30 Literature Review Examples
From Q1 Journals (2026 Templates)

Digital financial inclusion literature review

“The relationship between financial inclusion and economic development has been extensively documented (Demirguc-Kunt et al., 2018; Klapper et al., 2016). However, the emergence of mobile money platforms has created a new strand of literature examining digital pathways to financial inclusion (Suri & Jack, 2016; Mbiti & Weil, 2015).”

“Despite this growing body of work, existing studies predominantly examine Sub-Saharan Africa, leaving the digital financial inclusion dynamics of South and Southeast Asia — where mobile penetration has grown by 340% since 2015 — largely unexplored.”

ESG investing and firm performance

“Research on the relationship between environmental, social, and governance (ESG) criteria and financial performance has produced mixed results (Friede et al., 2015; Renneboog et al., 2008). Meta-analyses suggest a positive but modest relationship (Orlitzky et al., 2003), though methodological heterogeneity complicates cross-study comparison.”

“A critical limitation of existing research is its focus on large-cap institutional portfolios. The ESG-performance relationship among retail investors and small-cap stocks — which constitute the majority of market participants — remains poorly understood.”

Cryptocurrency volatility and market efficiency

“The efficient market hypothesis (Fama, 1970) has been tested extensively in traditional asset markets, but its applicability to cryptocurrency markets remains contested (Urquhart, 2016; Nadarajah & Chu, 2017). Early evidence suggested Bitcoin markets were inefficient (Cheah & Fry, 2015), though more recent studies indicate evolving efficiency as market maturity increases.”

“Existing studies treat cryptocurrency markets as homogeneous. The efficiency differential between major cryptocurrencies (Bitcoin, Ethereum) and altcoins — which exhibit fundamentally different liquidity profiles — has not been examined at the intraday level.”

Social media use and mental health in adolescents

“The relationship between social media use and adolescent mental health has attracted substantial research attention since 2015 (Twenge et al., 2018; Orben & Przybylski, 2019). Longitudinal studies have identified associations between passive social media use and depressive symptoms (Coyne et al., 2020), though effect sizes are typically small and context-dependent.”

“The literature has not adequately distinguished between platform types. The psychological mechanisms underlying Instagram use — centered on visual social comparison — are likely distinct from those of TikTok, where algorithmic content delivery rather than peer networks drives engagement. No study has tested these platform-specific pathways simultaneously.”

Mindfulness interventions in workplace settings

“Mindfulness-based interventions (MBIs) have demonstrated efficacy in clinical populations for reducing anxiety and depression (Hofmann et al., 2010; Khoury et al., 2013). The translation of these findings to workplace settings has been examined in a growing number of randomized controlled trials (Lomas et al., 2019; Bartlett et al., 2019).”

“While existing workplace MBI research demonstrates reductions in self-reported stress, the mechanisms linking mindfulness practice to specific job performance outcomes — particularly for roles requiring creative problem-solving — have not been examined with adequate mediator analysis.”

Decision fatigue and consumer behavior

“Decision fatigue — the deterioration of decision quality following an extended period of decision-making (Baumeister et al., 1998) — is well-established in laboratory settings. Field evidence from judges (Danziger et al., 2011) and medical professionals (Linder et al., 2014) has extended these findings to high-stakes real-world contexts.”

“No study has examined decision fatigue in online consumer contexts, where users may face hundreds of choice options within a single session. The interaction between choice overload (Iyengar & Lepper, 2000) and decision fatigue in e-commerce environments represents a significant gap.”

Gamification in higher education

“Gamification — the application of game design elements in non-game contexts (Deterding et al., 2011) — has been studied extensively in K-12 education (Dicheva et al., 2015), with evidence suggesting positive effects on motivation and engagement (Hamari et al., 2014). Application to higher education has grown significantly since 2016.”

“Existing higher education gamification studies focus predominantly on STEM disciplines in Western university contexts. The transferability of these findings to humanities and social science disciplines, where learning objectives are less easily quantified, has not been tested.”

AI writing tools and student learning outcomes

“The integration of AI writing assistance tools in academic contexts has generated considerable debate regarding academic integrity (Cotton et al., 2023; Perkins, 2023). Empirical research on learning outcomes — as distinct from integrity concerns — remains nascent, with most studies published after 2022.”

“The critical distinction between AI tools used as drafting aids versus revision aids has not been examined. Whether AI assistance at different stages of the writing process produces different effects on students' own writing development represents an unaddressed question with significant pedagogical implications.”

Peer assessment accuracy and reliability

“Peer assessment has been studied as both a learning tool and an assessment instrument for over three decades (Topping, 1998). Research consistently shows that peer marks correlate moderately with instructor marks (r = 0.60–0.75) and that the learning benefits of the process may exceed the accuracy of the resulting grades (Nicol et al., 2014).”

