Technology-based assessment of collaborative problemsolving skills: a bibliometric analysis and review

Journal of Computers in Education (2026) 13:389–410 https://doi.org/10.1007/s40692-025-00361-2 Technology-based assessment of collaborative problemsolving skills: a bibliometric analysis and review Langat Gilbert Cheruiyot1 · Gyöngyvér Molnár1 Received: 1 December 2024 / Revised: 25 February 2025 / Accepted: 15 March 2025 / Published online: 30 May 2025 © The Author(s) 2025 Abstract Classroom instruction in the 21st century embraces collaborative problem-solving as a vital student skill. This study systematically reviews the existing literature on this topic to identify and analyze the current trends in technology-based assessment of CPS skills research. For this study, bibliometric analysis was conducted using Vosviewer analysis. Moreover, visualization and bibliographic software programs were utilized to examine the present status of CPS research. The study sample consisted of Web of Science articles published between 2014 and 2024 (N = 2,247; exclusion criteria: non-English-language studies published before 2014; inclusion criteria: meta-analysis articles included). We used the following search keywords: CPS skills, collaborative, and technology-based assessment. The findings revealed that the publication in the field of CPS began declining after peaking in 2019, with a particularly sharp drop in 2024. The findings also showed that Gwo-Jen Hwang was the most influential author, having an h-index of 4, a g-index of 4, and a total citation of 94 citations across four publications. Furthermore, most research concentrated on junior secondary school learners, indicating a particular emphasis on this group, according to a systematic evaluation of 20 publications that satisfied the specific exclusion criteria. An examination of keyword patterns showed that, between 2014 and 2024, the emphasis of research on problem-solving and collaborative learning significantly changed, with distinct terms becoming less and less frequent. The most common method in CPS assessments was the human-to-human interaction approach. Gyöngyvér Molnár [email protected] Langat Gilbert Cheruiyot [email protected] 1 Institute of Education, MTA-SZTE Digital Learning Technologies Research Group, University of Szeged, Szeged, Hungary 13 390 Journal of Computers in Education (2026) 13:389–410 Keywords Assessment · Bibliometric analysis · Collaborative problem-solving skills · Trends · Review · Technology-based Introduction Complex social challenges facing the globe, such as climate change, demand swift action, which, in turn, has transformed human life. These difficulties, which require community action, highlight the significance of collaborative problem-solving (CPS) skills. Indeed, such skills are crucial in fostering a collaborative and creative atmosphere needed to tackle global health crises and resolve economic issues (Graesser et al., 2018; Stadler et al., 2019; Li et al., 2024). Collaborative problem-solving integrates social and cognitive skills, requiring free exchange of ideas and collective effort to solve issues (Hesse et al., 2015). Social skills facilitate collaboration, while cognitive skills are essential for conventional problem-solving (Andrews-Todd & Kerr, 2019). Meanwhile, integrating these skills might be perceived as reflective of the skills of the CPS. However, owing to the complexity of CPS skills, assessing these skills presents several challenges (Care et al., 2015). Traditional methods for evaluating CPS skills included self-reported surveys, observations, checklists, interviews, peer ratings, and think-aloud exercises (Aguado et al., 2014; Oliveri et al., 2017). Traditional assessment methods inadequately reflect the complex interactions essential for CPS, driven by communication and joint task interaction. The tendency of various observation methods to misinterpret joint task interactions results in unreliable assessments (Andrews-Todd et al., 2022). Moreover, self-reported questionnaires tend to assess meta-knowledge rather than actual skills, and therefore, they are less accurate than achievement tests (Andrews-Todd et al., 2022; Csányi & Molnár, 2023). In this context, technology-based assessments that utilize computer-based environments have emerged as an alternative; these assessments record and report every action and interaction during problem-solving as crucial evidence of individual and collaborative skills (Rosen, 2015). Assessing CPS requires participants to clearly explain their problem-solving methods and logically organize their contributions, which presents various difficulties for assessment (Care et al., 2015). To ensure fairness in the assessments, students with similar skill levels should be evaluated separately, as this allows for comparable performance metrics. Educators must design collaborative groups while considering gender, race, and cognitive abilities. Students should rotate roles within teams, including leadership positions, enabling all students to gain the experience necessary for ensuring effective teamwork (Rosen, 2015). This paper presents a review and bibliometric analysis of technology-based assessments of CPS skills. We compare and analyze various assessment approaches and the current state of technology-based assessments of CPS skills. Further, it identifies the most influential researchers in the field while pointing out the gaps in CPS skill assessment research. 13 Journal of Computers in Education (2026) 13:389–410 391 Definition and collaborative problem-solving assessment Definition of collaborative problem-solving There is no unified definition or model for CPS; however, numerous key perspectives have emerged. The OECD’s Programme for International Student Assessment (PISA, 2015) framework defines CPS as the competency to collaborate with others in sharing beliefs and resources to solve problems (OECD, 2017). According to Kyllonen (2012), CPS involves students collectively addressing a challenge as a team. Hesse et al. (2015) interpret CPS as a systematic collaborative effort to achieve a predefined outcome; meanwhile, O’Neil et al. (2003) describe CPS as an activity involving cooperation. For this study, we define CPS skills as individual active involvement in a group tasked with specific duties and responsibilities and sharing perspectives, ideas, and knowledge while solving a particular problem. Participants responsible for problem identification, information sharing, communication, and conflict management— must devise a solution to the problem presented (Rosen & Rimor, 2012). They must also maintain a positive atmosphere of collaboration. Each member should contribute to the group’s progress, evaluate the efficiency of the means, and make valid conclusions based on the analyzed information. Assessment of collaborative problem-solving skills Collaborative problem-solving entails using skills in a group setting, facilitating information exchange, and using knowledge to address an issue (OECD, 2013). Various methods of measuring these skills have been proposed, such as applying conventional methods using computers and other technological assessment tools. However, traditional assessment methods, questionnaires, and multiple-choice tests have proven ineffective in assessing these skills. Fuad et al. (2019) evaluated primary school teacher trainees’ participation in scientific learning using a self-report questionnaire completed with pencil and paper. Although collaborative approaches were shown to have positive effects (Andrews-Todd et al., 2022), Gonyea, (2005) found that these approaches provide only approximate estimations of participants’ actual abilities. Responses are subjective opinions (Yin & Abdullah, 2013); therefore, the ratings are neither reliable nor objective (Stephen et al., 2017). Furthermore, Pásztor-Kovács et al. (2023) confirmed that self-reported questionnaires for assessing CPS skills are subjective and provide limited insight, as they do not capture real-time interactions or the dynamics