INTRODUCTION
There are arguments for and against the role of teacher motivation, professional identity and passion for teaching (Iliya and Ifeoma, 2015). Drawing on self-determination theory, scholars suggest that teaching itself often fulfils educators’ needs for autonomy, competence, and relatedness, thereby sustaining commitment even under challenging conditions (Collie et al., 2016; Ryan & Deci, 2017). Instructional interactions, such as explaining concepts, responding to students, and witnessing learning progress, are consistently associated with positive emotions and professional fulfilment (Frenzel et al., 2016).
From this perspective, teaching passion functions as a protective factor. Studies indicate that intrinsic motivation for teaching can buffer against
568 Ndlovu and Iyobosa burnout and emotional exhaustion, particularly when educators perceive their instructional decisions as meaningful and impactful (Skaalvik & Skaalvik, 2018). Proponents of this view argue that teachers’ continued enthusiasm for teaching demonstrates professional resilience rather than contradiction, with passion enabling educators to cope with less rewarding aspects of their role (Beltman et al., 2016).
In contrast, critical theorists contend that educators’ reluctance toward marking is not an individual attitudinal issue but a rational response to the expansion and performativity of assessment practices. For example, (Ball, 2016; Perryman & Calvert, 2020) contend that marking has increasingly been framed as a mechanism for accountability rather than learning. Hardy, (2018), further contest that marking is frequently described as excessive, repetitive, and only weakly connected to pedagogical improvement, particularly when feedback is produced primarily to satisfy managerial or inspection demands. Teachers report that marking encroaches on personal time, diminishes work-life balance, and contributes to emotional exhaustion (Allen et al., 2018; Worth & Van den Brande, 2020). From this perspective, reluctance toward marking is interpreted as a form of professional resistance, reflecting educators’ critique of assessment practices that undermine their autonomy and distort the purposes of education (Benesch, 2017).
Critics also argue that framing marking reluctance as an individual deficit risks obscuring systemic responsibility. The problem, they suggest, lies not in assessment itself but in how assessment has been operationalised within high- stakes accountability systems (Torrance, 2017). Researchers argue that when assessment is meaningfully integrated into teaching, emphasising feedback dialogue, peer assessment, and reduced written marking, educators report greater coherence between instruction and evaluation (Boud et al., 2018; Carless & Winstone, 2023). These approaches challenge the inevitability of marking reluctance, suggesting that dissatisfaction emerges not from assessment per se, but from unsustainable assessment designs.
Taken together, the literature reveals that educators’ passion for teaching and reluctance toward marking are not contradictory but structurally produced experiences shaped by contemporary conditions of practice. The tension between these experiences reflects broader systemic pressures rather than individual inconsistency. Importantly, emerging work on sustainable and formative assessment suggests that this divide is not inevitable, but contingent upon how assessment is conceptualised, valued, and operationalised within educational systems. This study builds on these insights by empirically examining how teaching passion and assessment reluctance coexist and interact, and how this relationship shapes educators’ well-being.
Marking, Teacher Wellbeing, and Burnout Burnout is commonly conceptualised through three dimensions, emotional exhaustion, depersonalisation, and reduced personal accomplishment, and recent empirical studies consistently associate excessive assessment demands with all three components (Skaalvik & Skaalvik, 2018; Maslach & Leiter, 2016). From this perspective, marking is situated within organisational
Teaching With Passion, Assessing With Reluctance 569 conditions that expose teachers to sustained overload, limited recovery time, and diminished professional autonomy.
Workload-focused studies show that assessment and marking are among the most persistent contributors to work intensification, often extending beyond formal working hours and eroding work-life boundaries (Allen et al., 2018; Worth & Van den Brande, 2020). These demands are intensified within accountability-driven systems, where marking is required not only to support learning but also to evidence progress, compliance, and performance. Scholars drawing on performativity theory argue that such assessment regimes transform marking into high-stakes emotional and cognitive labour, increasing the likelihood of exhaustion and disengagement (Ball, 2016; Perryman & Calvert, 2020).
