as pedagogically promising and professionally threatening, revealing a tension between performance expectancy and concerns regarding surveillance and algorithmic fairness. Workload, utility, and time were challenges used to weigh the usefulness of LA compared to the effort needed (Amida et al., 2022).

The perceived compatibility of an innovation refers to how consistent it is with an individual’s values, experiences, and needs. The most prevalent positive emotions (psychological factors) are motivation, satisfaction, and expectations, whereas the most common negative emotions are anxiety and skepticism. A comparison of the observations and interview data sources reveals similarities throughout the dataset. Enthusiasm and satisfaction are two consistent concepts. To explore contradictions, some teachers expressed dissatisfaction, agony, and skepticism (the ‘anxiety’ item shown in Table 3). Next, we argue that four additional factors not captured by UTAUT and found to explain further LA adoption are human-centeredness, data culture, ethics, and social context.

Data culture and Ethics. Teachers need to accept the need for change and adapt their teaching, idiosyncrasies, and established practices to the needs of another culture. Moreover, we observed the importance of the cultural factor overtaking the data factor (Falcão et al., 2020). This complex socio-cognitive process can be described as a negotiation between two cultures (D. Tzimas & Demetriadis, 2024b). When participants were asked what they required to pursue their LA goals, there were three significant replies: clarity, training, and time. Numerous teachers indicated that self-efficacy and data literacy, including skills for reporting data, were significant barriers to adopting LA (Li et al., 2021).

Teachers’ comments raised ethical concerns, such as the obligation to act, data privacy, algorithmic bias, and surveillance (D. Tzimas & Demetriadis, 2021). The risk of excessive control over instructors’ work was mentioned, representing a threat to their autonomy. Furthermore, the instructors argued that they should maintain their right to autonomy and should not be obliged to intervene in every specific situation that the data might indicate. The instructors believed that their workload would increase and that some would resist change, opting to remain in their comfort zone (Falcão et al., 2020). Finally, a common understanding of the LA domain has not yet been established among educators. LA causes a transition from human agency to human and algorithmic agency.

Performance expectancy and Social context. The perceived relative advantage of an innovative intervention is the degree to which it is perceived as evolving from a previous innovation. Reflecting on the results of the ethnographic interviews, field notes, and observations, most teachers recognized LA’s usefulness. The monitoring concept emerged from teachers’ feedback for providing effective intervention (‘duty to act’ and ‘understanding learning’ items shown in Table 3) (Herodotou et al., 2021). Furthermore, the teachers believe that LA enhances students’ metacognitive competencies (‘effectiveness of the intervention’ item shown in Table 3). In contrast, some teachers claimed that the data failed to capture actionable aspects of learning (Tsai et al., 2018).

Teachers are interested in using LA to improve their professional profiles and social influence (Tsai et al., 2018). In addition, individualized and intensive professional development for various stakeholders is part of the introduction of any new technology. Furthermore, interpersonal communication channels are essential for spreading LA adoption through participatory design (‘shared experience’ item shown in Table 3) (Mavroudi et al., 2021). While facilitating conditions could be in place in the form of training, these conditions would not enable some teachers to engage with LA. Teachers mentioned a lack of training in data competence and mentoring from educational leadership (Amida et al., 2022) (‘needs’ item shown in Table 3). The interviews identified cost (time required for doing LA) as an important factor in their motivation to adopt LA; the observations and data fields confirmed this. Field observations further demonstrated that teachers’ responses to LA were socially negotiated within the everyday school culture. Informal staffroom discussions frequently revealed tensions between innovation-oriented and skeptical professional orientations.

Human-centered LA. Understanding teachers’ beliefs and constraints through inter-stakeholder communication leads to their agentic positioning (Dimitriadis et al., 2021). Unlike technologically deterministic views, teachers did not see LA as a unique agent of change. Instead, we identified agency in educators’ sensemaking about data (Cho & Wayman, 2014). Although we may want to reduce cognitive effort for users in other domains, this is not always desirable in learning. A broader view of human-centeredness could include designing for and evaluating qualities beyond the usability and efficiency of performance, such as well-being (Dimitriadis et al., 2021). Furthermore, the question of how instructors make sense of analytics is receiving increasing attention, along with the growing recognition that it is nontrivial for instructors to extract explainable and actionable insights from data (Barreiros et al., 2023). Furthermore, teachers emphasized the need for professional autonomy, interpretive flexibility, and pedagogical contextualization when engaging with analytics systems. The findings therefore support that sustainable LA adoption in K-12 education depends not only on technological infrastructure but also on trust and emotional acceptance.

Overall, participants frequently perceived analytics simultaneously as pedagogically promising and professionally disruptive. On the one hand, LA was associated with opportunities for identifying disengaged students. On the other hand, teachers expressed concerns regarding surveillance and emotional pressure. These tensions suggest that teacher acceptance of LA cannot be fully understood through linear adoption models alone but must also be interpreted as a complex process of professional and ethical negotiation.

6. Conclusions

In line with the UTAUT and studies on technology adoption (Herodotou et al., 2021), we identified the following factors facilitating LA adoption: (a) the performance benefits of LA (participation and differentiated teaching), (b) feelings (expectations and satisfaction), (c) social context (communication and professional development), and (d) human-centeredness (co-design and teacher agency). In antithesis, among the factors inhibiting adoption were (a) effort expectancy (complexity and workload), (b) ethical issues (privacy and surveillance), and (c) data culture (resistance and comfort zone). Although the UTAUT framework provided a useful interpretive structure for examining teachers’ perceptions regarding LA adoption, the findings suggest that the relationships among its dimensions are dynamic, context-dependent, and socio-culturally mediated. Performance expectancy, effort expectancy, social influence, and facilitating conditions were not experienced as isolated variables but were negotiated through teachers’ professional identities and ethical concerns.

Discussing contradictions among the participants, young teachers provided more positive responses and expectations concerning readiness for LA adoption. A reason for the replies could be the higher levels of ICT and data literacy skills of millennial teachers compared with those of older teachers. In this regard, Aldunate and Nussbaum (2013) found that early adopters who invest time in incorporating technology into their teaching practice are more likely to adopt new technology and contribute as change agents. Overall, the teaching staff needs timely feedback, data literacy skills, resources, and support to use LA with actionability, explainability, and anthropocentrism.

Study Limitations and Future Directions

This paper’s contributions to theory and practice include increasing the understanding of the factors influencing teachers’ adoption of LA and investigating the key constructs of UTAUT. We acknowledge the limitations to be considered. The sample size was small (ethnography), and the data covered a specific setting for a particular community of teachers. Therefore, caution is necessary when scaling the findings to other populations because cultural values and social contexts affect teachers’ perceptions of LA’s applicability.

Furthermore, participants may behave differently than usual when they are aware that they are being studied. As a result, the selected methods need to counteract this effect because the authenticity of cultural experiences is important for the use of ethnographic methodologies.

The purpose of the study was not statistical generalization but contextual and interpretive depth within an ethnographic setting. Following the logic of small-N qualitative inquiry, emphasis was placed on rich description and interpretive understanding rather than sample representativeness. Future research should expand beyond teacher perspectives and include additional stakeholders, such as school leaders, policymakers, parents, and LA designers, to better understand the socio-technical ecosystem surrounding LA adoption. Similar studies should be conducted in other types of schools and post-secondary education, and in different countries to explore further themes related to LA adoption. Finally, future research should investigate multi-sited ethnography to gain a comprehensive understanding of the adoption process and its viability.