of the CPS process. Indeed, these questionnaires focus on outcomes rather than the collaborative process, presenting difficulties in evaluating process-oriented skills (Krkovic et al., 2014). Additionally, there is often a low correlation between participants’ self-reported capability and their actual skills (Hodes & Thomas, 2020). These limitations have led to the adoption of technology-based achievement tests for CPS, incorporated into major international assessments, such as PISA (Rosen, 2015; Nouri et al., 2017; Andrews-Todd et al., 2023). Such technology-based assessments have proven advantages for evaluating cognitive skills compared to conventional approaches. Additionally, technologybased assessments also enable the use of multimedia elements (Parshall et al., 2010), 13 392 Journal of Computers in Education (2026) 13:389–410 avatars, interactive features (Molnár et al., 2017), and collaborative components (Pásztor-Kovács et al., 2023). Consequently, these technologies have become vital tools for effectively assessing CPS skills, producing more profound and engaging assessment experiences. Two main approaches to the test-based assessment of CPS skills can be articulated. That is, assessments focus either on either human-human (H-H) or humanagent (H-A) interactions, with each approach having benefits and drawbacks. Human-human interactions closely replicate real-world collaborative settings, and therefore, they facilitate the natural communication essential for effective problemsolving. For example, the globally recognized large-scale evaluation, PISA, uses the H-A technique to evaluate the student’s CPS skills (Scoular et al., 2017). In the PISA component, CPS skills assessment students collaborate with one or two virtual peers who offer feedback comparable to actual students. This strategy stresses collaboration and active participation by assigning students CPS tasks that involve the same resources and methods. Furthermore, the H-A approach allows for comprehensive scoring; nevertheless, it necessitates assigning each possible participant response to a distinct agent stimulus or issue scenario event. The H-A approach treats data as point estimates for aggregation and analysis rather than stages of developing competence, as in H-H evaluations; however, this approach may restrict the formative interpretation of findings (Rosen, 2015). Finally, assessments of CPS skills by human agents do not accurately reflect real-world CPS skills (Stephen et al., 2017). In contrast, Pásztor-Kovács et al. (2023) and Ras et al. (2014) have highlighted the validity and advantages of the H-H approach, which enables realistic interactions that imitate real-life collaborative situations in personal, educational, and professional settings (Rosen, 2015). Issues such as group size, composition disparities, and differences in individual perspectives may all influence the CPS process and outcome comparability (He, 2023). However, despite these challenges, the H-H approach, compared to the more standardized and controlled H-A approach (He, 2023), improves collaboration by capturing all aspects of human communication and social interaction, increasing student motivation and engagement in CPS tasks (Rosen, 2015). Human-human and human-agent strategies have vastly different pedagogical implications in the assessment of collaborative problem-solving skills. Though both emphasize collaboration in solving situations, the method by which they engage in interaction varies and has potential applications in the learning landscape. Humanhuman interaction engages in direct interpersonal interaction where knowledge is shared, solution-finding occurs, and input from different perspectives is welcome. This approach fosters basic soft skills of collaboration and interaction important in learning environments. Human-agent collaboration, by contrast, refers to the interaction of humans with a technology agent for the sake of problem-solving. This approach amplifies the group dynamics and also organizes the learning experience to adapt to the requirements of an individual student (Rosen, 2015). Integration of tools together with human-human interaction will foster collaboration on digital platforms, encourage group work and real-time peer feedback, and provide reflection mapping that will enhance the capacity of the learners to develop collaborative skills. Simulation tools provide for interaction between human participants and automated agents to facilitate the learning of complex systems (Rakić et al., 2020). Computer agents 13 Journal of Computers in Education (2026) 13:389–410 393 create interactive environments for the assessment of collaborative problem-solving skills as they engage students in collaborative tasks while monitoring responses during the interaction (Rosen et al., 2020). The further use of virtual agents is to counteract some of the biases that arise due to group dynamics and facilitate more advanced standardization in assessment procedures (Rosen & Mosharraf, 2014). Human-agent interactions involve attributes that assess both cognitive skills and collaborative behavior. A collaborative behavior assessment system evaluates the collaborative actions of participants during problem-solving tasks with a computer agent (Krkovic et al., 2016). These tools help merge the personalized support of human-human interaction with the scalability of a human-agent system in a hybrid model, thus complementing the strengths of both approaches. In summary, H-H interactions promote authenticity, while H-A interactions emphasize the consistency and scalability of assessment methods. This study focuses on recent developments, trends, and influential researchers in this area. To bridge this research gap, we reviewed the literature on CPS skill assessment using a bibliographic analysis. Aims and research questions The study aims to justify the present status in the field and examine trends in technology-based assessments of CPS skills via an analysis of the systematic review of the existing literature. The research questions for the study include: RQ1: What has been the ongoing state of studies on assessing collaborative problem-solving skills over a decade? (2014–2024)? RQ1a: What have been the publication patterns of studies assessing collaborative problem-solving skills over a decade (2014–2024)? RQ1b: Who has been the most influential author in collaborative problem-solving studies over the last ten years (2014–2024)? RQ1c: What have been the main features of sampling empirical studies on collaborative problem-solving over the last ten years (2014–2024)? RQ2: What have been the major trends in research based on technology assessments of collaborative problem-solving skills over the past ten years (2014–2024)? RQ2a: What have been the trends in keyword usage regarding the assessment of collaborative problem-solving skills over the last ten years (2014–2024)? RQ2b: What changes have emerged in keyword trends related to collaborative problem-solving over the past 10 years (2014–2024)? 