Transforming higher education: a decade of integrating wikipedia and wikidata for literacy enhancement and social impact

Abstract

This study examines a decade-long implementation of a course model leveraging Wikipedia and Wikidata as primary educational platforms in higher education. In alignment with the UNs’ SDG 4, this initiative emphasized inclusive, equitable education and lifelong learning opportunities. The study scrutinizes the formulation and deployment of three elective courses, rooted in this model, which were designed to augment students’ academic, digital, collaborative, and communication skills through the creation of Open Educational Resources (OERs), achieving significant social impact—evidenced by over 2000 new and 7000 edited articles, accruing 75 million public views. The research addresses three principal areas: the development and application of the model; course outcomes, including OERs produced, academic achievements, and students’ learning experiences; perceived challenges and benefits from the perspective of both students and faculty. A mixed-methods approach was employed to examine data from 17 iterations, involving 616 participants. Students’ learning experience was extracted from post-course questionnaires completed by 70% (n = 429). Findings demonstrate the role of Wikipedia and Wikidata in fostering knowledge creation, digital and data literacies and critical thinking, with the research contributing to the conversation surrounding Open Educational Practices. Findings include details on incorporating issues of diversity, equity, inclusion (DEI) and knowledge gaps into the curriculum design, and map challenges and benefit for students and faculty. This extensive study offers valuable insights into the effectiveness of embedding OERs in higher education, spotlighting the pedagogical implications and social impact of this approach. It discusses the relevance of this educational strategy in the context of Generative AI technologies.

Introduction

The advent of the digital era has markedly reconfigured the educational paradigms, with a pronounced impact on tertiary education. Central to this reconfiguration is the assimilation of open educational practices (OEP) and open educational resources (OERs) within academic syllabi. This article explores a decade-spanning initiative undertaken as part of undergraduate studies at a higher education institution (name removed for peer review), wherein Wikipedia and Wikidata have been integrated as principal pedagogical platforms in various academic courses. This strategy is congruent with the United Nations’ Sustainable Development Goal 4,¹ which advocates for universal access to inclusive, equitable quality education and lifelong learning opportunities. In 2013, a new course framework was developed, incorporating Wikipedia into the higher education milieu. This framework has been implemented in three academic courses: “Wiki-Med”, “Wikipedia: Skills for Producing and Consuming Knowledge”, and “From Web 2.0 to Web 3.0: From Wikipedia to Wikidata”. The objective of these courses was to refine students’ scholarly, digital, collaborative, and communicative competencies, whilst concurrently amplifying societal impact.

Despite a rich body of literature on the use of Wikipedia as an alternative assessment mechanism in classroom settings, there is a notable paucity of research focusing on its integration as a principal evaluative tool within a structured course framework (Davis et al., 2023). Additionally, the concepts of the Semantic Web, Wikidata, and their applications in educational contexts represent relatively nascent phenomena, only recently beginning to be scrutinized in pedagogical studies (Evenstein Sigalov & Nachmias, 2023; Farda-Sarbas & Müller-Birn, 2019; Mora-Cantallops et al., 2019). Given the scarcity of available literature and resources on this topic, this study endeavors to thoroughly examine the outcomes of three academic courses over a decade. The objectives of this investigation include: (1) Outlining the design principles of the course model; (2) Analyzing the outcomes and impacts of the courses; and (3) Investigating the advantages and challenges associated with the implementation of this course model in higher education, considering its implications for learners, instructors, and societal advancement.

This scholarly inquiry scrutinizes the formulation, implementation, outcomes, and impact of this course framework over an extensive period of a decade, probing the challenges and benefits experienced by both students and faculty members. This investigation is positioned within the broader ambit of the Open Knowledge Movement, exploring the educational potential of Wikipedia and Wikidata over a decade-long time span. It delves into the roles these platforms play in knowledge generation, digital literacy, critical thinking, and addressing educational disparities. By evaluating the long-term implementation of this course model, the study provides critical insights into the challenges and efficacy of integrating OERs into tertiary education curricula, offering an exhaustive overview of the pedagogical implications of this approach. Additionally, the paper underscores the necessity for future research and deliberates on the significance of Wikimedia platforms within educational contexts, particularly in the context of the burgeoning field of Generative AI applications.

Background

OERs in academia

In the milieu of UNESCO’s contemplation on humanity’s aspirations for 2030, the organization introduced a framework named the “Sustainable development goals” (SDGs) in 2015. These SDGs, a compilation of 17 global objectives, aim to chart a course towards a more sustainable and equitable future, receiving endorsement from the UN General Assembly in 2017. Notably, Goal 4 emphasizes the imperative of ensuring inclusive and equitable quality education and fostering lifelong learning opportunities for all, underscoring the significance of open and equal access to educational resources (UNESCO, SDG 4). This goal particularly accentuates the role of open education (OE), occasionally referred to as open education practices (OEP), open pedagogy (OP), or open educational resources (OERs), in fulfilling this SDG (Jha et al., 2019; Lane, 2017; Ossiannilsson, 2019; Tlili et al., 2020; Urbančič et al., 2019).

With the expansion of the Internet, the Open Knowledge Movement has gained momentum, fostering the adoption of OE, OEP, OP, and OERs in educational systems and academia (Cronin & MacLaren, 2018; Evenstein Sigalov & Nachmias, 2017; Hegarty, 2015; Lin, 2019). Academic literature presents a multitude of interpretations for these concepts (Cronin & MacLaren, 2018; Paskevicius & Irvine, 2019; Wiley & Hilton, 2018). Although a comprehensive exploration of these definitions is beyond the scope of this research, the review by Cronin and MacLaren (2018) provides foundational definitions. They discuss the challenges in precisely delineating Open Education, describing it as an umbrella term encompassing resources, tools, and practices aimed at enhancing educational access, effectiveness, and equality globally. Furthermore, they elaborate on the expansive nature of this term, encompassing not just educational policy, practices, resources, and pedagogy, but also the embedded values and the dynamics between educators and learners.

While Open Education has a longstanding history, the term Open Educational Practices emerged more recently, around 2007 (Cronin & MacLaren, 2018). The interpretations of OEP vary significantly, ranging from those focused on the creation and usage of OER to broader conceptualizations that include, but are not limited to, OER (Cronin & MacLaren, 2018). Regarding Open Pedagogy, scholarly consensus describes it as emergent academic practices advocating the use and production of OER, fostering open learning and teaching, collaboration through networked participation, and empowering learners in co-creating knowledge (Cronin & MacLaren, 2018).

Despite the diversity in definitions, there is a consensus in literature that advancements in technology, the Internet, and the Open Knowledge Movement have contributed to a significant surge in the adoption of OERs as a form of OE, OEP, and OP (Blumenstyk, 2015; Cronin & MacLaren, 2018; Hegarty, 2015; Lin, 2019; McDowell & Vetter, 2022; Paskevicius & Irvine, 2019; Wiley & Hilton, 2018). UNESCO defined OERs in 2002 as educational materials in the public domain or released under an intellectual property license permitting free use, adaptation, and distribution, with the Creative Commons licenses being predominant. One example would be the CC-BY-SA license used in Wikipedia, allowing unlimited usage, including commercially usage included, as long as credit is given (BY), and material is shared under original license it was released under (SA = Share Alike).