In addition to workload volume, research on emotional labour highlights marking as a source of moral distress. When teachers are required to enact assessment practices, they perceive as pedagogically ineffective or ethically misaligned with their professional values, marking can generate frustration, cynicism, and emotional withdrawal (Benesch, 2017; Richards, 2020). Over time, this values-based conflict is theorised as a pathway into depersonalisation, where educators emotionally distance themselves from assessment work as a coping strategy rather than disengaging from teaching more broadly.
The other side of the debate has “Burnout scholars” who contest that emotional exhaustion and disengagement arise from interacting stressors, including leadership quality, role conflict, classroom behaviour, and broader organisational climate, not workload alone (Maslach & Leiter, 2016). From this view, marking represents one component of a wider constellation of demands rather than a primary driver.
Assessment-focused researchers further argue that marking can enhance teacher efficacy and well-being when it is meaningfully aligned with pedagogy. Formative assessment and feedback practices are shown to support teachers’ sense of professional accomplishment, particularly when feedback results in observable student progress (Boud et al., 2018; Carless & Winstone, 2023). In such contexts, marking is not inherently draining but can be experienced as affirming and purposeful. Additionally, literature on teacher competence and assessment literacy suggests that professional knowledge moderates the relationship between marking and wellbeing. Teachers who feel confident in assessment design and feedback strategies report lower stress levels and greater control over marking processes (DeLuca et al., 2019). Proponents of this position argue that burnout risks stem less from marking itself and more from inadequate support, unclear expectations, and poorly designed assessment systems. Consequently, professional development is presented as a mechanism for mitigating negative wellbeing outcomes without reducing assessment demands.
One would therefore argue that the relationship between marking and teacher burnout is neither linear nor inevitable. On one hand, a substantial evidence base positions marking-related workload as a structural risk factor for burnout, particularly emotional exhaustion and depersonalisation, due to its cumulative, externally imposed, and performative character. From
570 Ndlovu and Iyobosa this standpoint, marking drains wellbeing not simply because it is time- consuming, but because it is embedded within accountability systems that restrict professional autonomy and generate moral dissonance. On the other hand, alternative perspectives caution against attributing burnout primarily to marking, emphasising that assessment can support professional fulfilment when it is pedagogically coherent, well-supported, and aligned with teachers’ expertise. These accounts suggest that marking becomes damaging primarily when systemic conditions distort its purpose. Taken together, the literature implies that burnout associated with marking reflects structural dysfunction rather than individual failure, reinforcing the need to examine how assessment workload is organised, valued, and regulated within contemporary education systems.
Recent developments in AI have intensified debates surrounding assessment workload by introducing technologies capable of automating elements of marking and feedback. Consequently, AI has emerged as both a potential solution to assessment burden and a source of new professional and ethical tensions.
Reframing the Transformative Narrative of AI in Education Artificial Intelligence (AI) is widely positioned as a transformative force in education, reshaping pedagogical design, delivery, and assessment practices. Contemporary scholarship highlights how AI-driven technologies, such as adaptive learning systems, intelligent tutoring platforms, and automated assessment tools, enable personalised learning pathways, real-time feedback, and data-informed decision-making (Holmes et al., 2022; OECD, 2021). From this perspective, AI is framed as an enabler of efficiency, scalability, and learner-centred pedagogy, aligning with broader discourses of digital innovation and educational reform.
However, this transformation narrative is not without contestation. Critics argue that the discourse surrounding AI in education is often technologically deterministic, overstating its capacity to enhance learning outcomes while underplaying the social, institutional, and pedagogical complexities of education (Hrastinski et al., 2019). While organisations such as UNESCO (2023) emphasise AI’s potential to democratise access and personalise instruction, they also caution against uncritical adoption, particularly in contexts where inequalities in digital infrastructure and access persist. Thus, rather than a neutral innovation, AI should be understood as socio- technically embedded, shaped by institutional priorities, policy frameworks, and power dynamics.
