Factors that promote research literacy in university students: A systematic literature review 2015-2025

Introduction

In recent years, higher education has faced the challenge of rethinking its educational models to meet the demands of the 21st century. Within this context, research literacy has taken on a significance that goes far beyond technical mastery of academic sources: it is a cross-cutting competency that integrates critical thinking, intellectual autonomy, and ethical responsibility regarding knowledge; According to Paul et al. (2024) and Mbandje et al. (2024), this competency involves the ability to identify complex informational needs, rigorously evaluate scientific evidence, synthesize knowledge from different disciplines, and communicate findings responsibly. All of this takes place within a university context that aspires to be a driver of sustainable and equitable innovation.

From this perspective, the research by Tachie-Donkor and Ezema (2023) is particularly revealing: while working with university students in Ghana, the authors found that those who develop strong information literacy skills adopt more proactive attitudes toward continuous learning and demonstrate better academic search patterns, which is evidenced by the fact that this is not an isolated finding; on the other hand, Makinde et al. (2025), in a study conducted with 392 library science students in Nigeria, found a statistically significant correlation (r = 0.67, p < 0.01) between information literacy and research competence, suggesting that these skills are not only relevant in Anglo-Saxon contexts but transcend cultural and disciplinary boundaries.

What contemporary research has made clear is that the development of research competencies does not occur spontaneously or linearly; furthermore, LeMire et al. (2024), after tracking more than 1,200 students throughout their university careers, observed that information literacy skills do not follow predictable trajectories, which challenges the notion that mere exposure to academic content alone produces competent research. Furthermore, this finding is reinforced by Sunny and Ramasamy (2025), who documented that most college students still exhibit moderate or low levels of digital literacy, even though they are young people with daily access to technology. The habitual use of digital devices, it seems, does not guarantee the acquisition of the skills necessary to navigate academic environments judiciously.

Added to this is the teaching dimension. Wu et al. (2022) demonstrated that both faculty characteristics and institutional conditions significantly influence teachers’ ability to foster information literacy in their students. For their part, Kadiri et al. (2024) identified significant differences based on the field of study: students in health sciences and those in other areas do not exhibit the same levels of self-efficacy or motivation regarding research tasks, highlighting that contextual factors cannot be ignored when designing pedagogical strategies.

The emergence of new technologies has added a new layer of complexity to the landscape. Yang et al. (2025), based on a bibliometric analysis of over 3,000 academic records, found that different disciplines prioritize distinct aspects of digital literacy, which hinders the development of unified conceptual frameworks. In this same vein, researchers such as Knoth et al. (2025) and Zhao (2025) have begun to explore the role of artificial intelligence in research training: educational chatbots, AI literacy, and functional resilience appear to have a significant impact on students’ motivation and critical thinking.

Given this landscape, and considering the gaps that still persist between social demands and actual educational outcomes, this article poses the following research question: What are the main factors that the scientific literature has identified as promoters of research literacy among university students during the 2015–2025 period, and how do they relate to the personal, pedagogical, and institutional dimensions? The general objective is to systematically analyze the academic output of this period to identify, characterize, and systematize these factors; the specific objectives include classifying them into the three aforementioned dimensions, describing the predominant methodological trends, identifying the most studied geographical and disciplinary contexts, and highlighting the knowledge gaps that require priority attention in future research.

The choice of a systematic review as the methodology responds to the need to rigorously synthesize the available empirical evidence. Furthermore, by following the PRISMA 2020 protocol and working with 50 articles selected from specialized databases, this study seeks to offer a comprehensive synthesis of the state of the art in university research literacy, which can guide both future research and the design of effective pedagogical interventions.

Methodology

Selection process

Two independent reviewers conducted the selection process using an Excel spreadsheet specifically designed to record decisions at each stage. The selection flow strictly followed the PRISMA diagram, comprising four consecutive phases: initial identification of 2,660 records in the databases; screening of 1,560 records by title and abstract after eliminating duplicates; eligibility assessment of 72 studies through full-text reading; and final inclusion of 50 articles that met all established criteria. Discrepancies between reviewers were resolved by consensus, and when necessary, a third evaluator was consulted.

Figure 1 shows the steps followed to ensure transparency, reproducibility, academic credibility, and traceability.

Figure 1. PRISMA 2020 flow diagram for the systematic literature review article selection process.

Note: PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-Analyses. The diagram illustrates the complete selection process, from the initial identification of 2,660 records across three database (Scopus, Web of Science, ScienceDirect) to the final inclusion of 50 studies. Exclusion criteria applied at each stage are detailed within the diagram.

Results

The analysis reveals exponential growth in research on university research literacy, with an annual rate of 26.3%. The acceleration from 2023 to 2025 (40% of total output) indicates consolidation of the field in higher education.

The 50 articles werw also analyzed acording to their categories: Q1, Q2 and Q3, which is a way of analyzing reserch excellence (See Table 3).

