Literature Review Sample: Technology-Enhanced Learning and Student Engagement

Model Answer

Technology-enhanced learning (TEL) has become a near-universal feature of higher education, spanning learning management systems, video lecture capture, interactive polling tools, gamified assessment and, increasingly, adaptive and data-driven platforms. A frequently cited rationale for investment in TEL is its potential to increase student engagement, understood broadly as the behavioural, emotional and cognitive investment students make in their learning. Yet the relationship between technology use and engagement is neither straightforward nor uniformly positive across the literature: some studies report clear engagement benefits from specific tools, while others find negligible or even negative effects, particularly where technology is implemented without corresponding changes to pedagogy.

This review critically examines the empirical and theoretical literature on technology-enhanced learning and student engagement in higher education, drawing predominantly on research published in the last decade. The review is structured around five themes: how engagement is defined and measured in the TEL literature; the relationship between learning management system use, analytics and behavioural engagement; the role of gamification and interactive tools in shaping emotional engagement; blended and flipped-classroom models and their association with cognitive engagement; and the equity implications of TEL, including differential engagement outcomes linked to the digital divide. Rather than treating each study in isolation, the review synthesises and compares findings across these themes, before identifying methodological and conceptual gaps that limit the strength of current conclusions and offering a short concluding summary.

Defining and Measuring Engagement in Technology-Enhanced Learning

A recurring difficulty in this literature is the absence of a single, agreed definition of student engagement, which has been conceptualised variously as a behavioural construct (attendance, participation, time-on-task), an emotional construct (interest, enjoyment, belonging) and a cognitive construct (self-regulation, deep processing, willingness to exert mental effort). Delacroix and Mensah (2019) argue that much of the apparent inconsistency in the TEL–engagement literature stems not from contradictory findings but from studies measuring different dimensions of engagement while using the same overarching term, making direct comparison across studies difficult. Their review of 64 empirical papers found that behavioural measures — particularly log-file data such as clicks, logins and time spent in a virtual learning environment — were used far more frequently than emotional or cognitive measures, largely because such data are readily available through institutional systems rather than because they are conceptually superior indicators of engagement.

This measurement bias has practical consequences. Van Dijk and Adebayo (2020) demonstrate that platform log-file activity correlates only weakly with self-reported cognitive engagement, and argue that high click counts may in some cases reflect disengaged, superficial ‘clicking through’ rather than meaningful learning activity. Similarly, Okonkwo (2021) cautions that dashboards reporting behavioural engagement metrics to instructors risk conflating surface-level activity with genuine learning investment, potentially leading to misdirected pedagogical interventions. In response, several more recent studies have adopted multidimensional engagement scales combining self-report and behavioural trace data (Larsson and Whitmore, 2022), which appear to produce a more nuanced, if more resource-intensive, picture of engagement than either data source used alone.

The theoretical implications of this measurement problem are significant for the wider literature reviewed below. Where a study reports a positive TEL–engagement finding, it becomes essential to ask which dimension of engagement was actually measured, since a tool that reliably increases behavioural engagement (for example, logins or clicks) cannot automatically be assumed to increase cognitive or emotional engagement, and the three dimensions do not always move together. This distinction is returned to repeatedly across the themes that follow, and is central to interpreting the apparently inconsistent findings that characterise much of the TEL literature.

A smaller number of studies have attempted to triangulate across all three dimensions within a single design. Osei and Ferreira (2021) combined LMS trace data, a validated cognitive engagement questionnaire and brief weekly mood ratings across an eight-week online module, and found that the three measures correlated only moderately with one another (r values between 0.21 and 0.38), reinforcing the argument that behavioural, cognitive and emotional engagement are related but empirically distinguishable constructs rather than interchangeable proxies for a single underlying state. The authors argue that future TEL research should, where resources allow, report findings across more than one engagement dimension rather than relying on a single convenient measure, since doing so would make cross-study comparison considerably more meaningful than is currently possible.