For many educators, the primary motivation for utilizing OERs is to alleviate the financial burden of textbooks; While others seek to create a pervasive, mobile learning environment, enabling access to materials anytime, anywhere (Hegarty, 2015; Lin, 2019). Additionally, some view the adoption of OERs as part of a broader pedagogical and ideological approach, valuing OERs not only for knowledge equity but also as a means to acquire essential skills, competencies, and literacies in a digitalized world (Cronin & MacLaren, 2018; Davis et al., 2023; Evenstein Sigalov & Nachmias, 2017; Hegarty, 2015; Lin, 2019; Wiley & Hilton, 2018).

Various stakeholders, including organizations, governments, educational institutions, policy makers, educators, researchers, and advocates of free knowledge, are actively engaged in creating OERs for teaching and learning purposes. Recent studies indicate positive trends in the use of OERs, although, as Lin points out, the mere availability of materials does not guarantee successful teaching and learning with OERs (Lin, 2019). Despite these promising trends, challenges persist, including low awareness of OERs among educators, difficulties in finding relevant OERs, assessing their quality, and addressing internet connectivity issues for learners from lower socio-economic backgrounds (Allen & Seaman, 2017; Lin, 2019).

As the utilization of OERs is a relatively novel phenomenon, there is a growing call within the educational research community for more empirical studies. These include investigations into the pedagogical effectiveness of OERs compared to traditional educational resources, understanding their efficacy and quality in teaching, and assessing their alignment with modern pedagogical approaches and methodologies. Hegarty’s, 2015 framework for open pedagogy, which lists eight attributes for successful OER integration in curricula, underscores the need for strategic planning for effective implementation. This framework includes participatory technologies, fostering openness and trust, encouraging innovation and creativity, sharing resources, building connected communities, creating learner-centered environments, practicing reflective teaching, and encouraging peer review. This strategic approach is deemed essential for the successful integration of OERs in educational contexts.

Wikipedia as a learning platform

Within the ambit of open pedagogy, the discourse around Wikipedia assumes critical relevance. Since its launch in 2001, Wikipedia has emerged as the largest endeavor of Open Knowledge in human history and, by virtue of its licensing, the most extensive open educational resource (OER). This volunteer-driven initiative, initially met with skepticism by many in the academic and educational spheres, has evolved into a highly frequented source of information, ranking among the top seven most visited websites globally. Wikipedia, alongside its 13 sister projects, serves millions of users, including those in offline settings through tools like Internet-in-a-box.²The COVID-19 pandemic further elevated Wikipedia’s stature, garnering acclaim from online organizations, researchers (McDowell & Vetter, 2020), activists, and individuals as a bastion of reliable information amidst a web plagued by misinformation, disinformation, fake news, and deep-fake technology. Wikipedia’s content has become increasingly prevalent on major social platforms, search engines, and within various AI agents.

The educational sector is not immune to this trend. Over the past two decade, a burgeoning number of educators have been integrating Wikipedia into their pedagogical and academic programs (Aibar et al., 2015; Davis et al., 2023; Dooley, 2010; Evenstein Sigalov & Nachmias, 2017; Konieczny, 2016; McDowell & Vetter, 2022; Ramjohn & Davis, 2020; Vetter et al., 2019). While Wikipedia is by no means infallible – entries can be inaccurate, biased, or absent, particularly in languages other than English–educators recognize its value as a platform for both teaching and learning, forming an integral part of the academic process. Initially, Wikipedia was employed to enhance information consumption skills, such as comparative research, critical thinking, and analysis. Subsequently, it has been utilized as a platform for collaborative knowledge creation, engaging a multitude of skills involved in this process (Evenstein Sigalov & Nachmias, 2017; Lauro, 2020). The pedagogical benefits of using Wikipedia as a teaching and learning platform warrant thorough academic investigation.

Wikipedia’s aspiration for high-quality, current, neutral, and well-referenced articles, created through a transparent and collaborative process, presents unique educational opportunities for both educators and learners (Aljawarneh, 2020; Arslan & Turk, 2023; Evenstein Sigalov & Nachmias, 2017; Herbert et al., 2015; Konieczny, 2007, 2016; Luo & Chea, 2020; Meseguer-Artola et al., 2020). As a quintessential Web 2.0 platform, Wikipedia’s pedagogical potential has been extensively studied, focusing on its capacity to actively and collaboratively engage learners in their knowledge construction (Aibar et al., 2013, 2015; Boulos et al., 2006; Konieczny, 2016; LaFrance & Calhoun, 2012; Mareca & Bordel, 2019; Minguillón et al., 2018; Naismith et al., 2011). This involves developing competencies such as digital literacy, collaborative skills, critical thinking, and academic literacy (Bordel & Mareca, 2019; Di Lauro & Johinke, 2017; Eteokleous et al., 2014; LaFrance & Calhoun, 2012; McKenzie et al., 2018; Selwyn & Gorard, 2016; Soler-Adillon et al., 2018; Staub & Hodel, 2016; Vetter et al., 2019; Zheng et al., 2015). Consequently, an increasing number of educators are exploring Wikipedia’s application in classroom settings, primarily as an alternative form of assessment. In this model, student performance is evaluated through knowledge construction on Wikipedia or its sister projects, offering an innovative alternative to traditional tests and assignments, a practice sometimes termed as “Wikidemic assignments”.

Although Wikipedia has been utilized in educational settings for at least 2 decades now (Davis et al., 2023), its integration into higher education remains relatively nascent (Evenstein Sigalov & Nachmias, 2017; Konieczny, 2014). As Evans noted, despite its pedagogical promise, the wiki phenomenon has largely not permeated classroom settings, either as a subject of research or a method of instruction (Evans, 2006). This observation was echoed by Konieczny in 2016, who noted that Wikipedia’s acceptance among academics and educators was increasing, albeit slowly and reluctantly (Konieczny, 2016). While there has been progress since 2006, and a growing contingent of educators are now endeavoring to embed Wikipedia into their curricula (Davis et al., 2023; Evenstein Sigalov & Nachmias, 2017; Konieczny, 2016; McDowell & Vetter, 2022), academia is only beginning to unveil its potential. The challenge remains to formalize strategies that implement the use of Wikipedia as a learning platform in a manner that promotes a deep sense of learning and enhanced skills through an engaging and positive learning process both in and out of class. Instructors are still grappling with how best to incorporate wikis into classroom settings and foster effective collaboration (Allwardt, 2011; Davis et al., 2023; Elgort et al., 2008; Johinke, 2020; Konieczny, 2014, 2016; Malik et al., 2023; Martin et al., 2023; Naismith et al., 2011; Ramanau & Geng, 2009; Smith, 2020; Vetter et al., 2019; Zou et al., 2020). Current research primarily focuses on specific case studies, with a dearth of long-term studies on the implementation of Wikipedia in higher education. This applies to its sister projects as well, with Wikidata, in particular, gaining recent attention from educators as a promising learning platform (Evenstein Sigalov & Nachmias, 2023; Evenstein Sigalov et al., 2023).