Rearding the identication of factors that promote literacy, they were identified by dimensión: personal (22 studies, 63%), pedagogical (8 articles, 23%) and institutional (5 articles, 14%). Within each dimensión, factors with varying effect sizes were identified (See Table 4).

The specific geographical análysis of higher education by región was also analyzed: it show that in Europe it has 26%, Asia 24%, North America 16%, Africa 12%, Latin America 14% and a tie between Oceania and the Middle East with 4%; in relation to the gaps it is evidente that in Sub-Saharan Africa and Central Asia the priority index is critical (See Table 5).

Regarding measurement instruments for university contexts, the IL Doc Model, IL- HUMASS Survey, DigComp 2.1 Adaptation, IL Engineering Scale, and the GAI Critical Literacy Scale were identified. Contexts, reliability, latent factors, and sample size with their units of analysis were also identified for each of the identified instruments.

The predominant methodologies for análysis in higher education identified were: Correlational, Quasi -experimental, Experimental. Longitudinal and Systematic Review designs. Representative studies and their advantages in the university context are also included (See Table 6).

It is considered important to establish tha thematic evolution by period used by university studente, wich were divided into three periods: 2015–2018: Fundations of University IL; 2019–2022: University Digitization and 2023–2025: AI and University Critical Thinking. Emerging factors include crital literacy with a potencial transformational impact, competency verification with a high potential impact, assisted academic writing with high potential, and academic digital collaboration with a médium – high potencial (See Table 7).

The quality criterio considered for this review were: University relevance, Methodological rigor, Sample size, Validated instruments and Practical applicability, classifield as Excellent, Good, aceptable. Poor (See Table 8).

A bibliometric analysis of 50 specific articles on research literacy in higher education (2015–2025) reveals a field undergoing rapid consolidation, with an annual growth rate of 26.3%. Institutional factors emerge as the most effective (average d = 0.83), particularly library-faculty collaboration (d = 0.94) and institutional IL frameworks (d = 0.87). Personal factors show robust consistency, led by information literacy (d = 0.82) and research self-efficacy (d = 0.76). Geographically, a Europe-Asia concentration persists (50%) with critical underrepresentation of Sub-Saharan Africa (3.5x) and Central Asia (6.0x). Emerging trends toward critical AI literacy (+410%) and verification skills (+280%) point to a fundamental paradigm shift.

The average methodological quality (77.5%) indicates disciplinary maturity, although longitudinal limitations persist (12% of studies) and require priority attention to understand skill development throughout the university experience.

Figure 2 shows the interconnected dimensions which show that most studies focused on the personal human factor (green cloud), followed by the pedagogical (blue cloud) and to a lesser extent on the institutional (red cloud); however, the three are connected making it impossible to understand them separately as the dimensions influence each other.

Figure 2. Integrative framework of the three interconnected dimensions of university research literacy: personal pedagogical and institutional.

Note: This framework synthesizes findings from the 50 reviewed studies. Circle size reflects the proportion of studies addressing each dimension: personal (63%, n = 22), pedagogical (23%, n = 8), and institutional (14%, n = 5). Bidirectional arrow indicate that dimensions mutually influence each other and cannot be interpreted in isolation. Visualization generated with VOSviewer 1.6.18.

The bibliometric analysis reveals that research literacy is a multidimensional competency; strong links with digital and media literacy are identified, indicating that the current university environment demands technological skills to access, evaluate, and produce scientific knowledge. Consequently, this dimension relates to teacher training and the use of ICT, demonstrating that training in digital environments is essential for fostering academic research. Likewise, a professional and applied dimension emerges, particularly in the health sciences, where research becomes a tool for evidence-based decision-making. Furthermore, bibliometrics also suggests that the development of research competencies begins before university: the progression from early childhood education reflects that this is a prolonged educational process requiring sustained pedagogical strategies (See Table 7 in the Annexes).

Finally, active methodologies—such as blended learning and problem-based learning—emerge as key elements in strengthening research skills (See Table 8 in the Annexes).

Discussion of results

The findings of this systematic review offer a clear yet complex picture of the factors influencing the development of university research literacy; the convergence of evidence from 50 studies, analyzed from quantitative, qualitative, and mixed perspectives, confirms that this is a systemic phenomenon in which the personal, pedagogical, and institutional dimensions do not operate independently but in constant interaction (Mbandje et al., 2024; Paul et al., 2024).

With regard to personal factors, the data are compelling: information literacy (d = 0.82) and research self-efficacy (d = 0.76) emerge as robust predictors of competency development, regardless of geographic or disciplinary context; large-scale studies such as those by Makinde et al. (2025) and Tachie-Donkor and Ezema (2023) converge on this point with experimental research, suggesting that these factors function as educational universals, albeit with contextual variations that cannot be ignored. In this regard, self-efficacy appears to act as a mediator between prior digital competencies and critical thinking, which has concrete implications for the design of training interventions.