13 394 Journal of Computers in Education (2026) 13:389–410 RQ2c: What have been the latest approaches in assessing collaborative problem-solving skills over the past ten years (2014–2024)? Methodology This review investigates the technology-based assessment of CPS skills using bibliometric analysis and systematic review methods. Bibliometric analysis quantitatively examines research trends and bibliographic data, providing insights into the historical development of research, with a particular focus on research output (Ellegaard & Wallin, 2015). We conducted the bibliometric analysis by selecting relevant articles from the database, collecting and analyzing data, and synthesizing key findings. We also employed statistical and computational techniques to identify the most prominent researchers in the field. Additionally, the authors developed the structure for a systematic review that ensured the established guideline using Preferred Reporting Items for Systematic Review, and Meta-analysis Protocols (PRISMA) were utilized for reporting and controlling methodologies. Web of Science (WoS) databases commonly used to collect data were the data source. The authors were responsible for interpreting the findings, drawing conclusions, and contributing to the discussion. The study’s findings provide new insights into current trends, influential researchers, and the latest approaches in assessing collaborative problem-solving skills that can inform future research and the development of collaborative problem-solving tools in education. We employed the h-index designed by Hirsch (2005) to ascertain the most prominent researchers in CPS skills. This index measures a researcher’s impact based on publication and citation metrics. However, the h-index has limitations, particularly in comparing researchers across different fields and career stages (Hirsch, 2007). To address this, Hirsch introduced the m-index, which evaluates a researcher’s long-term impact by incorporating the h-index and the period since the author’s initial publication release, providing a more detailed understanding of a researcher’s impact over the years. Simultaneously, the g-index evaluates the comprehensive citation achievement of a researcher’s publications, attributing more significance to frequently referenced works (Egghe, 2006). Data collection Bibliometric analysis is an essential tool for analyzing trends in technology-based research on CPS; it involves statistical and quantitative examinations of various attributes of published documents, including authorship, subject matter, cited authors, and publication details. During the data collection process, we utilized specific keywords to identify studies related to CPS, H-H interaction, H-A interaction, and technology-based assessment. Moreover, we accessed online articles using the WoS platform and searched titles, abstracts, and author names for relevant keywords. The search encompassed publications released between 2014 and 2024. 13 Journal of Computers in Education (2026) 13:389–410 395 Next, a dataset was generated utilizing the specified keywords, emphasizing CPS skills and technology. Education, educational research, computer science, and educational science were among the included disciplines. WoS indexes, such as the Social Science Citation Index, the Social Citation Index, and the Emerging Source Citation Index, were included, resulting in 241 articles being identified for analysis. This dataset provided a foundation for analyzing and investigating research patterns in technology-based assessments of CPS skills. The comprehensive review focused on sample size, participant age, and assessment methods. A WoS database search used the same terms from the bibliometric analysis. The preliminary search was carried out without constraints, followed by a modified search limited to articles in English from 2014 to 2024. A total of 20 articles meeting the criteria were identified and analyzed in terms of sample size, participant levels, group composition, assessment approaches, and challenges. Figure 1 presents a flow diagram outlining the document selection process, which followed the PRISMA 2020 guidelines for systematic reviews. Fig. 1 PRISMA flow chart for the bibliometric analysis of studies on the assessment of collaborative problem-solving skills using technology 13 396 Journal of Computers in Education (2026) 13:389–410 Data analysis Bibliometric tools commonly used for analysis include VOSviewer, BibExcel, SciMAT, and Cite Space (Aria & Cuccurullo, 2017). VOSviewer (Version 1.6.1) was employed for this study due to its network analysis capabilities, which can assess such variables’ word occurrences (Eck & Waltman, 2009). VOSviewer is recognized for its ability to visualize large bibliometric maps efficiently, and it facilitates network design based on connections, where item size reflects frequency and color indicates cluster membership. Co-occurrence analysis, a vital method in this context, evaluates the frequency of associated data within specific entities, often focusing on the author keywords of articles. Results This section outlines the results of a bibliometric analysis and literature review answering the main research question about the current trends and status of technology-based assessment of collaborative problem-solving skills. The sub-sections analyze the publishing trends, influential authors, and features of the empirical sample during the last decade. Furthermore, the analysis explains trends in keyword use and the latest approaches to technology-based assessment. Every sub-section discusses certain aspects of the current research, offering insights into the development and present focus of the field of study. What has been the ongoing state of studies on the assessment of collaborative problem-solving skills over a decade (2014–2024)? (RQ1) To answer RQ1, we investigated changes in the number of publications related to CPS for publications using both H-H and H-A techniques by analyzing the article’s explanatory data, including publication year, authors, and sample size. What have been the publication patterns of studies on the assessment of collaborative problem-solving skills over a decade (2014–2024)? (RQ1a) To facilitate the exploration of these research questions, we retrieved the dataset from the WoS in Bibtex format. This dataset contained comprehensive details on the identified journal articles, including authorship, document type, title, publication year, journal source, volume and issue numbers, page range, citation count, and reference lists (see Table 1). CPS publications increased continuously from 2014 until the COVID-19 pandemic, when the publication decreased, registering a significant drop. After countries returned to in-person instruction, the number of publications increased; however, publication numbers remained lower than before the COVID-19 pandemic (see Fig. 2). This trend illustrates the significance of CPS research integrating technology in assessment that ensures that collaborative skills remain essential for future edu- 13 Journal of Computers in Education (2026) 13:389–410 Table 1 Descriptive features of collected data from the articles Description MAIN INFORMATION ABOUT DATA Timespan Sources (Journals, Books, etc.) Documents Annual Growth Rate % Document Average Age Average citations per doc References DOCUMENT CONTENTS Keywords Plus (ID) Author’s Keywords (DE) AUTHORS Authors Authors of single-authored documents AUTHORS COLLABORATION Single-authored docs Co-Authors per Doc International co-authorships % DOCUMENT TYPES Article article; early access article; proceedings paper 397 Results 2014:2024 89 241 1.34 4.39 12.34 0 458 940 775 21 21 3.63 25.73 214 23 4 Fig. 2 Yearly number of published studies regarding assessments of collaborative problem-solving skills through technology cational strategies. Further, these findings suggest exploring alternative methods for assessing and fostering collaborative problem-solving skills. Alongside these publication frequency changes, we observed a shift in the methodology used from human-human and human-agent approaches, as seen in Fig. 3. Until COVID-19 restrictions were imposed, significantly more research was published that employed the H-H methodology. In contrast, research using the H-A methodology began to increase after 2020. This trend will be further advanced in 2022; perhaps due to PISA assessments that utilized the H-A approach. The transition thus emphasizes the significance of technology integration in assessments that lead to uniformity of assessments. An increasing interest in H-A approaches indicates a greater degree of consistency in the learning environment brought about by enhanced technology and sheds light on the changing collaborative aspect of education. The 13 398 Journal of Computers in Education (2026) 13:389–410 Fig. 3 Trends in human-human and human-agent interactions Table 2 Most influential authors between 2014 and 2024 Author h_index g_index m_index TC Hwang, GJ 4 4 0.8 94 Andrews Todd, J 3 4 0.429 205 Chien, Yh 