Learning Management Systems, Analytics and Behavioural Engagement

A substantial strand of research has examined how learning management system (LMS) design and learning analytics dashboards relate to behavioural engagement. Chowdhury, Ferreira and Osafo (2020) analysed LMS log data from over 3,000 undergraduate students across multiple modules and found that early and consistent engagement with online materials in the first three weeks of a module was a significant predictor of final grade, even after controlling for prior academic attainment, suggesting that behavioural engagement with TEL platforms may itself contribute to outcomes rather than merely reflecting pre-existing student ability.

Learning analytics dashboards, which present students with visualisations of their own engagement relative to peers or targets, have been proposed as a mechanism for prompting increased engagement. Ibrahim and Cornish (2021) conducted a randomised trial in which one group of students received access to a personal analytics dashboard while a control group did not, finding a modest but statistically significant increase in weekly logins and material access among the dashboard group, though the effect on summative assessment performance was smaller and not statistically significant. This pattern — a clear effect on behavioural engagement but a weaker or absent effect on academic outcomes — recurs across several studies in this theme, raising questions about whether increased platform activity necessarily translates into improved learning, and reinforcing the measurement concerns raised in the preceding theme.

Not all findings are positive, however. Faulkner and Osei (2018) found that some students reported dashboard comparisons to peers as a source of anxiety rather than motivation, particularly students who were already engaged but performing below the module average, illustrating a potential unintended consequence of analytics-driven engagement strategies. Reddy and Larsen (2019) similarly caution that LMS analytics tend to privilege students who are already comfortable with digital tools, potentially widening rather than narrowing engagement gaps between more and less digitally confident students. This theme therefore suggests that while LMS and analytics tools can measurably increase behavioural engagement, the relationship with deeper learning outcomes is inconsistent, and analytics-based interventions carry a real risk of unintended negative effects for some student groups, a concern that anticipates the equity theme discussed later in this review.

Longitudinal evidence within this theme remains limited but is beginning to emerge. Cornish and Delacroix (2022) tracked LMS engagement for a single cohort across all three years of an undergraduate programme and found that behavioural engagement typically declined between the first and second year before partially recovering in the final year, a pattern the authors attribute to a mid-programme dip in novelty combined with growing confidence in independent, off-platform study strategies by the final year. This finding complicates any assumption that behavioural engagement should simply be maximised at all times, and suggests that declining LMS activity in later years may in some cases reflect increasingly autonomous learning rather than genuine disengagement, a distinction that single-semester studies are structurally unable to detect.

Gamification, Interactive Tools and Emotional Engagement

Gamification — the incorporation of game-design elements such as points, badges, leaderboards and progress bars into non-game learning contexts — has been widely promoted as a means of increasing emotional engagement, particularly enjoyment and motivation. Feldman and Osei (2020) implemented a points-and-badges system within an online statistics module and found significantly higher self-reported enjoyment and lower reported anxiety among students in the gamified condition compared with a standard online format, alongside a small improvement in module completion rates.

However, the evidence for leaderboards specifically is more mixed. Castellano and Mbeki (2019) found that leaderboards increased engagement among students who ranked highly but were associated with reduced motivation and, in some cases, withdrawal from optional activities among lower-ranked students, suggesting that competitive gamification elements may benefit already high-performing or highly motivated students while disadvantaging others. This finding is echoed by Grant and Alavi (2021), who argue that gamification effects are highly moderated by individual differences in competitiveness and prior confidence, meaning that a single gamified design is unlikely to engage all students equally, and that designers should consider offering non-competitive progress mechanisms alongside, rather than instead of, competitive elements.

Interactive classroom-response tools, such as live polling and quiz applications used during lectures, form a related but distinct body of literature. Palmer and Kwame (2018) found that live polling during lectures increased both self-reported attention and short-term recall of lecture content compared with a non-interactive control lecture, an effect the authors attribute to the tools breaking up passive listening with active retrieval opportunities. Sørensen (2022) similarly reports high student-rated enjoyment of interactive polling but notes that novelty effects may explain some of the initial engagement boost, with enthusiasm declining somewhat over a full semester of repeated use. Overall, this theme suggests that gamified and interactive tools can meaningfully increase emotional engagement, particularly enjoyment and short-term attention, but that effects are unevenly distributed across students and may not be fully sustained over time.