The pedagogical domain, although less represented in the analyzed literature (23% of the studies), presents the highest effect sizes; collaboration between the library and the faculty (d = 0.94) and structured instruction (d = 0.78) emerge as the strategies with the greatest documented impact. This is no minor finding: it suggests that the most effective interventions are those that bring together different institutional actors around a shared goal, combining direct instruction with practical application in specific disciplinary contexts. The evidence from Reeves et al. (2024) and Kautonen and Gasparini (2024) in this regard is particularly consistent.

Institutional factors (14% of the studies) also deserve attention, precisely because, despite being underrepresented in the literature, they show considerable effects. Institutional conceptual frameworks (d = 0.87) and technological resources emerge as catalysts that amplify the impact of personal and pedagogical variables. This implies that, no matter how ly motivated and capable a student may be, without an institutional environment that facilitates access to resources and establishes clear policies for research training, competency development encounters structural barriers that are difficult to overcome.

From the perspective of methodological trends, the predominance of correlational designs (36% of the studies) reflects a consolidation of the field, albeit with evident limitations in establishing causal relationships. The scarcity of longitudinal studies (barely 12%) is one of the most significant gaps identified: understanding how research competencies develop throughout the university experience requires precisely this type of tracking, which cross-sectional designs cannot provide. This gap is even more relevant when considering, as LeMire et al. (2024) point out, that the development of these competencies does not follow predictable linear patterns.

The geographic distribution of the studies reveals another problematic asymmetry. The concentration of research in Europe and Asia (50% of the total) contrasts with the underrepresentation of Sub-Saharan Africa and Latin America, regions that are home to a significant proportion of the world’s university enrollment. This gap is not merely quantitative: it implies that the enabling factors identified in the literature largely reflect educational realities in high-income contexts, whose transferability to other cultural and socioeconomic settings cannot be assumed without specific contextual validation.

One of the most notable transformations emerging from the analysis is the paradigm shift that has been underway since 2023: the emergence of generative artificial intelligence has redefined the very concept of research literacy. Factors such as critical thinking take on new dimensions when applied to the evaluation of AI-generated information (Arthi & Gandhimathi, 2025), and the exponential growth of studies on critical AI literacy (+410% between 2023 and 2025) suggests that this will be one of the fastest-growing fields in the coming years.

Finally, the methodological tension between quantitative and qualitative approaches should be recognized as productive rather than as a limitation. While quantitative studies tend to operationalize research literacy in terms of measurable competencies, qualitative approaches highlight process-oriented and contextual dimensions that elude standardiz. This duality suggests that research literacy is a hybrid construct that requires conceptual frameworks capable of articulating both dimensions, guiding both future research and the design of pedagogical interventions based on robust empirical evidence.

Conclusions

This systematic review of 50 studies published between 2015 and 2025 confirms that university research literacy is a multidimensional, dynamic, and deeply contextualized competency; furthermore, personal factors—led by information literacy (d = 0.82) and research self-efficacy (d = 0.76)—show the most consistent evidence across different geographic and disciplinary contexts, confirming their role as fundamental predictors of competency development. At the same time, pedagogical factors—though less prevalent in the literature—exhibit the greatest effects, particularly library-faculty collaboration (d = 0.94), which invites a rethinking of the organization of university educational processes. Institutional factors, for their part, act as amplifiers of personal and pedagogical effects, and their underrepresentation in the reviewed literature constitutes in itself a call for future research.

From a methodological standpoint, the predominance of correlational designs and the scarcity of longitudinal studies are the most significant limitations in the field. Without the complete trajectories of competency development that only longitudinal studies can provide, it is difficult to understand how and when the most significant changes in university students’ research training occur. This gap must be addressed as a priority in the coming years.

Another important limitation of the present study is the geographical concentration of the analyzed literature. The fact that Europe and Asia account for 50% of the reviewed studies not only reflects imbalances in global scientific output but also limits the possibility of generalizing the findings to contexts such as Latin America, Africa, or the Middle East, which remain significantly underrepresented. The cross-cultural validity of the identified enabling factors needs to be verified in these settings.

In terms of practical implications, the results of this review point in a clear direction: the most effective interventions are those that combine explicit instruction in information literacy with active methodologies, collaborative work between the library and faculty, and institutional policies that guarantee adequate resources and structured research training programs. For faculty, this involves integrating the development of research competencies into their daily practices; for university administrators, it entails formalizing inter-institutional collaboration and ensuring relevant technological resources. For curriculum designers, it means incorporating these competencies across the board into the curriculum.

Finally, the paradigm shift driven by artificial intelligence in university educational settings cannot be ignored. Twenty-first-century research literacy increasingly includes the ability to critically evaluate information generated by AI systems, verify its reliability, and use it ethically. The 410% increase in studies on critical AI literacy between 2023 and 2025 suggests that this will be a central field of research in the coming years, and that available conceptual frameworks and assessment tools will need to be continuously updated to reflect a rapidly evolving educational reality.