3 3 0.3 61 Graesser, AC 3 3 0.429 168 Greiff, S 3 4 0.429 147 Herro, D 3 3 0.375 66 Andone, D 2 2 1 4 D’mello, Sk 2 2 1 4 Doughan 2 2 0.667 4 Jiang, Y 1 1 0.500 2 NP 4 4 3 3 4 3 2 2 2 3 PY_start 2020 2018 2015 2018 2018 2017 2023 2023 2022 2023 Country/Region Taiwan, Province of China England Taiwan, Province of China USA England USA Romania USA Canada China findings necessitate further exploration regarding the H-A methods in assessing collaborative skills and ramifications for furthering educational programs and policies. Who has been the most influential author in collaborative problem-solving studies over the last ten years (2014–2024)? (RQ1b) We selected the most reputable researchers in CPS skills based on their h-index, total citations, and publication count. Table 2 enumerates the top 10 most influential researchers who have contributed to the technology-based assessment of the CPS skills domain. Hwang, Andrew-Todd, and Greiff, with four publications each, and Chien and Graesser, with three publications each, emerged as the most productive authors in this field. Hwang had an h-index and g-index of 4, and the four articles Hwang has published since 2020 have received 94 total citations. Hwang also has an m-index of 0.8, making him the most influential author in this field. These indexes highlight each researcher’s productivity and impact on collaborative problem-solving research. Moreover, offering additional insights into understanding the contributions of each researcher in a collaborative problem-solving environment that fosters 21st–century skills. 13 Journal of Computers in Education (2026) 13:389–410 399 What have been the main features of sampling empirical studies on collaborative problem-solving over the last ten years (2014–2024)? (RQ1c) In addressing RQ1c, we analyzed studies based on the criteria of sample size, sample proportion, composition of groups, and grade levels (see Table 3). We identified a diverse range of sample sizes ranging from fewer than 100 participants, which corresponds to a proportional sample size of 40% of all the studies analyzed—to larger sample sizes of over 500 participants (Kuo et al., 2020), which accounted for a proportion of 5% of the studies analyzed (see Table 2). The participant’s grade levels ranged from elementary to college level. Junior secondary students were the most researched age group, representing 45% of all investigations (e.g., Nouri et al., 2017; Andrews-Todd et al., 2023). Meanwhile, elementary students represented 5% (e.g., Herro et al., 2021; Tsang et al., 2020), senior secondary students represented 5% (e.g., Siddiq & Scherer, 2017; Hao & Liu, 2015), and college students represented 5% (e.g., Song, 2020; Herro et al., 2021). Furthermore, the number of groups indicates the number of collaborative student groups involved in each assessment. The studies vary from those utilizing two groups (e.g., Nouri et al., 2017; Rosen, 2015) to those with three or more groups used (e.g., Kuo et al., 2020). CPS evaluations can be carried out with mixed groups or with single groups at different educational levels and for different ages; thus, they provide benefits for learners at all stages of learning. The variations in the different group compositions bring into focus the need to recognize how varied group dynamics can affect the interpretation of the CPS results. This indicates that consideration must be made of the group dynamics since the size and composition of the group could affect the CPS assessments. Such insight by researchers provides an understanding of how different groupings may compromise Table 3 Proportion of sampling features in empirical studies of CPS skills Variable Sample size Categories 0-100 101–200 201–300 301–400 400–500 Above 500 Not indicated Numbers 8 5 1 2 1 2 1 Proportion 40% 25% 10% 5% 5% 10% 5% Elementary Junior Secondary Senior Secondary College Not indicated 3 9 3 3 2 5% 45% 5% 5% 10% 2 3–4 Above 5 Multiple groups Not indicated 5 6 1 1 7 25% 30% 5% 5% 35% Level of learning Number of Groups 13 400 Journal of Computers in Education (2026) 13:389–410 the accuracy and effectiveness of CPS measures while ensuring that CPS assessments remain flexible and relevant across different educational contexts and learning stages. What have been the major trends in research based on technology assessments of collaborative problem-solving skills over the past ten years (2014–2024)? (RQ 2) In addressing this research question, the following sub-questions were examined: What have been the trends in keyword usage regarding the assessment of collaborative problem-solving skills over the last ten years (2014–2024)? (RQ2a) Figure 4 illustrates the 30 widely used terms, emphasizing the focal points of interest in CPS research. The selected documents were analyzed using bibliometric software to identify commonly occurring keywords within the WoS database, including author or “keywords plus.” While author keywords were contributed via authors, “keywords plus” offers a more comprehensive representation of the scope (Garfield, 1990). The present study measures keyword pluses and visualizes them using word clouds, whose different font sizes represent term frequency for deeper insights into each paper’s content and focus. The analyses done on research trends concerning CPS assessment have revealed themes such as collaborative learning, collaboration, technology, education, and problem-solving, further elucidating these on a statistical basis concerning the most commonly used keywords and their distribution. The importance of this analysis lies in the central themes emerging through this work, providing an understanding of field priorities for researchers and thus guiding future studies helping researchers to align their research with key focus areas and contribute toward the development of the collaborative problem-solving assessment. Figure 5 illustrates the results based on the co-occurrence analysis of 1,236, with the 76 highest-frequency words displayed in a word cloud. Each keyword is shown as a node, with lines depicting the connections between nodes; the depth of these lines indicates the magnitude of the association between keywords depending on the cooccurrence in publications. Keywords referenced frequently across studies are highlighted in pink to pinpoint key research areas in the technology-based assessment Fig. 4 Word cloud of the 30 most frequently used keywords generated using VOSviewer 13 Journal of Computers in Education (2026) 13:389–410 401 Fig. 5 Co-occurrence network of keywords generated using VOSviewer of CPS skills. This visualization not only gives a picture of the research structure, showing the central theme cluster, current research trends, and unexplored areas, but it also helps the researcher find new topics, gaps in the existing body of literature, and possible CPS assessment research options. What changes have emerged in keyword trends related to collaborative problemsolving over the past 10 years (2014–2024)? (RQ2b) The thematic evolution of CPS research was analyzed via the shift in keyword usage in the past 10 years. Figure 6 illustrates this evolution, with thicker lines representing more significant themes. The most popular keywords between 2014 and 2017 were collaborative learning and computer science. From 2018 to 2020, research introduced more integrated broader concepts such as flipped classrooms and ICT, with an interest in problemsolving and higher education. From 2021 to 2024, the emphasis shifted dramatically to CPS, problem-based learning, and topics such as ontology and computational thinking, reflecting a trend toward applying collaborative approaches to advanced, technology-driven educational frameworks. This trend illustrates that the emphasis on collaboration has shifted to real-world issues and the incorporation of ideas such as computational thinking, implying that collaborative learning is increasingly being used in more demanding and practical ways in education settings. This trend in the evolution of keywords is significant because it