Furthermore, the shift from teacher-centred to AI-mediated environments reflects not simply pedagogical evolution but a reconfiguration of epistemic authority, where knowledge production and validation increasingly involve algorithmic systems. This raises critical questions about whose knowledge is prioritised, how learning is quantified, and the extent to which AI may reinforce dominant norms through embedded biases.
Teaching With Passion, Assessing With Reluctance 571 Reconfiguration of Teacher Professional Identity: Tension and Transformation
Emerging scholarship suggests that AI is not merely augmenting teaching practice but reconstituting teacher professional identity itself (Shao and Sun, 2025). Teacher identity understood as a dynamic construct shaped by values, beliefs, and professional roles is increasingly destabilised within AI-mediated environments.
On one hand, AI offers opportunities for professional enhancement through access to rich instructional resources, analytics-driven insights, and innovative teaching tools (Xiaoyong, Zirou and Xiaohong, 2024). This aligns with a ‘transformative augmentation’ perspective, where AI complements pedagogical expertise and enables more responsive teaching.
On the other hand, a growing body of literature foregrounds a ‘crisis discourse’ surrounding teacher identity (Lianyu et al., 2024). Educators report experiences of marginalisation, technological anxiety, and diminished professional autonomy, particularly when AI systems assume functions traditionally associated with teacher expertise, such as knowledge delivery and assessment (Lebeničnik and Istenič, 2024; Conklin et al., 2024). This tension reflects a broader deprofessionalisation debate, where the expansion of automated systems risks reducing teaching to technocratic functions.
Moreover, the claim that AI challenges teacher authority warrants critical scrutiny. While access to AI-enabled information may reduce reliance on teachers as primary knowledge sources (Shao and Sun, 2025), this does not necessarily diminish their role. Instead, it suggests a shift from knowledge transmission to knowledge mediation, where teachers act as facilitators of critical thinking, ethical reasoning, and epistemic judgement. Thus, rather than eroding authority, AI may be reconfiguring it into more complex, relational, and interpretive forms.
Theoretical Framing: Identity, Emotional Labour, and AI-Mediated Work The integration of Teacher Professional Identity and Emotional Labour Theory provides a robust framework for understanding the complex implications of AI in education.
Teacher identity is increasingly shaped by technological, institutional, and epistemic shifts, requiring educators to renegotiate their professional roles. Simultaneously, emotional labour theory (Hochschild, 1983) foregrounds the affective demands of teaching, particularly the need to manage emotions in contexts of uncertainty and change.
AI intensifies these dynamics by introducing new tensions between:
- Efficiency and care
- Automation and autonomy
- Innovation and professional values These tensions require teachers to engage in heightened emotional regulation, particularly when navigating fears of replacement, loss of authority, or ethical concerns about AI use. Thus, AI’s impact extends beyond technical transformation to encompass identity reconstruction and emotional strain.
572 Ndlovu and Iyobosa Critically, this intersection reveals that AI adoption is not merely a pedagogical or technological issue but a deeply human and relational process, shaped by meaning-making, resistance, and adaptation. Teachers are not passive recipients of AI innovation but active agents who interpret, negotiate, and sometimes resist its implications.
FINDINGS
A multiple regression analysis was conducted to examine whether background information, ethical considerations, passion for teaching (PFT), and attitudes toward assessment (ATA) predicted perceptions of AI in assessment (PAI). The overall model was statistically significant, F(4,52) = 10.743, p < .001, explaining 45.2% of the variance in PAI (R² = .452, Adjusted R² = .410). PFT (β = .352, p = .007), ATA (β = .350, p = .020), and background information (β = .311, p = .023) were significant positive predictors of PAI. Ethical considerations were a significant negative predictor (β = –.332, p = .019), indicating that greater ethical concerns were associated with lower levels of PAI. No multicollinearity issues were detected, with all VIF values below 2.1.