Wikidata as a learning platform

In 1998, Tim Berners-Lee conceptualized an advanced iteration of the World Wide Web, known as the Semantic Web or Web 3.0, which is occasionally identified as ‘Linked data’ (Evenstein Sigalov & Nachmias, 2023). The Semantic Web envisions transforming the Web into a global data commons, enabling applications to function atop an unbounded array of data sources through standardized access protocols. This vision proposes an expansion of the traditional Web to a Web of Data, where connections exist not only between documents and links but also among diverse entities and relations, facilitating machine assistance in human tasks (Berners-Lee, 1998). In the Semantic Web, the integration of structured data from numerous sources and the meaningful linkage of information segments are pivotal. This innovative web concept theoretically empowers both humans and machines to exploit a high-quality, current, and well-referenced knowledge base of interconnected data.

Among the various endeavors to actualize this vision, Wikidata, a sister project of Wikipedia, stands out as exceptionally successful at scale. Initiated in 2012, Wikidata was born from the ambition of several Wikimedians to address queries beyond the scope of conventional search engines, such as identifying the "10 largest cities with a female mayor (Erxleben et al., 2014; Krötzsch et al., 2007). Dr. Denny Vrandečić championed the idea that free knowledge should encompass searchable, analyzable, and reusable data (Vrandečić & Krötzsch, 2014), leading to the creation of a multilingual, open database that stores structured, linked data—a free “big data” knowledge base accessible to both humans and machines. As of January 2024, over 108 million items and countless structured statements describing them, are curated in Wikidata, making it not only the largest semantic knowledge base, but also one of the world’s most expansive OERs due to its open license (Evenstein Sigalov & Nachmias, 2023). Beyond cataloging human knowledge and linking data within Wikimedia projects, Wikidata also interconnects with external databases and initiatives, an aspect only beginning to be fully explored. What renders Wikidata a database of profound significance, teeming with "exciting possibilities"? (Erxleben et al., 2014). Beyond its integral connection to Wikipedia, Wikidata presents a vast structured and linked database amenable to querying, providing precise and relevant responses, surpassing the capabilities of traditional Web 2.0 search engines like Google. The valuable insights offered by Wikidata render it a fertile ground for research across various fields, particularly in science, technology, and the arts (Vrandečić & Krötzsch, 2014).

The Semantic Web, and Wikidata in particular, appear to offer humanity tools to navigate and manage the deluge of information characteristic of the digital era, facilitating meaningful data engagement previously unattainable. Wikidata has revolutionized the interaction between people and knowledge, introducing new learning possibilities (Evenstein Sigalov & Nachmias, 2023). This includes not only providing accurate answers to complex queries, but also enabling the visual representation of information. For example, one can instantly generate a timeline of Leonardo da Vinci’s works (Fig. 1) or a geographical mapping of the birthplaces of female physicists throughout history (Fig. 2).

Fig. 1

W or ks by Leonardo Da Vinci in a timeline, from Histropedia, a platform based on WD

Fig. 2

A map of the birthplace of women physicists in history

Thus, Wikidata emerges as a potent learning platform (Evenstein Sigalov & Nachmias, 2023), marking a significant stride towards realizing Berners-Lee’s Semantic Web vision. However, with its growing scale and success, Wikidata assumes a considerable responsibility. It is already being utilized by individuals, researchers, organizations, and AI agents like Siri and Alexa. In an era where younger generations routinely receive immediate, presumed-accurate responses from AI agents and Generative AI platforms such as ChatGPT, it becomes imperative for educators to ensure that learners are equipped with appropriate skills as digital citizens. In a world inundated with Information Explosion, Big Data, Machine Learning, and GenAI, fostering critical thinking, Data Literacy, and GenAI Literacy among learners is crucial. Engagement with Wikidata could address challenges related to data modeling, analysis, verifiability, completion, and systemic biases (Evenstein Sigalov & Nachmias, 2023).

Research goals and questions

In an academic environment where educators are increasingly recognizing Wikipedia’s potential as a teaching tool (Bayliss, 2013; Davis et al., 2023; Janio, 2014; Konieczny, 2016; McDowell & Vetter, 2022; Vetter et al., 2019), a novel course framework integrating Wikipedia into higher education syllabi was developed and executed at an Israeli University (name removed for peer review) in 2013. In this pioneering course, a global first, offered a for-credit, semester-long program where the creation of knowledge on Wikipedia and its affiliated projects constituted the primary mode of assessment. The course was designed to utilize Wikipedia and related Wikimedia platforms to enhance students’ academic and digital literacies, collaborative and communication skills, and data literacy, all while amplifying social impact. A second course was then designed to expand upon the original format, making it accessible to undergraduate students across various disciplines. Then a third course further evolved the model, incorporating not only Wikipedia but also Wikidata, with a special emphasis on data literacy, marking a global precedent in an academic setting. The three courses offered at the university since 2013 are:

• 1) “Wiki-Med” The School of Medicine; 2013 – 2022 (9 iterations).

• 2) “Wikipedia: Skills for producing and consuming knowledge” campus-wide; 2015 – 2018 (3 iterations).

• 3) “From Web 2.0 to Web 3.0, From Wikipedia to Wikidata” campus-wide; 2018 – 2023 (5 iterations).

This research examines the course model as implemented in these three academic courses over a long time period of a decade. It aims to share details of the course design, outcomes, and application insights. Consequently, the research questions are:

• 1) What were the design guidelines in developing a course model that integrates Wikipedia and Wikidata as learning platforms in a higher education curriculum?

• 2) What were the courses’ outcomes, in terms of students’ output and grades, as well as their perceived learning experience?

• 3) Considering an integrative examination of the course model implementation over a decade, what were the perceived challenges and benefits of implementing the course model in higher education for learners and instructors?

Methodology

Research design

Our research delves into the formulation and execution of a specialized course model centered on Wikipedia and Wikidata within the realm of higher education. It aims to critically assess this course model over an extensive duration of ten years, applied across three distinct courses, to gauge its effects on students, educators, and the broader society. Such an analysis necessitates a comprehensive approach, drawing from a variety of information sources (Creswell, 1998). A mixed-methods approach was employed to collate and analyze data from 17 iterations of these courses. Over this span, 616 students participated in the courses. The evaluation encompassed an examination of the OERs produced by the students and their academic performance metrics. Furthermore, a post-course questionnaire, responded to by 70% of the students (n = 429), was analyzed both quantitatively and qualitatively to extract insights into the pedagogical experiences, challenges, and benefits perceived by the students. The research additionally delves into the challenges and benefits for faculty, offering perspectives essential for future adaptation of this educational model.