shows that cooperative learning is shifting from a theoretical idea to a valuable tool for solving real-life educational problems. Increased interest in problem-solving and computational thinking indi- 13 402 Journal of Computers in Education (2026) 13:389–410 Fig. 6 Thematic evolution of keywords in collaborative problem-solving skill research Table 4 Frequency of the approach to assessment Approach of Number Proportion Article assessment Human-human 11 55% Andrews-Todd et al. (2023); Li et al. (2023); PásztorKovács et al. (2023); Mashuri et al. (2021); Rojas et al. (2021); Song (2020); Yuan et al. (2019); Song (2018); Polyak et al. (2017); Rosen (2015); Hao & Liu (2015) Human-agent 5 25% Li et al. (2023); Kuo et al. (2020); Nouri et al. (2017); Siddiq and Scherer (2017); Ras et al. (2014) Questionnaire4 20% Li et al. (2023); Fuad et al. (2019); Pásztor-Kovács based approach (2018); Aguado et al. (2014) cates how these keywords seem to draw the line toward the preparation of learners with the skills of the 21st century crucial in STEM education. In addition, there is collaborative problem solving (CPS) in an up-to-date, technology-driven learning environment, as suggested by the emergent focus on computational thinking and ontology, bringing about a growing overlap between collaboration, technology, and cognitive development. What have been the latest approaches in assessing collaborative problem-solving skills over the past ten years (2014–2024)? (RQ2c) Using technology, we identified three major methods of assessing CPS skills: H-H interaction, H-A interaction, and questionnaires. Studies with H-H interactions involving direct collaboration between individuals accounted for 55% of the studies (see Table 4). H-A interaction assessments using virtual agents accounted for 25% of the studies, while questionnaire-based studies using self-reporting instruments to measure CPS abilities made up approximately 25% of the published papers. These important findings indicate that human-human collaboration is still prevalent, gaining interest in technology-based assessment approaches. This shift allows the refinement and blending of automated assessment tools by combining the benefits of actual teamwork with the scalability and efficiency of a technology-driven approach. 13 Journal of Computers in Education (2026) 13:389–410 403 Discussion The assessment of CPS skills has garnered heightened interest in recent years, mainly due to the greater use of technology in learning settings. The present investigation sought to systematically assess the literature by evaluating the present status of publications in the field, identifying the most prominent authors, and analyzing the essential characteristics of the samples used in empirical studies. This study analyzed keyword use patterns, documented changes in keyword trends, and investigated current approaches for CPS assessment. This research used the WoS database, a reliable resource, to gather comprehensive and up-to-date data. This will indicate essential insights regarding assessing CPS skills and shifts in assessment methods and highlight gaps in the existing literature. Collaborative problem-solving assessment methods are classified into three main methods: Human-Human, Human-agent, and questionnaire-based assessment methods. Human-human assessment refers to a direct collaborative interaction between two people in either face-to-face or computer-mediated integration. Human-agent assessment is where humans and technology-based or computerized agents work together to emulate real-world collaborative problem-solving occasions. The third refers to the questionnaire-based evaluation method, which as its name suggests, measures collaborative skills using self-reported questionnaires, but with confirmation already added on technology, increases preference for technology and thus the scalable and more objective assessment approaches. Bibliometric analysis of publishing patterns over the past ten years (2014–2024) (RQ1a) indicates a notable rise in interest in assessing CPS skills. The number of publications about CPS constantly rose from 2014 until the COVID-19 pandemic triggered a temporary drop in education achievement in 2020. Temporary school closures restricted in-person teaching, and challenges in conducting in-person assessments disrupted traditional CPS evaluation methods. This led to the adoption of technology-based approaches to assessment, as remote learning environments required new methodologies. This resurgence indicates that CPS is a relevant and active field of investigation within educational research. Finally, since 2020, H-A interactions have been the most explored research topic, denoting an increasing interest in applying technology to simulate human interactions during CPS tasks. This transition underscores the progress in educational technology and the need for innovative methods to evaluate critical 21st-century skills, such as CPS skills. These findings correspond with the extensive implementation of the OECD’s 2015 PISA assessments and the ATC21S framework (Care et al., 2018), underscoring the global focus on CPS in educational assessments. This review further explicated the current state of CPS skills assessment by identifying the most influential author in this field over the past ten years (2014–2024) using the h-index, g-index, and m-index metrics (RQ1b). A comparison of the ten most influential authors in CPS skills revealed that Hwang was the most influential, having an h-index and g-index of 4, with a cumulative citation count of 94 across four publications since 2020. Additionally, the analysis indicated that several studies concentrated on junior secondary school students, with sample sizes lower than 100 participants (RQ1c). Most studies have focused on junior secondary schools (grades 13 404 Journal of Computers in Education (2026) 13:389–410 6–8), with fewer focusing on upper secondary schools (grades 10–12). This results in a knowledge gap regarding CPS skills development within upper secondary education settings. These findings align with the observation presented by Molnar et al. (2013), which highlights the enhancement of cognitive skills throughout middle school. The literature on the development of CPS skills focusing on older learners has not been addressed; this creates a research gap that underlines the need for further investigation into how these skills are enhanced during later stages of education. This gap presents considerable challenges in advancing cognitive and collaborative problem-solving skills beyond early secondary school. The insufficient representation of elementary school students in CPS assessments could limit the understanding of collaborative problem-solving development. Further, the underrepresentation of senior secondary school students in CPS assessments adversely affects the validity of evaluations and the comprehensive development of critical skills. This deficiency may impede students’ capacity to effectively participate in university-level collaborative endeavors, where such skills are essential for academic and professional achievement. Furthermore, inadequate CPS training in undergraduate programs may lead to graduates entering the job without essential collaborative skills. If higher education institutions disregard CPS evaluation, students may forfeit technology developments that might greatly improve their collaboration abilities. Moreover, poor evaluation of CPS at the senior secondary and undergraduate tiers may lead to students obtaining insufficient feedback to enhance vital 21st-century competencies, such as communication, which are needed in academic settings. The insufficient emphasis on CPS may impede the creation of effective assessment tools, resulting in education systems using less relevant ways for assessing collaborative problem-solving skills. The keyword frequency analysis indicated a clear focus on concepts that influence the present framework of CPS assessment (RQ2a). The recurring concepts demonstrate a network of