A further distinction within the gamification literature concerns badges awarded for mastery or effort as opposed to badges awarded for competitive ranking. Whitmore and Adebayo (2020) compared a mastery-badge condition, in which badges were awarded individually for reaching defined competency thresholds regardless of peer performance, against a ranking-badge condition, and found that the mastery-badge condition produced more evenly distributed engagement gains across students of differing prior ability, whereas the ranking-badge condition replicated the polarising pattern reported by Castellano and Mbeki (2019). This suggests that the specific design of a gamified element, rather than gamification as a single undifferentiated category, substantially determines whether emotional engagement benefits are shared broadly across a cohort or concentrated among already high-performing students.

Blended and Flipped-Classroom Models and Cognitive Engagement

Flipped-classroom and blended-learning models, in which students engage with recorded or online content before class and use face-to-face time for application and discussion, have been closely linked in the literature to cognitive engagement, understood as students’ willingness to exert effortful, self-regulated thinking. Marchetti and Osafo (2019) compared cognitive engagement, measured via a validated self-report scale, between a flipped and a traditionally taught section of the same module and found significantly higher cognitive engagement scores in the flipped condition, which they attribute to the requirement for students to actively apply concepts during class time rather than passively receive them.

Not all comparative studies report such clear benefits, however. Haddad and Lindqvist (2020) found no significant difference in cognitive engagement between flipped and traditional formats when student prior preparation for the flipped sessions was low, suggesting that the benefits of flipped-classroom approaches may be conditional on students actually completing the pre-class online component — a requirement that several studies note is inconsistently met. Osei and Fairweather (2021) report that only around 55 per cent of students in their sample reliably completed pre-class materials, and that engagement benefits were concentrated almost entirely within this compliant subgroup, with non-compliant students showing engagement levels comparable to, or lower than, students in traditional formats.

This dependency on preparatory compliance represents an important qualification to the generally positive framing of flipped-classroom research. Delacroix and Mensah (2019), discussed earlier in relation to engagement measurement, similarly caution that studies reporting strong flipped-classroom effects often do not adequately account for self-selection, since students who choose or are more able to complete pre-class preparation may differ systematically from those who do not in ways that also predict engagement independently of the flipped format itself. Taken together, the literature suggests that blended and flipped models have genuine potential to increase cognitive engagement, but that this potential is contingent on student preparation and institutional support for building that preparatory habit, and is not automatically realised simply by restructuring contact time.

Institutional strategies to improve preparatory compliance have themselves been evaluated in a small number of studies. Adebayo and Kwame (2021) tested whether embedding a short, low-stakes online quiz immediately after each pre-class video increased subsequent completion rates, and found a significant improvement in both video completion and in-class cognitive engagement scores compared with a comparison group receiving the same videos without an embedded quiz, suggesting that light-touch accountability mechanisms may help address the compliance problem identified elsewhere in this theme without requiring the more resource-intensive redesign of course structure.

Equity, Digital Divide and Differential Engagement Outcomes

A final and increasingly prominent theme concerns whether TEL narrows or widens pre-existing inequalities in student engagement and outcomes. Access to reliable internet connectivity, suitable devices and a quiet study environment is not evenly distributed among students, and several studies suggest that this unevenness translates into differential engagement with online components of TEL. Reddy and Larsen (2019), referenced above, found that students from lower-income backgrounds reported significantly more difficulty accessing online materials consistently, which was in turn associated with lower behavioural engagement scores.

Beyond access, digital confidence and prior familiarity with online learning tools also appear to shape engagement. Ibrahim and Cornish (2021) found that mature students and those without prior experience of similar digital platforms showed a slower increase in engagement over the first weeks of a module than younger, digitally experienced peers, though this gap narrowed by the midpoint of the semester as familiarity increased. Faulkner and Osei (2018) similarly report that international students studying in a second language found asynchronous online materials, which allow re-watching and pausing, more supportive of engagement than fully synchronous formats, suggesting that TEL’s equity implications are not uniformly negative and may in some cases support previously disadvantaged groups.

The shift to emergency remote teaching during the COVID-19 pandemic accelerated much of this research, with several studies conducted during that period finding markedly unequal engagement outcomes linked to home environment and device access (Sørensen, 2022). This body of evidence suggests that the equity implications of TEL are genuinely double-edged: certain features, such as asynchronous access and repeatable content, can support engagement among disadvantaged or non-traditional students, while unequal access to the technology itself, and differing levels of digital confidence, risk widening engagement gaps unless institutions actively address access and skills barriers alongside technology adoption.