Participants

This study encompassed participants who were students enrolled in 17 iterations of three distinct courses spanning from 2013 to 2023, totaling 616 students. A detailed enumeration of student participation per cohort is presented in Table 1. The initial course predominantly engaged first-year medical students. In contrast, the subsequent two courses attracted a more heterogeneous group of students from various disciplines, faculties, and programs, ranging from the Exact Sciences to Arts and Humanities. Owing to the courses’ status as cross-campus electives, the student demographic varied from first-year undergraduates to those in their fourth year. The linguistic diversity of the classes was notable, with most students being native Hebrew speakers, alongside individuals whose first languages included Arabic and Russian among others, thereby fostering a diverse learning environment. Participation in the end-of-course questionnaire was voluntary, with a specific query included to ascertain consent for involvement in the research. To maintain student privacy and address ethical concerns regarding the obligatory creation of knowledge or research participation, all personal data was rendered anonymous.

Table 1 Students per cohort, per course, per year between 2013 and 2023

Data collection

Data collection for this study spanned a decade, with analysis phases conducted from October 2021 to April 2022, and subsequently from August 2023 to October 2023. To gather and scrutinize the data, a mixed-method approach was employed, focusing on three primary sources: (1) The courses’ profiles on the Programs & Events Dashboard, a Wikimedia tool utilized for organizing and monitoring student work and impact, primarily addressing the second research question concerning the outcomes and influence of the course model; (2) Assessment of final course grades, chiefly to explore the second research question pertaining to course results; and (3) Post-course questionnaire, principally aimed at addressing the second and third research questions, focusing on student perspectives regarding their learning experience, encompassing benefits and challenges. Research Design principles were employed to address the initial research question, drawing on the experiences and insights of the course faculty in developing the course structure. The experiences, insights, and perceptions of the course faculty were also integral to exploring the third research question.

Data analysis

Performance assessment

An assessment of students’ comprehensive performance was executed at the conclusion of each semester, encompassing an analysis of the quality of articles produced in Wikipedia, contributions to Wikidata projects, items included, and formulated queries. Given that Hebrew Wikipedia lacks an article quality ranking system, unlike its English counterpart, and considering the extensive criteria necessary for a detailed quality assessment, conducting a thorough qualitative analysis of the articles was beyond the faculty’s capabilities and resources. Consequently, quality assessment was based on a rubric initially devised for the first course and subsequently modified for the second. This rubric was also employed by students for peer and self-assessments. For the third course, the rubric was further extended to evaluate engagement with Wikidata, aligning with the specific requirements of the assignment.

The rubric emphasized the use of high-quality references in Wikipedia and Wikidata entries and the adoption of an encyclopedic writing style. This assessment considered various elements covered in the course, including maintaining a neutral, unbiased, and inclusive perspective; constructing balanced and coherent arguments in an accessible language; adhering to Wikimedia community norms for encyclopedic writing; mastering technical aspects of the platforms, and respecting copyright laws. Students were encouraged to produce meticulous, well-considered work, given the immediate public availability of the information, with special attention to the repercussions of presenting non-referenced or inaccurate content.

In evaluating the quality of students’ submissions, particular attention was paid to the finer details that distinguish a well-crafted Wikipedia article or a meticulously modeled Wikidata item. For Wikidata projects, the evaluation of students’ work quality encompassed the precision and comprehensiveness of the queries, the use of visualization tools for data exploration, and the quality of the items added to Wikidata, including high-quality references and well-structured modeling of items. In addition to the rubric-based quality assessment, which focused on the practical application of the course content, students’ overall performance also included an evaluation of their collaborative endeavors. This encompassed peer reviews, group projects, contributions to both in-class and online discussions, classroom presentations, and overall active engagement. It is noteworthy that non-native Hebrew speakers faced inherent challenges in contributing to Hebrew Wikipedia. For these students, greater emphasis was placed on their effort and adherence to encyclopedic writing standards, rather than on linguistic fluency or grammatical accuracy. Conversely, in the Wikidata project, multilingual proficiency was advantageous, as it enabled students to contribute data in various languages, thereby enriching the items with more comprehensive information.

Post-course questionnaire

The end-of-course questionnaires were initially developed during the inaugural course (Evenstein Sigalov & Nachmias, 2017), and subsequently modified for subsequent courses (Evenstein Sigalov & Cohen, 2020; Evenstein Sigalov et al., 2023). Out of the 616 participants across 15 iterations, 429, representing 70%, completed the questionnaire. Two iterations were excluded from this data collection. During the 2019–2020 iteration of the third course, the outbreak of COVID-19 led to a decision to not increase students’ workload with the questionnaire. Furthermore, the second cohort of the first course (2014–2015) was also omitted. This cohort faced internal administrative challenges, resulting in only six students enrolling and participating. Due to the small class size, the course was conducted in a manner akin to a private course or seminar, deviating from the standard elective format, and thus not accurately reflecting the typical course model and operations. To maintain methodological integrity, the results from this iteration were excluded to prevent any potential distortion of the overall findings. Appendix Table provides a year-wise breakdown of the number of students who completed the questionnaire. Table 2 presents a summary of the gender and native language of the respondents who filled out the questionnaires, except for the first iteration, where socio-demographic data were not collected as part of the questionnaire.

Table 2 Socio-demographic details on students who filled out the questionnaire

The questionnaire utilized for this study was bifurcated into two distinct sections. The initial segment concentrated on the students’ evaluation of their educational experience, their perceptions of contributing to Wikipedia and Wikidata, their self-assessed knowledge post-course, and their overall appraisal of the course. In this segment, students rated various aspects of the course using a 1–5 Likert scale, supplemented by optional open-ended queries. The latter half of the questionnaire solicited student feedback on the course faculty and their instructional sessions, again allowing for additional insights through optional open-ended questions. Quantitative responses were consolidated to present data from each individual course and cumulatively across the 15 iterations. From the qualitative responses, 679 statements concerning the students’ perceived educational experience were extracted. An iterative methodology was employed to categorize similar codes and refine a hierarchical structure elucidating the qualitative remarks about students’ learning experiences, with allowance for statements to recur in multiple categories.

To ensure the reliability of the coding process, 30% of the statements underwent secondary analysis by an alternate coder. This process yielded a high level of agreement, evidenced by a Cohen’s Kappa of 0.94. The amassed data was organized into categories and, where applicable, sub-categories, followed by a statistical analysis. Select quotations were chosen to exemplify the various categories and sub-categories that emerged from the coding process. The qualitative data primarily addressed the second research question, enhancing the quantitative data and providing deeper insights into the students’ learning experiences. From the 679 statements about students’ learning experiences, those explicitly addressing perceived benefits and challenges were segregated for separate analysis, employing a similar iterative process to develop a category tree, aimed at addressing the third research question.

Each cohort also included participation in a session devoted to student presentations, where they showcased their work and reflected on their learning journey throughout the course. These presentations, a mandatory course component, were not all systematically coded and categorized, but served as a reflective tool for the faculty. Notably, these sessions often revealed anecdotal information not captured elsewhere, aiding in a more nuanced understanding of students’ perceptions, particularly in comparison with the data gleaned from end-of-course questionnaires. These presentations also provided insights into what aspects of the course were effective and which were not. Alongside questionnaire data, these presentations informed faculty assessments of the course model, guiding decisions for modifications in subsequent iterations and contributing to the faculty’s understanding of the course model’s strengths and challenges. Since the presentations were not methodically analyzed, the insights derived from them are shared anecdotally, either to reinforce certain conclusions or to offer additional perspectives.