interconnected ideas, emphasizing the significance of collaborative learning and the impact of technology on learning. This distribution provides insight into key research topics and serves as a guiding framework for the discipline. Accordingly, such a trend analysis would allow researchers to identify unexplored alternatives while bridging the literature gaps and exploring emerging research domains that will shape the future assessment approaches of CPS. (RQ2b). Our review indicated that the latest approaches used in CPS skill assessment include the H-H approach, the H-A approach, and the questionnaire-based approach (RQ2c). This study revealed that the H-H approach dominates the field, indicating rising scholarly interest in collaboration. Meanwhile, the H-A interaction indicates an emerging interest in technology. Questionnaire-based evaluations are insufficiently employed, and difficulties associated with these assessments mainly arise from insufficient detailed documentation of collaborative dynamics, which may be mitigated via technology. With further development, educational technology can serve as an enabling method to overcome some of these challenges, thereby allowing for the practical assessment of CPS skills with appropriate design questionnaires. The results of this study have important implications for education, particularly for STEM because collaborative problem-solving skills better prepare students for advancement in a rapidly changing digital workplace. CPS in curriculum may contribute to enhancing 13 Journal of Computers in Education (2026) 13:389–410 405 students’ critical thinking, collaboration, and problem-solving proficiencies, all of which are integral to success in educational and professional spheres. Those abilities equip students to convey ideas and solve problems that matter to them during the learning process. Conclusion In conclusion, the bibliometric analyses used in this study enabled us to provide a comprehensive and systematic summary of the present status and trends in technology-based assessments of the CPS skills field. Additionally, this study shed light on the shift from traditional assessments to technology to foster collaborative environments for assessing CPS skills, a change that previous research failed to highlight. The results demonstrate that CPS skill assessment has emerged as a research area of increasing interest over the past 10 years, with particular interest in H-A interactions. The analysis indicated that Hwang is the outstanding author in the field with the most significant citation index. The study also showed that most research has focused on small sample sizes (0–100 participants) of students in lower secondary education, often with small groups. Furthermore, keyword frequency analysis emphasized group problem-solving and its potential integration with technology and education. This highlights essential connections and emerging trends while demonstrating possible gaps in technology-based skill assessments. Although the assessed studies constituted meaningful contributions to the domain, additional study is required to enhance the assessment of CPS skill development. Therefore, we strongly encourage researchers to consider the following recommendations: (a) expand study samples beyond lower secondary students to include students from different educational levels while also including both cross-sectional and longitudinal research; (b) develop a technology-driven approach to gauge the advancement of CPS abilities in long-term research; (c) evaluate the efficacy of technological tools that emphasize human agents supporting CPS tasks for assessment, such as virtual agents; and (d) focus on designing a more effective questionnaire to assess CPS in recognition of the underutilization of the questionnaire approach. Limitations Although bibliometric analysis helps assess academic contributions, this research acknowledges that different publishing patterns may present difficulties in accurately identifying prominent authors. Authors can use several identities and initials, complicating proper citations and publishing. Moreover, several authors may possess identical names, which complicates citation counts, leading to variability in the assessment of author metrics, including h-index, g-index, and total citations. Such metrics may generate confusion when interpreting an author’s citation effect, as it may be unclear which citations pertain to the intended author. Publications for analysis were sourced exclusively from the WoS; other databases, such as Scopus, Google Scholar, and ERIC, were excluded. Despite its focus on high-impact, peer-reviewed articles, the Web of Science (WoS) has a language bias making it incomplete in explicitly cov- 13 406 Journal of Computers in Education (2026) 13:389–410 ering other research outputs and literatures done in other languages, regional journals, or articles from open-access publishers. This limitation can cause a bias toward perspectives that are dominant linguistically or geographically thereby disregarding valuable perspectives coming from educational contexts that are less represented. WoS’s bias against gray literature parameters like conference proceedings and dissertations, which frequently contain substantial exploratory ideas, can also constitute a principal drawback. A suggestion proposed to salvage these gaps is to include multiple databases, for instance, Scopus, ERIC, Google Scholar, PubMed, and local databases, in the process of data collection. The inclusion of preprints and gray literature would help reduce publication bias and increase coverage with the dataset. This way, future research would develop a more broadly reduced view of collaborative learning practices that is more- inclusive and broadly relevant through the selective inclusion of data sources. Acknowledgements This research was supported by a Hungarian National Research, Development, and Innovation Fund grant (under the OTKA K135727 funding scheme), a Hungarian Academy of Sciences Research Programme for Public Education Development grant (KOZOKT2021-16), and the Humanities and Social Sciences Cluster of the Centre of Excellence for Interdisciplinary Research, Development, and Innovation of the University of Szeged. The authors are members of the Digital Learning Technologies Incubation Research Group. Author contributions Langat Gilbert: Writing an original draft, Writing– review & editing, Data curation Conceptualization. Gyöngyvér Molnár: Writing– reviewing & editing, Methodology, and supervision. Funding Open access funding provided by University of Szeged. Data availability ​h​t​t​p​s​:​​/​/​d​r​i​​v​e​.​g​o​o​​g​l​e​.​​c​o​m​/​d​​r​i​v​e​/​​f​o​l​d​e​r​​s​/​1​u​​n​s​V​v​k​​u​0​D​y​m​​c​h​s​1​t​7​​v​S​A​l​​Z​0​N​q​C​O​D​m​O​F​O​?​u​ s​p​=​d​r​i​v​e​_​l​i​n​k. Declarations Conflicts interests The author discloses no conflicting interests, either direct or indirect, in this work. Over the last three years, the author has not made any financial, professional, or personal decisions that would have affected the findings of this study or its submission for publication. Ethical statement We hereby declare that this manuscript is the result of our independent creation under the reviewers’ comments. Except for the quoted contents, this manuscript does not contain any research achievements that have been published or written by any individuals or groups. We are the only manuscrip authors. The legal responsibility for this statement should be borne by us. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit ​h​t​t​p​:​/​/​c​r​e​a​t​i​v​e​c​o​m​m​o​n​s​.​o​r​g​/​l​i​c​e​n​ s​e​s​/​b​y​/​4​.​0​/​​​​.​​ 13 Journal of Computers in Education (2026) 13:389–410 407 References Aguado, D., Rico, R., Sánchez-Manzanares, M., & Salas, E. (2014). Teamwork competency test (TWCT): A step forward in measuring teamwork competencies. Group Dynamics: Theory Research and Practice, 18(2), 101–121. https://doi.org/10.1037/a0036098 Andrews-Todd, J., & Kerr, D. (2019). Application of ontologies for assessing collaborative problem-solving skills. International Journal of Testing, 19(2), 172–187. ​h​t​t​p​s​:​​/​/​d​o​i​​.