DISCUSSION
The present study examined whether background information, ethical considerations, passion for teaching (PFT), and attitudes toward assessment (ATA) predicted educators’ perceptions of artificial intelligence in assessment (PAI). The regression model was statistically significant, explaining 45.2% of the variance in perceptions of AI in assessment. This finding indicates that these factors collectively play a substantial role in shaping educators’ attitudes towards AI-enabled assessment practices.
Passion for teaching emerged as a significant positive predictor of PAI, suggesting that educators with greater enthusiasm and commitment to teaching are more likely to view AI favourably. One plausible explanation is that intrinsically motivated educators perceive AI as a supportive tool that enhances instructional effectiveness by reducing administrative burdens
Teaching With Passion, Assessing With Reluctance 573 associated with assessment and feedback. This finding aligns with existing literature indicating that educators with strong intrinsic motivation are more receptive to innovations that enable them to focus on meaningful pedagogical engagement. Moreover, it challenges narratives of technological displacement, instead supporting the view that AI can augment rather than replace teacher expertise (Shao and Sun, 2025).
Attitudes toward assessment also significantly predicted perceptions of AI in assessment. Educators with more positive views of assessment were more likely to hold favourable perceptions of AI. This relationship suggests that individuals who value assessment processes may be more open to technological solutions that enhance efficiency, consistency, and the quality of feedback. Given that assessment is often associated with workload intensification and emotional strain, AI may be perceived as a mechanism for streamlining assessment-related tasks while maintaining, or potentially improving, educational quality.
Background information (demographic and experiential factors) was another significant positive predictor of AI perceptions. This finding suggests that familiarity with AI concepts and applications contributes to greater acceptance of AI in educational assessment. Educators with stronger background knowledge, as well as those more recently entering the higher education sector, may possess a clearer understanding of both the opportunities and limitations of AI technologies. Such familiarity can reduce uncertainty and enhance confidence in their use. This result is consistent with prior research highlighting technological literacy and professional competence as key determinants of technology adoption in educational contexts.
In contrast, ethical considerations emerged as a significant negative predictor of PAI. Educators expressing greater concern about ethical issues were less likely to report favourable perceptions of AI. This finding reflects ongoing debates surrounding algorithmic bias, transparency, accountability, privacy, and fairness in AI-driven educational systems. While educators may recognise the potential benefits of AI, ethical concerns appear to temper their willingness to fully embrace its use. This result reinforces arguments that successful AI implementation requires not only technological effectiveness but also robust ethical governance frameworks that address issues of trust, transparency, and professional responsibility. Furthermore, the negative association may reflect the emotional labour involved in navigating uncertainties around AI adoption, as educators balance efficiency gains with their commitment to fairness, transparency, and student-centred assessment practices.
Collectively, these findings highlight the complex and multidimensional nature of educators’ perceptions of AI in assessment. Positive perceptions appear to be driven by pedagogical commitment, constructive attitudes towards assessment, and technological understanding, while ethical concerns act as a significant constraint on acceptance. The results suggest that institutional efforts to promote AI adoption should extend beyond technical training to include explicit engagement with ethical considerations, supported by clear policies, transparency mechanisms, and targeted professional development.
574 Ndlovu and Iyobosa Findings indicate that while AI can reduce certain burdens associated with assessment, it simultaneously introduces new professional, ethical, and cognitive challenges, offering critical insights for developing sustainable teaching practices and informing institutional support strategies in an AI-enhanced educational environment.
