Reasoning and Problem-Solving in Disciplinarily Diverse Teams: Implications for Research and Practice

Structuredness

One of the ways that problems have been described is in relation to the structure of the problem space, and whether that space can be described as well-structured or ill-structured (Alexander & Judy, 1988; Davidsen et al., 2020). Well-structured problems tend to have clear, algorithmic solution paths that result in a widely accepted answer. In contrast, ill-structured problems are approached more heuristically and can have multiple acceptable solutions (Alexander, 2007; Frederiksen, 1986; Simon, 1973). Research on solving ill-structured problems has become increasingly important, especially in teamwork studies within engineering and design fields (Dringenberg & Purzer, 2018; Singh & Chakrabarti, 2024). In contrast, how teams solve well-structured problems has received less attention. This systematic review aims to enhance understanding of how teams solve problems across the spectrum of structuredness given that these types of problems arise in all domains. Additionally, we will explore how these problem structures relate to team reasoning and thinking processes.

Knowledge Demands

The ability to reason and think invariably requires knowledge related to the problem at hand. When it comes to interdisciplinary teamwork, individuals have specialized knowledge of their specific discipline as well as understanding pertinent to related fields. They should be able to integrate and transfer their knowledge across disciplinary domains in order to effectively collaborate with team members representing other areas of expertise (Klein, 2011; Spiro & Jehng, 1990). In empirical studies examining problem characteristics, researchers have often examined two types of problems: those that require minimal knowledge (knowledge-lean; Holyoak, 1991) and those that demand extensive expertise (knowledge-rich; Anderson, 1987). Findings in educational psychology indicate that reasoning processes vary significantly based on the knowledge requirements of the problem at hand (Kulikowich & DeFranco, 2003). One goal of this review is to explore how these knowledge demands affect the reasoning and thinking processes of teams with diverse disciplinary backgrounds.

Task Directives

Assessing task directives is essential for understanding team decision-making processes and their navigation through problem spaces. Several researchers have conceptualized problem spaces as consisting of several stages (Gick, 1986; Goldman, 1983; Polya, 1945) that we generalize here as: first, understanding the problem; second, devising a plan to test the initial understanding; third, carrying out the test; and fourth, identifying or producing a solution or decision. Researchers in educational psychology have explored how specific directives, such as prompting information-gathering for initial sensemaking or creative directives for solution-building, can affect individuals’ reasoning and thinking processes involved within specific stages of problem-solving (Xun & Land, 2004; Puntambekar, 2022). Notably, Lajoie (2003) has found that analyzing tasks can reveal important differences in reasoning and strategic planning among individuals with varying levels of knowledge and experiences throughout the problem-solving process. This review explores the influence of task directives (information-gathering, procedural, creative, prioritization, and decision) on interdisciplinary teams, focusing on how those directives vary based on the knowledge and experience of team members.

Team Composition

Diversity in age (Luksyte et al., 2022), race (Hattery et al., 2022), and gender (Borgonovi et al., 2023) has been shown to significantly impact collaborative team problem-solving. For instance, Borgonovi et al. (2023) conducted a cross-country study on gender performance in problem-solving. They found that in countries where social and political conditions support women, women often outperform their male counterparts. In contrast, in countries where societal factors limit educational opportunities for women, the performance gap is wider. Moreover, both the area of expertise (Raine et al., 2014) and individual’s level of expertise (Alexander et al., 2009) should be key considerations when assessing team dynamics. Traditionally, research on expertise has compared novices directly against experts within a dichotomous framework (Chi et al., 1981; Newell & Simon, 1972). However, in this review, we adopt a broader perspective on expertise development that accounts for various competency levels over time (Alexander, 2003). We explore how team characteristics differ across disciplines and analyze whether these differences have a bearing on team reasoning and thinking.

Direct Observation

Laboratory-like, or in vitro, methods have been the leading method of choice in much of the research related to reasoning, decision-making, and problem-solving (Dumas, 2017). The popularity of these methods lies in the controlled environment that researchers can create, allowing for the study of outcomes that could arguably go unnoticed in more naturalistic settings (Johnson-Laird & Wason, 1977). However, limitations of studying interdisciplinary teams in these settings include reliance on pre- or post-data rather than real-time interactions (Schunn, 2017).

In contrast, there are researchers who employ in vivo or naturalistic methods (Dunbar, 1998; Dunbar, 2001; Frey, 1994; Salas et al., 2008) that allow for direct observation of reasoning and thinking within an ecologically valid environment. While these types of studies are rarer in an educational context (Dumas, 2017), they are more common in professional studies and applications of cognitive science and psychology (Chan et al., 2012; Kuo et al., 2019; Paletz et al., 2011; Price et al., 2022; Yong et al., 2014). Although there are noteworthy movements in this area, such as naturalistic decision-making (Brown et al., 2023; Klein, 2011) and team mental models (DeChurch & Mesmer-Magnus, 2010), many publications lack an explanation of how discrete forms of reasoning were involved (W. Zhang & Pastel, 2015; D. Zhang & Shen, 2015). Therefore, for this review, we intentionally searched for studies that examined explicit forms of reasoning and thinking occurring in teams to understand the underlying processes that influence real-time collaborative problem-solving efforts.