​o​r​g​/​1​​0​.​1​0​​8​0​/​1​5​​3​0​5​0​5​​8​.​2​0​1​9​​ .​1​5​7​​3​8​2​3 Andrews-Todd, J., Steinberg, J., Flor, M., & Forsyth, C. M. (2022). Exploring automated classification approaches to advance the assessment of collaborative problem-solving skills. Journal of Intelligence, 10(3), 39. ​h​t​t​p​s​:​​/​/​d​o​i​​.​o​r​g​/​1​​0​.​3​3​​9​0​/​j​i​​n​t​e​l​l​​i​g​e​n​c​e​​1​0​0​3​​0​0​3​9 Andrews-Todd, J., Jiang, Y., Steinberg, J., Pugh, S. L., & D’Mello, S. K. (2023). Investigating collaborative problem-solving skills and outcomes across computer-based tasks. Computers and Education, 207 (September), 104928. ​h​t​t​p​s​:​​/​/​d​o​i​​.​o​r​g​/​1​​0​.​1​0​​1​6​/​j​.​​C​o​m​p​e​​d​u​.​2​0​2​​3​.​1​0​​4​9​2​8 Aria, M., & Cuccurullo, C. (2017). Bibliometric: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007 Care, E., Griffin, P., Scoular, C., Awwal, N., & Zanetti, N. (2015). Collaborative problem-solving tasks. In P. Griffin, & E. Care (Eds.), Assessment and teaching of 21st century skills (pp. 85–104). Springer Netherlands. https://doi.org/10.1007/978-94-017-9395-7_4 Care, E., Griffin, P., & Wilson, M. (Eds.). (2018). Assessment and teaching of 21st-century skills: Research and applications. Springer. https://doi.org/10.1007/978-3-319-65368-6 Chai, H., Hu, T., & Wu, L. (2023). Computer-based assessment of collaborative problem-solving skills: A systematic review of empirical research. Educational Research Review, 100591. ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​.​1​ 0​1​6​/​j​.​e​d​u​r​e​v​.​2​0​2​3​.​1​0​0​5​9​1​​​​ Chang, Y. H., Yan, Y. C., & Lu, Y. T. (2022). Effects of combining different collaborative learning strategies with problem-based learning in a flipped classroom on program Language learning. Sustainability, 14(9), 5282. https://doi.org/10.3390/su14095282 Csányi, R., & Molnár, G. (2023). How do test-takers rate their effort? A comparative analysis of self-report and log file data. Learning and Individual Differences, 106 (July 2022). ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​.​1​0​1​6​/​j​.​l​i​ n​d​i​f​.​2​0​2​3​.​1​0​2​3​4​0​​​​ Dignum, F. (2012). Agents for games and simulations. Autonomous Agents and Multi-Agent Systems, 24(2), 217–220. https://doi.org/10.1007/s10458-011-9169-2 Eck, N. J. V., & Waltman, L. (2009). How do I normalize co-occurrence data? An analysis of some wellknown similarity measures. Journal of the American Society for Information Science and Technology, 60(8), 1635–1651. https://doi.org/10.1002/asi.21075 Egghe, L. (2006). Theory and practice of the g-index. Scientometrics, 69(1), 131–152. ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​.​ 1​0​0​7​/​s​1​1​1​9​2​-​0​0​6​-​0​1​4​4​-​7​​​​ Ellegaard, O., & Wallin, J. A. (2015). The bibliometric analysis of scholarly production: How significant is the impact? Scientometrics, 105(3), 1809–1831. https://doi.org/10.1007/s11192-015-1645-z Fu, H. Z., Ho, Y. S., Sui, Y. M., & Li, Z. S. (2010). A bibliometric analysis of solid waste research during the period 1993–2008. Waste Management, 30(12), 2410–2417. Fuad, A. Z., Alfin, J., Fauzan, Astutik, S., & Prahani, B. K. (2019). Group science learning model to improve primary school teacher candidates’ collaborative problem-solving skills and self-confidence. International Journal of Instruction, 12(3), 119–132. https://doi.org/10.29333/iji.2019.1238a Garfield, E. (1990). Keywords Plus®: ISI’s breakthrough retrieval method. Part 1. Expanding your searching power on Current Contents on Diskette. Current Contents®, 1(32), 5–9. Gonyea, R. M. (2005). Self-reported data in institutional research: Review and recommendations. New Directions for Institutional Research, 127, 73–89. https://doi.org/10.1002/ir.156 Graesser, A. C., Fiore, S. M., Greiff, S., Andrews-Todd, J., Foltz, P. W., & Hesse, F. W. (2018). Advancing the science of collaborative problem solving. Psychological Science in the Public Interest, 19(2), 59–92. https://doi.org/10.1177/1529100618808244 Graesser, A. C., Greiff, S., Stadler, M., & Shubeck, K. T. (2020). Collaboration in the 21st century: The theory, assessment, and teaching of collaborative problem-solving. Computers in Human Behavior, 104(September 2019)), 2019–2021. https://doi.org/10.1016/j.chb.2019.09.010 13 408 Journal of Computers in Education (2026) 13:389–410 Hao, J., Liu, L., von Davier, A., Kyllonen, P., Lindwall, O., Hakkinen, P., et al. (2015). Assessing collaborative problem-solving with simulation-based tasks. Paper presented at the 11th International Conference on Computer Supported Collaborative Learning: Exploring the Material Conditions of Learning, CSCL 2015. HE, Q. (2023). Collaborative problem-solving design in large-scale assessments: Shedding lights in sequential conversation-based measurement. International Journal of Assessment Tools in Education, 10(Special Issue), 194–207. https://doi.org/10.21449/ijate.1407315 Hesse, F., Care, E., Buder, J., Sassenberg, K., & Griffin, P. (2015). A framework for teachable collaborative problem-solving skills. In P. Griffin & E. Care (Eds.), Assessment and Teaching of 21st Century Skills (pp. 37–56). Springer Netherlands. https://doi.org/10.1007/978-94-017-9395-7_2 Herro, D., Quigley, C., Abimbade, O. (2021). Assessing elementary students’ collaborative problem-solving in maker space activities. Information and Learning Sciences, 122, 774–794. ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​.​ 1​1​0​8​/​i​l​s​-​0​8​-​2​0​2​0​-​0​1​7​6​​​​ Hirsch, J. E. (2005). An index to quantify an individual’s scientific research output. Proceedings of the National Academy of Sciences, 102(46), 16569–16572. https://doi.org/10.1073/pnas.0507655102 Hirsch, J. E. (2007). Does the h index have predictive power? Proceedings of the National Academy of Sciences, 104(49), 19193–19198. https://doi.org/10.1073/pnas.0707962104 Hodes, L. N., & Thomas, K. G. F. (2020). Inaccuracy of self-reports and influence of psychological and contextual factors. Computers in Human Behavior. https://doi.org/10.1016/j.chb.2020.106616. 106616. Krkovic, K., Greiff, S., Pásztor-Kovacs, A., & Gyöngyvér, M. (2014). New technologies in psychological assessment: The example of computer-based collaborative problem-solving assessment. International Journal of E-Assessment, 4, 1–13. https://doi.org/10.18293/IJEA.2014.4001 Krkovic, K., Wüstenberg, S., & Greiff, S. (2016). Assessing collaborative behavior in students: An experiment-based assessment approach. European Journal of Psychological Assessment, 32(1), 52–60. https://doi.org/10.1027/1015-5759/a000329 Kuo, B. C., Liao, C. H., Pai, K. C., Shih, S. C., Li, C. H., & Mok, M. C. (2020). Computer-based collaborative problem-solving assessment in Taiwan. Educational Psychology, 40 (9(SI), 1164–1185. ​h​t​t​p​s​:​​/​/​ d​o​i​​.​o​r​g​/​1​​0​.​1​0​​8​0​/​0​1​​4​4​3​4​1​​0​.​2​0​1​8​​.​1​5​4​​9​3​1​7 Kyllonen, P. C. (2012). Measurement of 21st-century skills within the common core state standards. In Invitational research symposium on technology-enhanced assessments (pp. 7–8). Li, J. F., Wang, M. H., & Ho, Y. S. (2011). Trends in research on global climate change: A science citation index Expanded-Based analysis. Global and Planetary Change, 77(1–2), 13–20. Li, S., Pöysä-Tarhonen, J., & Häkkinen, P. (2023). Students’ collaboration dispositions across diverse collaborative problem-solving skills in a computer-based assessment environment. Computers in Human Behavior Reports, 11, 100312. https://doi.org/10.1016/j.chbr.2023.100312 Li, M., Liu, H., Cai, M., & Yuan, J. (2024). Estimation of individuals’ collaborative problem-solving ability in computer-based assessment. Education and Information Technologies, 29(1), 483–515. ​h​t​t​p​s​:​/​ /​d​o​i​.​o​r​g​/​1​0​.​1​0​0​7​/​s​1​0​6​3​9​-​0​2​3​-​1​2​2​7​1​-​w​​​​ Mashuri, N., Hermanto, I. M., Sinaga, P., & Hasanah, L. (2021). Evaluating collaborative problem-solving skills: Students’ social and cognitive skills on the parabolic motion material. Journal of Physics: Conference Series, 1806(1), 012038. ​h​t​t​p​s​:​​/​/​d​o​i​​.