CONCLUSION
The findings indicate that educators’ perceptions of AI in assessment are neither uniformly supportive nor resistant. Rather, these perceptions reflect a balance between perceived pedagogical benefits and ethical concerns. While AI is viewed as a potentially valuable tool for enhancing assessment efficiency and reducing workload, concerns relating to fairness, transparency, and accountability remain significant barriers to its widespread acceptance. These findings carry important implications for higher education institutions seeking to integrate AI into assessment practices. Strategies to support AI adoption should prioritise the development of educator competence through training, assessment literacy, and increased awareness of AI applications. Equally important is the need to address ethical concerns through transparent governance frameworks and responsible implementation strategies that foster trust and accountability.
Future research should extend this work by examining these relationships in larger and more diverse samples, as well as exploring how educators’ perceptions evolve as AI technologies become further embedded in educational practice. The findings challenge the assumption that educators who care deeply about assessment are inherently resistant to technological innovation. Instead, educators with more positive attitudes towards assessment may be more inclined to view AI as a tool that reduces administrative burden while preserving, and potentially enhancing, the pedagogical value of assessment practices. The results also support a human-centred approach to AI adoption in education. Rather than positioning AI as a replacement for professional judgement, the findings suggest that acceptance is highest when AI is perceived as augmenting educators’ work supporting their practices while maintaining human oversight and ethical responsibility.
Overall, this study contributes to the growing body of literature on AI in education by demonstrating that acceptance of AI in assessment is shaped not only by technological familiarity but also by educators’ professional values, beliefs about assessment, and ethical considerations.
Implications for Practice
- Institutions should provide AI literacy training.
- Ethical guidelines should accompany AI implementation.
- AI should support rather than replace educator judgement.
- Assessment redesign should focus on reducing marking burden.
- Educators should be involved in AI governance decisions.
ACKNOWLEDGMENT
The authors would like to acknowledge Mrs Sitha Ndlovu and St. Charles.
REFERENCES
Allen, R., Jerrim, J., & Sims, S. (2018). How did the 2008 recession impact teachers’ workload and wellbeing? Educational Evaluation and Policy Analysis, 40(1), 5–30. https://doi.org/10.3102/0162373717724447 Ball, S. J. (2016). Education, justice and democracy: The struggle over ignorance and
opportunity. Polity Press. Beltman, S., Mansfield, C.F. and Price, A. (2016) ‘Thriving not just surviving: A review of research on teacher resilience’, Educational Research Review, 18, pp. 77–91. https://doi.org/10.1016/j.edurev.2016.03.001 Benesch, S. (2017) Emotions and English Language Teaching: Exploring Teachers’
Emotion Labor. New York: Routledge. Boud, D., Ajjawi, R., Dawson, P., & Tai, J. (2018). Developing evaluative judgement in higher education: Assessment for knowing and producing quality work. Routledge. https://doi.org/10.4324/9781315109251 Carless, D., & Winstone, N. (2023). Designing effective feedback processes in higher
education. Routledge. Collie, R.J., Shapka, J.D. and Perry, N.E. (2016) ‘School climate and social-emotional learning: Predicting teacher stress, job satisfaction, and teaching efficacy’, Journal of Educational Psychology, 108(8), pp. 1189–1204. https://doi.org/10.1037/ edu0000136 Conklin, S., Dorgan, T. and Barreto, D. (2024) ‘Is AI the new course creator’, Discover
Education, 3(1), p. 285. doi:10.1007/S44217-024-00386-2. DeLuca, C., Coombs, A. and LaPointe-McEwan, D. (2019) ‘Assessment mindset: Exploring the relationship between teacher assessment literacy and student learning’, Assessment in Education: Principles, Policy & Practice, 26(3), pp. 247– 272. https://doi.org/10.1080/0969594X.2019.1593102 Frenzel, A. C., Taxer, J. L., Schwab, C., & Kuhbandner, C. (2016). Independent and
joint effects of teacher enthusiasm and motivation on student motivation and experience of boredom. Educational Psychology, 36(4), 1–19. https://doi.org/10.1 080/01443410.2015.1066498 Hardy, I. (2018). Governing schooling through “what works”: The promise and perils of educational research. Educational Philosophy and Theory, 50(1), 54–66. https://doi.org/10.1080/00131857.2016.1259823 Hochschild, A.R. (1983) The Managed Heart: Commercialization of Human Feeling.