Setting

To understand the nature of interdisciplinary reasoning and thinking, it is imperative to examine the context in which teams are situated. Evidence suggests that the context of a study significantly influences the reasoning and cognitive processes exhibited by individuals during problem-solving (Jurdak, 2006; Rhodes et al., 2024). Moreover, individuals operating in professional workplaces are more likely to grapple with ill-structured problems, whereas students collaborating in educational environments typically encounter more well-structured problems (Danaher & Schoepp, 2020). More recently, considering current societal demands presented in our introduction, an increasing number of researchers are investigating how students navigate ill-structured problems within educational contexts (D. Jonassen et al., 2006; McNeill et al., 2016). Consequently, this review seeks to delineate the distinctions in reasoning and problem-solving approaches across these varied settings.

Data Sources

Given that the setting and research designs described are broad in scope, researchers’ choices as to the sources of data to be investigated become highly critical in the interpretation of findings (Payne & Westerman, 2003). For instance, within the naturalistic studies of teamwork, there are clearly two camps. One camp views these naturalistic designs as an opportunity to collect rich qualitative data, often emphasizing ethnographic and phenomenological approaches to understanding team processes (Steffensen et al., 2016). The other camp views these as an opportunity to quantify team processes using systematic coding schemes that are later incorporated in quantitative models for predictive or correlational analyses (Chiu & Khoo, 2005). Within this systematic review, we want to examine sources of data used in both qualitative and quantitative methods to understand how these data contribute to our understanding of thinking and reasoning in disciplinarily diverse teamwork.

Inclusion Criteria

In addition to the delimitators set in the search parameters, documents included in this systematic review needed to meet the following criteria: First, the participants in these studies:

Although we recognize the importance of understanding interdisciplinary teamwork in adolescents, we only focused on late adolescents and adults to potentially draw insight from the outcomes of more mature working groups. Second, to be included in this review, the documents had to:

Exclusion Criteria

The following criteria excluded documents from the final analysis: (a) the participants were adolescents in K–12 school settings; (b) problem characteristics or reasoning data pertained to individuals working alone; (c) participant data came only from individual interviews conducted; (d) reasoning or thinking was not analyzed in the study; (e) there were no dedicated data on the characteristics of the problem the groups were tasked with; and (f) documents that were systematic reviews, meta-analyses, proposals in curriculum instruction and design, or program evaluations. For example, in Andrews et al. (2020), the researchers followed a team of professional engineers and medical professionals developing a device for patient benefit in pain management. However, no data on the reasoning or thinking processes underlying this collaboration were reported. Therefore, this work was excluded.

Abstract and Title Review

Each of the 1,117 documents identified in the initial search was exported from each database and imported into an online software program called Rayyan (rayyan.ai). The information imported into Rayyan included the title, year of publication, author(s), keywords, and abstract. The first author proceeded with an abstract and title review to ascertain whether they addressed each of the inclusion criteria. If not, or if the abstract mentioned one of the exclusion criteria, it was excluded from further analysis. One of the most frequent reasons for exclusion (38.67%; n = 432) was that the participants in these studies were not interdisciplinary. Rather, while documents referenced the terms “interdisciplinary,” “multidisciplinary,” or “cross-disciplinary,” these terms were only descriptors of researchers’ orientation or approach to their study and not characteristics of the teams involved. Further, despite having excluded adolescent populations within our search parameter, we still identified 43 (3.84%) such documents during title and abstract review. Other reasons for excluding documents during the abstract and title review included publications did not observe group or teamwork (2.33%; n = 26) and publications whose only source of data were interviews or surveys (8.50%; n = 95) or that did not examine a reasoning or thinking process (.90%; n = 10) or were characterized as literature reviews, program evaluations, or proposals in curriculum instruction and design (35.81%; n = 400). In total, we excluded 1,006 documents at this step in the culling process.

Full Text Analysis

The remaining 111 documents were analyzed in their entirety to ensure that the inclusion criteria were met and that there were no exclusionary conditions present. For instance, an article by Kodkanon et al. (2018) examined decision-making within a team of interdisciplinary educators with a focus on teachers’ experiences and how they communicated and supported one another. However, because there was no explicit data on the reasoning processes involved in participants’ decision-making, it was excluded.