Findings

Course model & design

The formulation and execution of an innovative course model centered on knowledge creation proved to be an intricate and ongoing endeavor. This complexity arose from several factors: firstly, the absence of any global precedent for an academic course of this nature. While there have been instances of Wikipedia assignments being used in classroom settings, systematic efforts to institutionalize Wikipedia engagement through a semester-long course in the academic sphere have been scarce. Secondly, the course model demanded a design that was both adaptable and scalable. This was evident as the inaugural course was tailored to a specific academic discipline (medicine), but subsequent courses were designed to be inclusive of students from a diverse range of fields. Thirdly, there was a need to address administrative considerations. In all iterations of the course, resources were constrained, influencing communication methods beyond the classroom, the support framework, and the assessment processes. Lastly, the course model necessitated the navigation of various pedagogical challenges, some of which were unique to an academic context. To ensure the course’s effectiveness and positive reception, its design had to reconcile the sometimes divergent objectives and distinct requirements of three key stakeholders: the faculty (and its commitment to academic rigor), the students, and the Wikimedia community, which played a supportive role in the course.

Main goals and learning objectives

The course was structured around the following educational objectives and aims:

• Introduce students to various stages of the Web’s development, actively involving them in both the consumption and creation of knowledge through different Wikimedia projects

• Enhance students’ scholarly and digital proficiencies, cultivate skills in collaboration and communication, and bolster data literacy

• Augment students’ capabilities in evaluating online information, encompassing aspects such as copyright, gender and knowledge disparities, and bias, while underscoring themes of diversity, inclusion, and knowledge equity

• Encourage students to generate high-quality content for Wikimedia projects, thereby contributing to the augmentation of open educational resources (OERs) for future learners and the wider public

• Foster a positive societal impact and endorse the ethos of community contribution

• Create an enriching and positive educational experience for the students

Design guidelines

In pursuit of these objectives, the course was founded on a set of guiding design principles:

• Emphasis on active learning and collaborative engagement The global trend of declining lecture attendance, exacerbated by COVID-19 restrictions, coupled with the challenges of online learning such as disconnection, ‘zoom-fatigue’, and concentration difficulties, necessitated an interactive approach. The course was designed to be engaging, with student participation in discussions, collaborative work in small groups, and peer assessment to provide constructive feedback.

• Student-centric approach with a focus on depth, iteration, and reflection The inherent complexity of Wikimedia projects often presents a steep learning curve that can deter newcomers. To counter this, the course adopted a comprehensive learning process, allowing ample time for students to absorb, experiment, enhance, and introspect their learning journey. This approach stands in contrast to courses that merely use Wikipedia assignments as an alternative form of assessment.

• Integration of the wikimedia community Both local and global Wikimedia contributors played a vital role in the course, serving as guest speakers and mentors. This interaction not only provided students with exposure to the community’s diverse volunteer base but also ensured community support for the course.

• Commitment to diversity, combating bias, and advancing knowledge equity Despite Wikipedia’s efforts towards inclusivity and neutrality, biases persist (Ford & Wajcman, 2017; Hargittai & Shaw, 2015; Konieczny & Klein, 2018; Wagner et al., 2015), with a contributor demographic predominantly composed of Western white males and only 13% women³⁴⁵⁶. This imbalance affects the range of topics covered and leads to inherent biases in the content. The course design, therefore, emphasized recognizing and addressing these knowledge gaps and inequities, with a particular focus on the Gender Gap (Ramjohn & Davis, 2020).

• Promoting reusability The course was designed to be a replicable and scalable model that could be adopted by educators, faculties, and academic institutions globally, aligning with the principles of the Open Education movement.

Course structure

The inaugural and subsequent courses were structured with an initial introductory session, 11 core sessions to impart essential Wikipedia skills, and a concluding session dedicated to student presentations. These presentations facilitated reflection on the learning process and informed improvements for future course iterations. Given limited faculty resources, a system of peer evaluations and, in certain instances, a single self-assessment per article, was implemented. Research suggests this approach yields comparable outcomes to traditional teacher assessments (Sadler & Good, 2006) and has been adopted in massive open online courses (MOOCs) by platforms like Coursera (Piech et al., 2013). Additionally, two faculty-led sessions were introduced to analyze select examples in class. In the third course, which divided its focus between Wikipedia and Wikidata, the format consisted of two main modules of six sessions each, followed by a final presentation session. A period for revision and resubmission of work was allocated, given the public visibility of students’ contributions, with an emphasis on creating quality materials for future learners and the general public. Each course possessed unique characteristics, detailed below.

Cours 1: Wiki-Med

The Wiki-Med course, initiated in 2013 as a pioneering venture worldwide, had nine iterations at TAU, available annually to medical students. In 2015, to address the Gender Gap on Wikipedia, a new category for Women’s Health was introduced by students and has been continuously populated since. From 2015, an introduction to Wikidata was included, evolving from a single session to a three-session module.

The course’s eighth iteration coincided with the COVID-19 outbreak, necessitating a shift to an online format. A hybrid model of synchronous and asynchronous sessions was adopted, with adaptations such as small group assignments via Zoom, communication through WhatsApp, and reduced workload with group submissions. Though this resulted in fewer student-generated articles, it maintained a positive learning experience and high content quality. Primary research in 2016 examined the first iteration’s impact, with follow-up questionnaires and interviews conducted two years later (Evenstein Sigalov & Nachmias, 2017). Mendes et al. (2021) extended this research to other iterations, confirming the enduring relevance of the initial findings and the combined social impact (Mendes et al., 2021).

Course 2: wikipedia: skills for producing and consuming knowledge

The second course, open to all undergraduates at TAU, expanded the model for a multidisciplinary audience, maintaining workload comparability to other electives through group and individual assignments. In 2019, after three iterations, further research was undertaken to evaluate the course’s outcomes, effectiveness, and students’ learning perceptions, with findings presented at the 2020 AERA conference (Evenstein Sigalov & Cohen, 2020).

Course 3: from web 2.0 to web 3.0, from wikipedia to wikidata

The third course, influenced by the increasing focus on Wikidata, split evenly between Wikipedia and Wikidata modules. It encompassed data curation and extraction on Wikidata, addressing topics such as data verifiability, completeness, and biases. This course, unique in featuring a Wikidata project, adapted to an online, hybrid format during the COVID-19 pandemic. Research on using Semantic Networks and Wikidata as learning platforms is scarce. Thus, this course served as a vital exploration in this area. The first three iterations were examined, with results published in “Open Educational Resources in Higher Education” by Springer Nature (Evenstein Sigalov et al., 2023). Additional research, based on aggregated data from all five iterations and focused on faculty’s data-driven decision-making, is underway.