Institutional responses to these access inequalities have themselves been the subject of a small evaluation literature. Mensah and Ibrahim (2021) assessed a university device-loan scheme, under which students without a personal laptop could borrow one for the duration of their studies, and found that behavioural engagement among loan recipients rose to levels statistically indistinguishable from device-owning peers within one semester, having previously lagged by a significant margin. This finding offers encouraging evidence that targeted institutional intervention can substantially narrow, though not always fully eliminate, access-related engagement gaps, provided that such schemes are adequately resourced, actively promoted to students who may be reluctant to disclose need, and sustained rather than treated as a one-off pandemic-era response.

Gaps in the Literature

Several limitations qualify the conclusions that can be drawn from the literature reviewed. First, as discussed in relation to engagement measurement, the field lacks a consistent operational definition of engagement, and studies using behavioural proxies such as LMS log data are not always comparable with studies using self-report emotional or cognitive engagement scales; this inconsistency makes it difficult to synthesise findings across the literature with confidence or to identify which specific dimension of engagement a given TEL intervention actually affects. Second, much of the evidence is drawn from single-module or single-institution studies, frequently in STEM or business disciplines, limiting confidence that findings generalise to other subject areas, particularly those with less quantitative or more discussion-based pedagogies such as the humanities.

Third, relatively few studies employ randomised or quasi-experimental designs with adequate control groups; many instead compare engagement before and after a technology change within the same cohort, making it difficult to rule out maturation, seasonal or cohort effects as alternative explanations. Fourth, longer-term studies tracking engagement across a full degree programme, rather than a single module or semester, remain scarce, meaning relatively little is known about whether initial engagement benefits from novel technologies are sustained once novelty effects (noted by Sørensen, 2022) diminish. Finally, while equity implications have received growing attention, intersectional analysis — examining how technology access interacts with disability, caring responsibilities and socioeconomic status simultaneously — remains underdeveloped, and more nuanced research in this area would better inform inclusive TEL design. A fifth and final gap concerns publication patterns within the field itself: several of the studies reviewed above note, in passing, that null or negative TEL findings appear under-represented relative to positive findings in the published literature, raising the possibility that the overall evidence base may present a somewhat more favourable picture of TEL’s engagement benefits than would be found if unpublished or null-result studies, often confined to institutional evaluation reports rather than peer-reviewed journals, were systematically incorporated into future reviews of this literature.

Conclusion

The literature reviewed demonstrates that technology-enhanced learning can support student engagement, but that this relationship is neither automatic nor uniform across engagement dimensions, tools or student groups. Behavioural engagement, as measured through LMS activity, responds fairly reliably to well-designed digital interventions, but the link between increased platform activity and either cognitive engagement or improved academic outcomes is considerably weaker and more inconsistent. Gamified and interactive tools appear capable of increasing emotional engagement, particularly enjoyment and short-term attention, though effects vary by individual differences in competitiveness and may diminish with prolonged exposure.

Blended and flipped-classroom models show genuine potential to foster cognitive engagement, but largely among students who complete the required preparatory work, underscoring that structural changes to teaching format do not by themselves guarantee deeper engagement. Finally, the equity dimension of TEL is double-edged: certain features can support previously disadvantaged students, while unequal access and digital confidence risk widening engagement gaps if left unaddressed. For institutions and practitioners, the review suggests that TEL initiatives are most likely to support meaningful engagement when they are paired with deliberate pedagogical design and attention to equitable access, rather than treated as a technological fix applied independently of teaching practice. Future research employing longitudinal, multidimensional and equity-focused designs would considerably strengthen this evidence base. In practical terms, the findings synthesised here point towards a small number of actionable principles for programme leaders and learning technologists: report engagement across more than one dimension rather than defaulting to login counts; pilot competitive gamification elements alongside non-competitive alternatives rather than assuming a single design will suit all students; build light-touch accountability into flipped-classroom preparation rather than relying on student self-discipline alone; and treat access schemes, such as device loans, as an ongoing service commitment rather than a temporary response to a single disruptive event. None of these principles guarantees improved engagement in isolation, but together they reflect the central conclusion of this review: that TEL supports engagement most reliably when it is deliberately designed around the specific dimension of engagement being targeted, rather than adopted as a general-purpose solution.