​o​r​g​/​1​​0​.​1​0​​8​8​/​1​7​​4​2​-​6​5​​9​6​/​1​8​0​​6​/​1​/​​0​1​2​0​3​8 Molnár, G., Greiff, S., & Csapó, B. (2013). Inductive reasoning, domain-specific and complex problem solving: Relations and development. Thinking Skills and Creativity, 9, 35–45. ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​.​1​0​ 1​6​/​j​.​t​s​c​.​2​0​1​3​.​0​3​.​0​0​2​​​​ Molnár, G., Greiff, S., Fischer, A., & Wüstenberg, S. (2017). An empirical study of computer-based assessment of domain-general complex problem-solving skills. In B. Csapó, & J. Funke (Eds.), The nature of problem solving: Using research to inspire 21st-century learning (pp. 129–150). OECD Publishing. https://doi.org/10.1787/9789264273955-10-en Nouri, J., Åkerfeldt, A., Fors, U., & Selander, S. (2017). Assessing collaborative problem-solving skills in technology-enhanced learning environments—the PISA framework and modes of communication. International Journal of Emerging Technologies in Learning, 12(4), 163–174. ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​.​3​9​ 9​1​/​i​je​ ​t​.​v​1​2​i0​ ​4​.​6​7​3​7​​​​ O’Neil, H. F., Chuang, S. H., Sabrina, & Chung, G. (2003). K. W. K. Issues in the computer-based assessment of collaborative problem-solving. Assessment in Education: Principles, Policy. OECD. (2013). Pisa 2015. Draft a collaborative problem-solving framework. OECD Publishing. OECD. (2017). PISA 2015 assessment and analytical framework: Science, reading, mathematic, financial literacy, and collaborative problem-solving. OECD. https://doi.org/10.1787/9789264281820-en 13 Journal of Computers in Education (2026) 13:389–410 409 Oliveri, M. E., Lawless, R., & Molloy, H. (2017). A literature review on collaborative problem solving for college and workforce readiness. ETS Research Report Series, 2017(1), 1–27. ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​ .​1​0​0​2​/​e​t​s​2​.​1​2​1​3​3​​​​ Parshall, C. G., Harmes, C., Davey, T., & Pashley, P. J. (2010). Innovative items for computerized testing. In van der W. J. Linden, & C. A. W. Glas (Eds.), Computerized adaptive testing: Theory and practice (2nd ed., Vol. 216, p. 215). Kluwer Academic. Pásztor-Kovács, A. (2018). The assessment of collaborative problem solving (Doctoral dissertation). Doctoral School of Education, Faculty of Arts, University of Szeged. Pásztor-Kovács, A., Pásztor, A., & Molnár, G. (2023). Measuring collaborative problem solving: Research agenda and assessment instrument. Interactive Learning Environments, 31(8), 5159–5179. ​h​t​t​p​s​:​​/​/​d​o​ i​​.​o​r​g​/​1​​0​.​1​0​​8​0​/​1​0​​4​9​4​8​2​​0​.​2​0​2​1​​.​1​9​9​​9​2​7​3 Polyak, S. T., von Davier, A. A., Peterschmidt, K. (2017). Computational Psychometrics for the Measurement of Collaborative Problem-Solving Skills. Frontiers in Psychology, 8, 2029. ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​.​ 3​3​8​9​/​f​p​s​y​g​.​2​0​1​7​.​0​2​0​2​9​​​​ Rakić, K., Rosić, M., & Boljat, I. (2020). A survey of agent-based modeling and simulation tools for educational purposes. Tehnicki Vjesnik, 27(3), 1014–1020. https://doi.org/10.17559/TV-20190517110455 Ras, E. (2014). Moving towards the assessment of collaborative problem-solving skills with a tangible user interface. The Turkish Online Journal of Educational Technology, 13(4). Rojas, M., Nussbaum, M., Chiuminatto, P., Guerrero, O., Greiff, S., Krieger, F., & Van Der Westhuizen, L. (2021). Assessing collaborative problem-solving skills among elementary school students. Computers and Education, 175(September 2020)), 104313. ​h​t​t​p​s​:​​​/​​/​d​o​​i​.​o​r​​g​/​​1​0​.​​1​0​​1​​6​​/​j​.​c​o​m​​p​​e​d​u​.​​​2​0​2​​1​.​1​0​4​3​1​3 Rosen, Y. (2015). Computer-based assessment of collaborative problem solving: Exploring the feasibility of the human-to-agent approach. International Journal of Artificial Intelligence in Education, 25(3), 380–406. https://doi.org/10.1007/s40593-015-0042-3 Rosen, Y., & Mosharraf, M. (2014). New methods in online assessment of collaborative problem solving and global competency. Paper presented at the International Association for Educational Assessment Conference, May 25–30, 2014, Singapore. Rosen, Y., & Rimor, R. (2012). Teaching and assessing problem-solving in an online collaborative environment. In R. Hartshorne, T. Heafner, & T. Petty (Eds.), Teacher education programs and online learning tools: Innovations in teacher Preparation (pp. 82–97). Information Science Reference, IGI Global. Rosen, Y., Wolf, I., & Stoeffler, K. (2020). Fostering collaborative problem-solving skills in science: The animalia project. Computers in Human Behavior, 104(June 2018)), 1059–1222. ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​.​1​ 0​1​6​/​j​.​ch​ ​b​.​2​0​1​9​.​0​2​.​0​1​8​​​​ Scoular, C., Care, E., & Hesse, F. W. (2017). Designs for operationalizing collaborative problem solving for automated assessment. Journal of Educational Measurement, 54(1), 12–35. ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​.​1​ 1​1​1​/​j​e​d​m​.​1​2​1​3​0​​​​ Siddiq, F., & Scherer, R. (2017). Revealing the processes of students’ interaction with a novel collaborative problem-solving task: An in-depth analysis of think-aloud protocols. Computers in Human Behavior, 76, 509–525. https://doi.org/10.1016/j.chb.2017.08.007 Song, Y. (2018). Improving primary students’ collaborative problem-solving competency in project-based science learning with productive failure instructional design in a seamless learning environment. Educational Technology Research and Development, 66(4), 979–1008. ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​.​1​0​0​7​/​s​1​1​ 4​2​3​-​0​1​8​-​9​6​0​0​-​3​​​​ Song, M. H., Park, J. A., & Park, J. (2020-08-06).Measuring collaborative problem-solving capability in a creative problem-solving situation, [Paper presentation]. 21st ACM international conference on supporting group work, Sanibel Island, United States of America. ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​.​1​1​4​5​/​3​3​2​3​9​9​4​ .​3​3​6​9​8​8​4​​​​ Stadler, M., Herborn, K., Mustafić, M., & Greiff, S. (2019). Computer-based collaborative problem-solving in PISA 2015 and the role of personality. Journal of Intelligence, 7(3), 15. ​h​t​t​p​s​:​/​/​d​o​i​.​o​r​g​/​1​0​.​3​3​ 9​0​/​j​in​ ​t​e​ll​​i​g​en​ ​c​e​7​0​3​0​0​1​5​​​​ Stephen, M. F., Graesser, A., Greiff, S., Griffin, P., Gong, B., Kyllonen, P., Massey, C., O’Neil, H., Pellegrino, J., Rothman, R., & von Davier, H. S. A. (2017). Collaborative problem solving: Considerations for the National assessment of educational progress. Instructional-Design Theories and Models: A New Paradigm of Instructional Theory, 2, 241–267. Tsang, H. W. C., Liu, Y., & Ying Law, N. W. (2020). An in-depth study of assessment of students’ collaborative problem-solving (CPS) skills in both technological and authentic learning settings. ComputerSupported Collaborative Learning Conference CSCL, 3, 1381–1388. 13 410 Journal of Computers in Education (2026) 13:389–410 Yin, K. Y., & Abdullah, A. G. K. (2013). The collaborative problem-solving questionnaire: Validity and reliability test. International Journal of Academic Research in Business and Social Sciences, 3(1). Yuan, J., Xiao, Y., & Liu, H. (2019). Assessment of collaborative problem solving based on process stream data: A new paradigm for extracting indicators and modeling dyad data. Frontiers in Psychology, 10 (FEB), 1–14. https://doi.org/10.3389/fpsyg.2019.00369 Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Langat Gilbert Cheruiyot is a PhD student at the University of Szeged, Hungary. Research Interests include cognitive skills development, collaborative learning, technology learning, and assessment. Gyöngyvér Molnár is a full professor and the head of the Institute of Education at the University of Szeged, Hungary. Her main areas of interest include technology-based assessment, improving cognitive skills, studying the quality of school learning, and the potential for using ICT in education. She heads eDia, an online diagnostic testing system used in numerous countries. 13