Assessment model

The structural design and evaluative framework of the course are elaborated upon in Figs. 3, 4 below. Across its various iterations, the course model largely maintained a consistent format, with the notable exception being the heightened emphasis on Wikidata during the third course, where the capstone assignment transitioned from a Wikipedia article to a Wikidata project. The practice of peer review was a constant feature across all course iterations. However, self-assessment was selectively employed, being incorporated only in the initial iterations of the first course and throughout the second course, but it was omitted in the third course. When evaluated against Hegarty’s framework for the successful integration of open educational resources (OERs), the course model appears to encompass all eight attributes identified as crucial for success. The course was indeed strategically planned and designed with success in mind (Hegarty, 2015).

Fig. 3

Course model

Fig. 4

Assessment model

Outcomes

Course outputs and grades

A decade of instruction culminated in the development of three courses, spanning 17 iterations, with the involvement of 616 students. This period saw the creation of 2240 Wikipedia articles and Wikidata items, in addition to the editing of 7308 articles by students. These student-contributed articles have amassed over 75 million views from the public. Notably, since 2015, at least half of the Wikipedia articles generated were selected from the Women-in-Red list⁷, thereby contributing to reducing the gender disparity on Wikipedia. Comprehensive data concerning the number of articles created and edited, the corresponding page views, and final student grades are outlined in Table3 below. For a more detailed breakdown, Appendix Table provides in-depth data, including a methodology for the calculation of aggregated page views.

Table 3 Course outcomes: dashboard metrics & final grades

Students’ learning experience

The evaluation of students’ learning experiences was initially grounded in the analysis of quantitative data from questionnaires completed by participants across 15 iterations (N = 429). While a core set of questions remained consistent, adaptations were made for certain iterations. The qualitative feedback provided further insights into the learning experiences, illuminating specific observations.

The course structure, encompassing organization, efficiency, logical flow, and balance, was highly rated by students, scoring between 4.3 and 4.7 on a 1–5 Likert scale. The course assignments received ratings in the range of 4.0–4.5, with students acknowledging their significance in the learning process and the value of class and peer feedback. One student remarked, “The peer-review process was extremely beneficial and will be a crucial skill in our future careers, so it’s advantageous to start learning it now.” The self-assessment component of the assignments, however, received a lower rating of 3.8, being seen as valuable by some but unnecessary by others. Consequently, this aspect was phased out after six iterations.

Students’ evaluation of their learning outcomes averaged between 4.0 and 4.1, with slightly lower ratings for the third course due to the more technical nature of Wikidata. Their perception of Wikipedia as an innovative, collaborative platform was high (4.0–4.3), while their view of it as a reliable and neutral source was slightly lower (3.8–3.9). This variance was anticipated due to the differing initial levels of trust among students. The change in perception of Wikidata was rated at 4.1–4.2, expectedly higher since many were unfamiliar with Wikidata before the course.

In assessing their skill enhancement, students gave the lowest ratings, ranging from 2.7 to 3.6. Most noted minimal improvement in computer skills (2.7), which aligns with many already possessing these skills upon enrolling. As one student expressed, “My pre-existing computer skills meant that the course offered enhancement but not a major transformation”. The improvement in other skills, including digital, academic, collaborative, and data literacy, received slightly higher ratings (3.3–3.6). However, these were lower compared to other course ratings, contrasting with written and oral feedback indicating significant skill development in these areas. This discrepancy suggests the possibility of misinterpretation of questions or difficulty in attributing skill development directly to the course.

In the overall course evaluation, students rated the course highly (4.0–4.3) for being engaging, inspiring, satisfying, community-oriented, and offering a positive learning experience. Over half shared their course outcomes with family and friends, with some considering the course a pivotal part of their undergraduate studies. The likelihood of continued Wikipedia contributions was rated at 3.5, reflecting the time and effort required. For the Wikidata-focused course, many students expressed intentions to integrate Wikidata into their academic pursuits. The overall quality of the course was rated at 4.2, indicating a generally positive learning experience. All aggregated quantitative questionnaire data per course is included in Appendix Table.

To further elucidate the rated questions, qualitative comments were systematically coded and analyzed. As outlined in the methodology, all comments (N = 679) underwent an iterative coding process, resulting in a refined category tree, which was then subjected to statistical analysis. A chi-square goodness-of-fit test yielded significant results (X² (4) = 622.692), with additional tests on sub-categories also showing statistical significance in most cases. The qualitative feedback generally paralleled the themes of the rated questions. Five main categories emerged: course structure (n = 488), course effect (n = 462), improved skills (n = 205), assignments (n = 120), and collaborative learning (n = 40). Each category was further divided into sub-categories focusing on aspects to retain or improve.

Regarding course structure, the most frequent comments (n = 488) advocated preserving certain elements. Students praised the course’s organization, efficiency, and clarity, with weekly summary emails being particularly appreciated. The course was commended for intriguing lectures, guest speakers, and a well-balanced mix of synchronous and asynchronous sessions, which included supportive self-paced learning materials. Suggestions for improvement varied, with some requesting more practical in-class exercises or less technical content, while others desired more depth on certain topics. Preferences for synchronous versus asynchronous sessions diverged, and some students struggled with guest lectures in English, requesting translations. Requests for more time dedicated to Wikidata were addressed by incrementally adding sessions focused on it, leading to the introduction of a third course in 2018 with an equal emphasis on Wikipedia and Wikidata (see 5.3.2 for additional information on faculty’s perception).

Feedback on the course’s effect (n = 462) highlighted aspects to maintain and enhance. Students described the course as engaging, enjoyable, and enriching, noting it broadened their perspectives and skill sets, making it a distinctive educational experience. Personal growth, overcoming challenges, and the opportunity to contribute to the community were particularly valued. Suggestions for improvements included the introduction of an advanced course, particularly on Wikidata, and some students expressed a desire for the course to be mandatory.

In terms of skill development (n = 205), students reported enhancements in digital and academic skills, writing proficiency, and data literacy. While some entered with strong digital competencies, they still acknowledged gains in other areas. Feedback on assignments (n = 120) revealed mixed views. While many found the assignments beneficial for learning, others perceived them as stressful or too demanding in terms of time and effort. Comments on collaborative learning (n = 40) mainly supported group work and peer review, with a few suggesting improvements. Group work was largely seen as conducive to the learning experience, though some faced challenges with unengaged group members. Peer review was praised for its role in fostering critical assessment skills and constructive feedback. Table 4 presents the category tree mapping these comments, with example quotes for each category/sub-category.

Table 4 Statistical analysis of Students’ comments on their learning experience in the course

The overall student experience was aptly summarized by a graduate from the first Wiki-Med course, who reflected on the broader academic context and the course’s role in addressing contemporary educational challenges:

“Overall, I am dissatisfied with my academic experience. I feel that the system struggles to adapt to the reality of freely available information for our generation and the challenges of 21st-century teaching and learning. I believe this course, despite being elective, is instrumental in driving this necessary change. Its elective nature provides faculty with the flexibility to explore innovative teaching methods and allows students to engage more patiently with experimental pedagogies, extracting valuable lessons from the experience”.