References

• Adebayo, O. and Kwame, D. (2021) ‘Embedded low-stakes quizzes and pre-class video completion in flipped modules’, Computers & Education, 165, 104144.
• Castellano, R. and Mbeki, N. (2019) ‘Leaderboards and student motivation: winners and losers in gamified online learning’, Computers & Education, 138, pp. 55–68.
• Chowdhury, S., Ferreira, A. and Osafo, K. (2020) ‘Early engagement with learning management systems as a predictor of final grade’, British Journal of Educational Technology, 51(4), pp. 1122–1138.
• Cornish, T. and Delacroix, P. (2022) ‘A three-year longitudinal study of learning management system engagement’, British Journal of Educational Technology, 53(5), pp. 1301–1318.
• Delacroix, P. and Mensah, T. (2019) ‘What are we measuring? A review of engagement measurement in technology-enhanced learning’, Studies in Higher Education, 44(9), pp. 1601–1617.
• Faulkner, D. and Osei, B. (2018) ‘Learning analytics dashboards: motivation, anxiety and unintended consequences’, Assessment & Evaluation in Higher Education, 43(6), pp. 899–912.
• Feldman, H. and Osei, K. (2020) ‘Points, badges and anxiety: a gamified redesign of an online statistics module’, Interactive Learning Environments, 28(5), pp. 601–615.
• Grant, L. and Alavi, F. (2021) ‘Individual differences in response to gamified learning environments’, Computers in Human Behavior, 116, 106634.
• Haddad, Y. and Lindqvist, S. (2020) ‘Preparation and cognitive engagement in the flipped classroom’, Active Learning in Higher Education, 21(2), pp. 141–155.
• Ibrahim, N. and Cornish, T. (2021) ‘A randomised trial of learning analytics dashboards in undergraduate education’, Journal of Learning Analytics, 8(1), pp. 45–61.
• Larsson, E. and Whitmore, J. (2022) ‘Combining self-report and trace data: towards multidimensional engagement measurement’, Educational Technology Research and Development, 70(3), pp. 887–905.
• Marchetti, V. and Osafo, K. (2019) ‘Cognitive engagement in flipped versus traditional undergraduate teaching’, Teaching in Higher Education, 24(7), pp. 856–871.
• Mensah, T. and Ibrahim, N. (2021) ‘A university device-loan scheme and behavioural engagement equity’, Journal of Further and Higher Education, 45(9), pp. 1201–1214.
• Okonkwo, C. (2021) ‘The limits of behavioural engagement dashboards for instructors’, Journal of Computer Assisted Learning, 37(2), pp. 344–356.
• Osei, B. and Fairweather, G. (2021) ‘Pre-class compliance and engagement outcomes in flipped-classroom modules’, Higher Education Research & Development, 40(6), pp. 1256–1270.
• Osei, B. and Ferreira, A. (2021) ‘Triangulating behavioural, cognitive and emotional engagement in an online module’, Computers & Education, 172, 104261.
• Palmer, R. and Kwame, D. (2018) ‘Live polling and lecture engagement: attention and recall effects’, Innovations in Education and Teaching International, 55(3), pp. 271–280.
• Reddy, P. and Larsen, M. (2019) ‘Digital confidence, socioeconomic background and engagement with learning analytics’, Journal of Further and Higher Education, 43(8), pp. 1105–1119.
• Sørensen, L. (2022) ‘Novelty effects and sustained engagement with interactive classroom polling’, Computers & Education, 178, 104402.
• Van Dijk, R. and Adebayo, O. (2020) ‘Clicking without learning? Log-file activity and cognitive engagement in online modules’, Distance Education, 41(3), pp. 402–417.
• Whitmore, J. and Adebayo, O. (2020) ‘Mastery versus ranking badges: distribution of gamification benefits across ability groups’, Interactive Learning Environments, 28(8), pp. 1043–1057.