“The reliability of peer assessment has been studied almost exclusively in face-to-face or synchronous online contexts. How asynchronous, anonymous peer assessment — increasingly common in large online courses — affects both accuracy and student engagement has not been examined longitudinally.”

Telemedicine adoption post-COVID

“The COVID-19 pandemic accelerated telemedicine adoption globally, with utilization rates increasing by 3,000% in some healthcare systems (Wosik et al., 2020). Research during the pandemic documented patient satisfaction and clinical outcomes comparable to in-person care for many conditions (Lam et al., 2020; Nouri et al., 2020).”

“Post-pandemic telemedicine research has focused on urban, digitally connected patient populations. The long-term adoption barriers and clinical outcome differences for rural and elderly populations — who exhibit systematically lower digital health literacy — have not been examined in the post-emergency context.”

Vaccine hesitancy and information sources

“Vaccine hesitancy — defined by the WHO as the delay in acceptance or refusal of vaccination despite availability — has been recognized as one of the top ten threats to global health (WHO, 2019). Research has identified multiple determinants including trust in health authorities, social norms, and perceived risk (MacDonald, 2015).”

“Existing research treats information source exposure as a static variable. The cumulative effect of repeated misinformation exposure over time — and whether correction messaging can reverse established hesitancy — has not been tested with longitudinal designs tracking the same individuals across multiple vaccination campaigns.”

AI diagnostic tools in radiology

“Artificial intelligence applications in diagnostic radiology have demonstrated performance comparable to radiologists on specific tasks including chest X-ray interpretation (Rajpurkar et al., 2017) and diabetic retinopathy screening (Gulshan et al., 2016). Systematic reviews suggest AI performs well in controlled research settings (Liu et al., 2019).”

“The translation gap between research performance and clinical deployment remains underexplored. Specifically, AI diagnostic tool performance under real-world conditions — including image quality variation, patient population diversity, and workflow integration constraints — differs systematically from benchmark dataset performance in ways not yet quantified.”

Urban green space and wellbeing

“The relationship between urban green space and resident wellbeing is supported by a substantial body of epidemiological and experimental evidence (Hartig et al., 2014; van den Berg et al., 2015). Proposed mechanisms include stress reduction (Kaplan, 1995), physical activity facilitation, and air quality improvement (Nowak et al., 2014).”

“Existing research predominantly examines large parks and green corridors in high-income cities. The wellbeing effects of micro-green spaces — pocket parks, green rooftops, urban gardens — in dense, lower-income urban environments are not adequately characterized, despite their prevalence as the primary green space typology in developing city contexts.”

Carbon pricing policy effectiveness

“Carbon pricing mechanisms — including carbon taxes and cap-and-trade systems — have been studied as tools for emissions reduction since Pigou (1920). Empirical evaluations of implemented systems including the EU Emissions Trading System (Ellerman & Buchner, 2008) and British Columbia's carbon tax (Murray & Rivers, 2015) provide evidence of modest emissions reductions.”

“The distributional effects of carbon pricing on low-income households have been studied primarily in developed economy contexts. How carbon pricing policies affect energy poverty and household welfare in emerging economies — where energy access remains incomplete — represents a critical gap in the policy literature.”

Large language model hallucination

“Hallucination in neural text generation — the production of factually incorrect but fluent text — was identified as a challenge in early sequence-to-sequence models (Wiseman et al., 2017). With the emergence of large language models (LLMs), hallucination has become a central concern for deployment in high-stakes applications (Maynez et al., 2020; Ji et al., 2023).”

“Hallucination benchmarks predominantly use general knowledge tasks. The domain-specific hallucination profile of LLMs in medical, legal, and financial applications — where factual errors carry significantly higher consequence — has not been systematically characterized across model families.”

Explainable AI in high-stakes decision making

“Explainable AI (XAI) research has proliferated following recognition that opaque algorithmic systems create accountability problems in consequential domains (Doshi-Velez & Kim, 2017; Lipton, 2018). The field has produced numerous explanation methods including LIME (Ribeiro et al., 2016) and SHAP (Lundberg & Lee, 2017).”

“XAI research has focused primarily on technical explanation quality — measured by proxy metrics — rather than human comprehension and decision quality. Whether explanations that score well on technical metrics actually improve human decision-making in real deployment contexts remains poorly evidenced.”

Federated learning for privacy-preserving AI

“Federated learning (McMahan et al., 2017) enables model training across distributed data sources without centralized data collection, addressing privacy concerns that limit AI deployment in sensitive domains including healthcare (Rieke et al., 2020) and finance (Yang et al., 2019).”

“Federated learning performance has been characterized on relatively homogeneous data distributions. Real-world deployment involves extreme data heterogeneity across participating institutions — different patient populations, imaging protocols, and data quality — the effect of which on model convergence and bias has not been adequately studied.”