In a few cases, we found it necessary to detangle the authors’ writing pertaining to how the sampled participants engaged in interdisciplinary reasoning and thinking. For instance, Ertl et al.’s (2021) title, “Encouraging communication and cooperation in e-learning: solving and creating new interdisciplinary case histories,” and purpose in exploring clinical students’ engagement in case-based learning activities online to help them develop the skills of synthesizing information from multiple disciplines initially led us to consider this paper. What the full-text analysis revealed, however, was that these students’ interactions with the scenario-based cases were performed entirely independently. The entire procedure for the full-text analysis resulted in the exclusion of 88 documents, with 40 documents retained in the final document pool.

Forward and Backward Searching

In forward searching, we searched to find what other documents cited the publications identified in our initial search as a potential source of other relevant studies. Through this process, we found that the databases we selected did not produce a comprehensive record of where the documents identified in the initial search had been cited. For that reason, we decided to incorporate Google Scholar to aid in locating the source of document citations. In backward searching, we read the references of identified documents to determine if other publications might be considered. This forward and backward searching procedure resulted in an additional six documents being added to the initial pool.

Researcher Checking

In the next step in the search process, we looked up the published works of the first authors of documents identified through the selected databases. First, we reviewed any promising titles against the specified inclusion criteria. Second, for those authors whose publications included several relevant documents, we expanded our search by locating their curriculum vitae and checking whether any additional works warranted examination. This multistep procedure added three documents to the initial pool.

Journal Hand-Searching

For journal hand-searching, we used Excel to identify all duplicates within the Journal Name column to determine whether any journal names appeared multiple times. International Journal of Technology and Design Education and Education Sciences were two journals that appeared on multiple occasions in our pool. Therefore, we hand-searched the volumes and issues of these two journals for the period 2012 through 2024 to locate any potential publications missed by the other search procedures. This journal hand-searching process resulted in an additional eight articles being added to the document pool.

Codes for Team Characteristics

The following designations for participants’ level of expertise were used: undergraduates (UG), graduate students (GS), undergraduate and graduate students (U|GS), graduate students and professionals (GS|P), and professionals (P). Gender was coded as male (M), female (F), or nonbinary (NB). Participants’ race and ethnicity were based on self-identified categories of White or Caucasian (C), Black or African American (B), Hispanic or Latino (H), and Asian or Asian American (A). Some researchers chose to report nationalities instead of race. In that case, we coded them as citizens from the United States (US) or as representing multiple nationalities (MN). Many documents had several unique disciplines working together in teams, and so we coded the number and type of disciplines authors described. Given the vast array of disciplines compiled, we chose to collapse those into broader characterizations (formal, natural, social, and applied sciences, and humanities). Citations for included works related to these codes can be found in Table S3 of our supplemental material (online).

Codes for Reasoning and Thinking Processes

Reasoning (R) and thinking (T) were the cognitive processes that researchers designated in their studies. We did not impose those categories on the processes mentioned in the studies we analyzed but relied on the researchers’ determinations. Further, in parentheses, we denote the specific form of reasoning and thinking that the researchers chose to examine. For example, Hanum et al. (2023) were interested in the clinical reasoning abilities of an interdisciplinary team of students, and so we denoted this as R (clinical) to represent this focus on clinical reasoning. Only a few researchers did not provide a specific form of reasoning or thinking in their papers. Noroozi et al. (2013) discussed the reasoning underlying transactivity in collaborative learning and problem-solving but did not identify any particular form of reasoning per se, so we denoted this paper with an R with no parentheses.

Codes for Problem Characteristics

We organized codes related to the characteristics of problems into four sections centered on structuredness, knowledge demands, task directives, and team composition. The citations for these can be found in Table S4 (online only).

Codes for Methodological Approaches

Bias Assessment

To mitigate bias in the document coding and analysis process, a research assistant, uninformed about the purpose of the study, was trained using the document coding manual. Once the research assistant demonstrated competency with training materials using 5% of the included documents, they coded 45% of the remaining documents independently. A high degree of agreement (85.7–100%) was achieved (see Table S6 in supplemental materials, online.)

Team Characteristics

Descriptive findings for team composition are organized into three sections: team demographics, levels of expertise, and disciplines. Figure 2 provides visual representations of the frequencies found for team and problem characteristics.

OPEN IN VIEWER

Figure 2.

Descriptive findings for team member and problem characteristics.

Problem Characteristics

Methodological Approaches

Reasoning and Thinking Processes

Findings regarding the frequency of reasoning and thinking processes studied in disciplinarily diverse teams show that thinking processes were considered more often (n = 26; 65%) than reasoning processes (n = 14; 35%). The most common thinking process that appeared was critical thinking (n = 9; 22.5%). Design thinking was often under investigation in studies that were focused on understanding how concepts like insight, creativity, and novel solution production manifest within design-based problem-solving (n = 7; 17.5%).