Challenges and benefits for learners and faculty

An exhaustive analysis of course outcomes, student feedback, and end-of-course presentations revealed both challenges and benefits in integrating Wikipedia and Wikidata into an academic setting for students and faculty.

Challenges & benefits to students

Of the 679 statements regarding student learning experiences, 210 (31%) focused on learning benefits and 156 (23%) on learning challenges. A rigorous process of coding and categorization followed, resulting in a well-defined category tree, which underwent statistical analysis. Chi-square goodness of fit tests confirmed the significance of these findings (Benefits: X² (5) = 215.369, p = 0.000; Challenges: X² (4) = 150.789, p = 0.000). The analysis identified 190 statements specifically about challenges, categorized into five key areas: course management (n = 98), noting sessions as overly technical, calls for more practical time, ineffective group work, redundant tasks, and the need for timely faculty feedback; resource intensity (n = 41), highlighting the course’s demanding nature compared to other electives; technical challenges (n = 40), predominantly with Wikipedia’s translation tool and adapting to Wiki syntax and Wikidata queries; language barriers (n = 7), mainly for Arabic-speaking students; and limited topic flexibility (n = 4), with some resistance to mandated focus on the Gender Gap. Subsequently, students were given the option to opt-out of this focus. Table 5 presents these challenges with exemplifying quotes.

Table 5 Statistical analysis of Students’ comments on the course’s challenges

In contrast, 520 statements related directly to course benefits, classified into six categories: enhanced capabilities (n = 183), including digital, academic, writing skills, and critical thinking; unique learning experience (n = 144), describing the course as engaging and horizon-expanding; effective course management (n = 59), with well-organized, efficient sessions and meaningful feedback mechanisms; community contribution (n = 49), where students valued the opportunity to give back; personal interest (n = 45), appreciating work on personally relevant topics; and personal growth (n = 40), such as increased self-efficacy. Table 6 presents these benefits with exemplifying quotes.

Table 6 Statistical analysis of Students’ comments on the course’s benefits

A student reflected two years post-completion:

“In our third year, we must write a research paper, crucial for our academic and research careers, but we lack formal training in academic writing and research. Only now do I realize the course’s relevance to these skills, advocating for its mandatory inclusion for teaching academic and research literacy.”

Challenges & Benefits to faculty

This research, primarily centered on the learning outcomes and experiences of students, also offers valuable insights into the teaching experiences within the course, particularly the challenges and benefits encountered by the faculty over a decade of executing the course model. Faculty confronted a diverse array of hurdles to overcome, broadly categorized into administrative, academic, and unforeseen challenges. Each category posed distinct obstacles that needed to be surmounted to ensure effective learning and teaching experiences. A key administrative challenge was the intensive requirement of engaging students in active learning, demanding a proactive teaching approach and consistent weekly efforts to maintain student participation and engagement. Managing the peer-review process presented another significant challenge, requiring meticulous coordination to effectively involve all students in a timely manner. Furthermore, monitoring student progress and interaction across a variety of platforms including Moodle, emails, social media, Wikipedia, and the Dashboard, added layers of complexity to the administrative workload, necessitating significant organizational and time management skills.

Academically, the pioneering nature of the course design, devoid of established precedents or models, presented its own set of challenges. These included convincing academic authorities of the course’s merit and securing their endorsement, as well as fostering a successful collaboration with the Wikimedia community. Adapting to the diverse needs of student cohorts with varying backgrounds, academic levels, and technical abilities required a nuanced approach to session design and student support. Additionally, faculty grappled with balancing the course’s ambitious learning objectives against the differing needs of stakeholders, including students, academia, and the Wikimedia community. A notable challenge was the absence of campus-level support for non-Hebrew speakers, compounded by the lack of teaching assistants due to resource constraints. This particularly impacted feedback provision on student work, necessitating practical in-class interventions and an extensive peer evaluation process. Addressing feedback about the course’s demanding workload while maintaining learning objectives led to the adaptation of assignments into small group tasks, reducing content production but fully addressing the issue. Continuously implementing changes based on student feedback and faculty observations required annual adaptability and an open-minded approach.

Unforeseen challenges, such as the global pandemic, demanded a swift pivot from in-person to online teaching, entailing a complete redesign of the course for synchronous and asynchronous delivery, while also addressing students’ diminished capacities and emotional stress. Catering to a group of first-year Arabic-speaking students presented unique requirements, necessitating specific adjustments and additional support. The emergence of Generative AI introduced a novel challenge, prompting the allocation of class time to discuss its implications on Wikimedia platforms and fostering Generative AI literacy among students. Table 7 below demonstrates the different challenges that emerged from 17 iterations of the course model.

Table 7 Challenges to implementing the course model from faculty’s perspective

Despite these challenges, the course model also yielded significant benefits identified by faculty across academic, social, and professional domains. Academically, the course achieved its goals and learning objectives, leading to observable improvements in students’ knowledge consumption and production, digital and academic skills, data literacy, and exposure to critical topics such as knowledge construction processes, copyright issues, gender gaps, systematic bias, and DEI (diversity, equity, and inclusion). The course’s unique assessment model, encompassing group work and peer evaluation, not only enhanced student skills but also reduced faculty workload.

Socially, the course promoted the development and use of Open Educational Resources (OER), advancing open knowledge and educational practices in higher education. It served as a practical exemplar for championing DEI in academia, addressing issues like the Gender Gap and systematic bias, and supporting a diverse student body, especially Arabic native speakers. The academic output from the course had a significant social impact, as evidenced by the 75 million pageviews garnered by student-created content. Additionally, the course model inspired similar educational initiatives globally, further amplifying its social benefits.

Professionally, faculty experienced numerous benefits. The course facilitated their professional development through specialization in a unique subject area, enabling the design and scaling of three interconnected courses that expanded their experience and experimentation. Faculty’s engagement in academic research related to the course allowed them to connect with a global community of professionals with similar interests, supporting faculty worldwide in similar endeavors. The successful implementation of the course model, aligned with faculty values and having a clear social impact, provided self-motivation, encouraging them to persistently enhance the course and expand their work both locally and globally. Table 8 highlights the varied benefits experienced by faculty across the 17 iterations of the course model.

Table 8 Benefits to implementing the course model from faculty’s perspective

Discussion, conclusions and reflections

Despite initial growing pains in early iterations of each course, the course model is widely regarded as a success by faculty, students, as well as the Wikimedia community. It effectively achieved its primary objectives, including enhancing students’ information literacy and digital citizenship, honing various skills, producing high-quality online content, and reducing knowledge gaps and biases. The course also promoted active learning and student engagement, offered a novel pedagogical model scalable and adaptable, and facilitated a rewarding learning experience for students. However, as Boulos et al. (2006) suggest, continual systematic evaluation is essential to understand the benefits and limitations of new technologies (Boulos et al., 2006). Our collective experiences (Evenstein Sigalov & Cohen, 2020; Evenstein Sigalov & Nachmias, 2017, 2023; Evenstein Sigalov et al., 2023) have informed the refinement of the course model, impacting teaching methods, assessment models, and consequently, student learning experiences. Each year, the courses evolved to accommodate more efficient learning experiences, emerging technologies, and unforeseen needs, such as those brought on by the global pandemic .