Berkeley, CA: University of California Press. Holmes, W., Bialik, M. and Fadel, C. (2022) Artificial Intelligence in Education: Promises and Implications for Teaching and Learning. Boston, MA: Center for Curriculum Redesign. Hrastinski, S., Olofsson, A.D., Arkenback-Sundström, C., Utterberg, M. and
Jaldemark, J. (2019) ‘Critical imaginaries and reflections on artificial intelligence and robots in postdigital K–12 education’, Postdigital Science and Education, 1(2), pp. 427–445. https://doi.org/10.1007/s42438-019-00046-x Iliya, A. and Ifeoma, L.G. (2015) ‘Assessment of teacher motivation approaches in less developed countries’, Journal of Educational Psychology, 8(2), pp. 45–58. Lebeničnik, M. and Istenič, A. (2024) ‘Dataset higher education students’ use of
online learning resources and related beliefs’, Cogent Education, 11(1). doi:10.10 80/2331186X.2024.2385122.
study on the identity construction of local university teachers in the situation of performance management’, Journal of Educational Studies, 20(5), pp. 100–112. doi:10.14082/j.cnki.1673-1298.2024.05.009. Maslach, C. and Leiter, M.P. (2016) Burnout. In: G. Fink (ed.) Stress: Concepts, Cognition, Emotion, and Behavior. London: Academic Press, pp. 351–357. OECD (2021) Digital Education Outlook 2021: Pushing the Frontiers with Artificial
Intelligence, Blockchain and Robots. Paris: OECD Publishing. https://doi. org/10.1787/589b283f-en. Perryman, J., & Calvert, G. (2020). What motivates people to teach, and why do they leave? Accountability, performativity and teacher retention. British Journal of Educational Studies, 68(1), 3–23. https://doi.org/10.1080/00071005.2019.15 89417 Richards, G. and Richardson, R. (2018) Reducing teachers’ marking workload and
developing pupils’ learning. Abingdon: Routledge. Ryan, R.M. and Deci, E.L. (2017) Self-Determination Theory: Basic Psychological Needs in Motivation, Development, and Wellness. New York: Guilford Press. Shao, J. and Sun, S. (2025) ‘Teacher identity in the era of intelligence’, Journal of
Educational Theory and Practice, 2(1). https://doi.org/10.62177/jetp.v2i1.141. Skaalvik, E. M., & Skaalvik, S. (2017). Motivated for teaching? Associations with school goal structure, teacher self-efficacy, job satisfaction and emotional exhaustion. Teaching and Teacher Education, 67, 152–160. https://doi. org/10.1016/j.tate.2017.06.006 Skaalvik, E.M. and Skaalvik, S. (2021) ‘Teacher burnout: relations between job
demands, job resources, motivation and job satisfaction’, Teachers and Teaching, 26(7–8), pp. 602–616. Torrance, H. (2017) ‘Blaming the victim: Assessment, examinations, and the responsibilisation of students and teachers in neo-liberal governance’, Discourse: Studies in the Cultural Politics of Education, 38(1), pp. 83–96. https://doi.org/10. 1080/01596306.2015.1104854 UNESCO (2023) Guidance for Generative AI in Education and Research. Paris:
UNESCO. Worth, J., & Van den Brande, J. (2020). Teacher workload and wellbeing. National Foundation for Educational Research. Xiaoyong, H., Zirou, L. and Xiaohong, L. (2024) ‘Integrating artificial intelligence
into education: Global trends and China’s path’, e-Education Research, 45(12), pp. 13–22. doi: 10.13811/j.cnki.eer.2024.12.002.