The number of participants directly influenced the overall contribution to the class, impacting the morale and support of all stakeholders—faculty, students, Wiki community, and policy decision-makers. Therefore, innovative strategies are needed to scale up, increase student enrollment per course, and enhance its social impact (Dees et al., 2004), all while maintaining manageability. As shown in academic literature, well-managed peer evaluation methodologies can scale up, alleviate faculty workload, and enhance student learning processes (Alqassab et al., 2023; Double et al., 2020; Jonsson et al., 2015). However, other studies argue that peer evaluations do not necessarily reduce faculty workload but may increase it (Bouzidi & Jaillet, 2009; Popova & Kolesova, 2019; Zlabkova et al., 2021). This is indicative of the need for further refinement of peer evaluation practices. Additional research is also needed to optimize self-assessment techniques (Andrade, 2019; Yan et al., 2020), which our students found ineffective and was subsequently removed in later iterations. While peer assessment enhanced learning, considering the course workload, alternative engagement and assessment methods are crucial. These new models should aim to reduce workload and assignments, while maintaining the benefits of collaborative learning and constructive feedback. The course model’s applicability in other academic settings, such as MA and PhD studies, and its expansion to Wikipedia sister projects, particularly Wikidata, and potentially Abstract Wikipedia and WikiFunctions, should also be explored. Research on Wikidata’s potential as a learning platform suggests its significant ability to revolutionize learning, teaching, and research while promoting data literacy (Atenas et al., 2023; Evenstein Sigalov & Nachmias, 2023; Evenstein Sigalov et al., 2023), a crucial skill in today’s data-driven digital landscape.

Data literacy and related skills enhanced or gained through engagement with Wikipedia and its sister projects, are increasingly vital, especially in the context of the rise of Generative AI (GenAI). Artificially intelligent machines in the form of applications or online services, such as ChatGPT, seem to offer quick answers in an easier, more accessible manner, which mimics interactions with humans better than ever before (Pataranutaporn et al., 2023; Pavlik, 2023; Susarla et al., 2023; Van Noorden & Perkel, 2023). Large Language Models (LLMs) powering GenAI applications, are trained on Wikipedia (Touvron et al., 2023.), and according to Selena Deckelman, CPTO at the Wikimedia Foundation, “every LLM is trained on Wikipedia content, and it is almost always the largest source of training data in their data sets”⁸. It is therefore even more essential that the information available through the platform is accurate and reliable, as it is being re-used by third-party applications.

In this context, it is worth mentioning that a new type of literacy is needed for today’s digital citizens to effectively navigate information – GenAI literacy. As AI is a not new phenomenon, academic literature exists on AI literacy (Ng et al., 2023); however GenAI is relatively new, and requires a different set of skills and capabilities. As this emerging new literacy is being defined and explored by academics around the world (Kreinsen & Schultz, 2023), skills gained by working on Wikimedia platforms are helpful in the context of this new literacy for a variety of reasons. GenAI applications currently operate as a black box. When asking it a question, the provenance of the information is mostly unknown (even in paid versions). Thus, in essence, GenAI applications create another barrier between users and the sources of knowledge, unlike a good Wikipedia article, where every sentence is referenced, so users can verify the validity of each piece of information. When asked to provide references, GenAI is known to fabricate these. The phenomenon, at times called “hallucinations” (Fui-Hoon Nah et al., 2023), means that GenAI can invent academic references that look legitimate, but are non-existent.

While companies that power GenAI applications are working on solving the issue of information provenance, there is the issue of GenAI perpetuating bias and knowledge gaps that are already well recorded online, including in Wikipedia (Alasadi et al., 2022; Evenstein Sigalov et al., 2023; Ferrara, 2023; Ford & Wajcman, 2017; Kirk et al., 2021; Konieczny & Klein, 2018; Perrotta et al., 2022; Stokel-Walker & Van Noorden, 2023; Thakur, 2023). As LLMs are trained on massive amounts of information from the internet, regardless of reliability of sources, it is also trained on quite a bit of misinformation, disinformation, fake news, deep fake, and does not consider missing information. Additional issues are privacy (Behnia et al., 2022; Peris et al., 2023; Staab et al., 2023), copyrights (Abd-Alrazaq et al., 2023; Park, 2023; Peng et al., 2023; Sallam, 2023), and perpetuating socio-economical differences. The latter is a notable issue, as GenAI platforms are mostly run by for-profit companies, with many requiring a fee for access, advanced features or accurate and up-to-date information. This practically means that only resource-rich entities and individuals have access to the most accurate and high-quality information. This, in itself, stands in sharp contrast with the values of Wikipedia, the Wikimedia movement, and the free and open knowledge movement, which promote access to information to all as a basic human right. Critical thinking, critical ignoring (Kozyreva et al., 2023), and a deep understanding of how these machines work and how to properly work with knowledge and data, are critical for survival in today’s world. These skills are exactly those that are improved by engaging with Wikipedia, both as a consumer and as a producer of knowledge. Notwithstanding various challenges posed by GenAI applications, they also hold various opportunities to learners, and additional research is required to explore how GenAI could be properly utilized to enhance work with Wikipedia and Wikidata in the classroom.

In conclusion, our ongoing investigation into this course model is motivated by its proven effectiveness and significant influence on both students and broader society. However, it’s important to acknowledge that this approach to integrating Wikipedia and Wikidata into academic programs, and fostering the creation of Open Educational Resources (OERs) with social impact, is not the sole method available. Despite its effectiveness and value to students, this model remains relatively uncommon, likely due to the challenges in securing academic approval and sustaining it without adequate resources and support. This model’s development followed years of less ambitious experiments with alternative models, including different assessment methods. It is persistence and a touch of fortune, that allowed us to expand these efforts.

Stepping back to evaluate this course model, it is essential to recognize it as just one among many viable approaches. Alternative models could involve appointing a Wikimedian-in-Residence, as practiced by the University of Edinburgh and other institutions in the UK and the US, or organizing edit-a-thons to address knowledge gaps, in partnership with local academic libraries or other cultural institutions (Galleries, Libraries, Archives, and Museums). The choice of model for engaging with Wikipedia and Wikidata to create OERs and generate social impact should be guided by a commitment to continuous exploration, experimentation, learning from failures, and iterative improvement. Despite its limitations, this study aims to motivate educators, OER experts, and academic institutions to adopt either this model or explore new ones that yield a positive societal impact. We especially hope to inspire the integration of not only Wikipedia but also Wikidata in educational and academic settings, as well as the utilization of Generative AI applications to enhance these educational endeavors, all of which must be further examined in future research.

Appendix

See Tables 9, 10 and 11.

Table 9 Number of students per cohort, per course, who filled out the questionnaire between 2013 and 2023

Table 10 Course outcomes per cohort

Table 11 Students’ rating of quantitative questions in the questionnaires per course