The greater part of the studies incorporated in this review focused on different thinking processes (n = 26). The most common thinking processes examined were critical thinking (n = 9), followed by design thinking (n = 7), creative thinking (n = 6), and convergent and divergent thinking (n = 2). Critical thinking was defined by several authors as a process that involves both convergent and divergent thinking, questioning and analyzing information, and making informed decisions regarding the information (Kiernan et al., 2020b; McMahon & Bhamra, 2017). Throughout our review of these papers, we also noted that while strategic thinking was only explicitly mentioned by a few papers, the underlying concept of strategic thinking was evident in many studies. For instance, the goal of several studies was to identify how interdisciplinary teams think through problems that necessitated certain strategies and steps that these experts were previously trained for (n = 3; Hjörne & Saljo, 2014; C.-Y. Lee et al., 2025; Soukup Ascencao, 2017). A case of this is found in Hjörne and Saljo’s (2014) study, which examined the reasoning processes of a multiprofessional team tasked with evaluating and addressing student cases within a school system. Given prior training and competencies of these professionals, one aspect of Hjörne and Saljo’s (2014) study was to first evaluate their process for handling such cases. Thus, when their observations deviated from an expected or anticipated action, the researchers then delved into the nature of the teams’ reasoning within those particular instances of deviation.

The most common forms of reasoning studied in this body of literature were analogical reasoning (n = 4) and clinical reasoning (n = 4), followed by team reasoning (n = 1) and then general reasoning (n = 3). The researchers who utilized analogical reasoning each defined it as the process where a source and a target are linked to one another by a systematic mapping of attributes and relations, allowing for the transfer of knowledge to the target (Chan & Schunn, 2015; Christensen & Ball, 2016; Goel et al., 2014; Paletz et al., 2013). Clinical reasoning in these studies focused on cognitive processes that involve particular teams composed of nurses, doctors, and physicians, among others who are engaged in collecting, processing, and disseminating patient case information aiding in decision-making (Hanum et al., 2023; C.-Y. Lee et al., 2025; Seif et al., 2014; Ward et al., 2016).

Our findings regarding each of the thinking and reasoning processes identified are summarized in Table 1. This table lists each key thinking and reasoning process identified in the review and summarizes the most commonly coded team, problem, and method characteristics used when investigating that particular process. Table 1 sets the stage for the third research question on the ways in which these thinking and reasoning processes interact or coincide with these characteristics.

OPEN IN VIEWER

Table 1

Problem Characteristics by Forms of Thinking and Reasoning

			Mean	Most Frequently Occurring Code
	n a	Levels of Expertise Considered b	Subdomains Represented	Discipline c	Knowledge Demand d	Structure e	Task Directive f	Context g
Forms of Thinking
Critical	9	UG–GS–P	4.44	AS	KR	IS	CR	E
Design	7	UG–GS–P	4.83	AS	KR	IS	CR	E
Creative	6	UG–GS–P	4.67	AS	KR	IS	CR	E
Convergent–divergent	2	UG–GS	11.00	AS, FS, NS	KL, KR	WS, IS	CR	E
Logical and empathetic	1	UG–GS	3.00	AS	KL	IS	PR	E
Reflective and strategic	1	P	3.00	SS	KL	WS	PR	W
General	3	GS–P	8.00	AS	KR	IS	IG	W
Forms of Reasoning
Analogical	4	UG–GS–P	5.50	AS	KR	IS	CR	W
Clinical	4	UG–GS–P	5	AS	KR	WS	DM	E
Team	1	UG	6	AS	KR	WS	DM	E
General	4	UG–GS–P	2.5	AS	KL	IS	PR	E

Note. The data used to generate these descriptives comes from Table S1 in online supplemental material.

a

Number of documents examining one particular form of thinking or reasoning.

b

Undergraduate (UG), graduate (GS), and professional (P).

c

Applied science (AS), formal science (FS), natural science (NS), social science (SS), and the humanities (HU).

d

Knowledge-lean (KL) and knowledge (KR).

e

Well-structured (WS) and ill-structured (IS).

f

Information gathering (IG), decision-making (DM), creative (CR), and procedural (PR).

g

Educational setting (E) and workplace setting (W).

Disciplines by Setting by Reasoning Processes

We first discuss the disciplines, the contexts in which they were studied, and the insights we gained about the reasoning processes that occurred in these situations. Specifically, we found that most studies with a focus on reasoning processes examined teams from the applied sciences observed in educational settings (n = 9; Table S2, online). Moreover, the educational setting was a clear indicator of the focus and purpose of the research. For instance, studies examining clinical reasoning in educational settings with individuals in the applied sciences (n = 3) were focused on interprofessional teamwork (Hanum et al., 2023; Seif et al., 2014; Ward et al., 2016) and how students’ exposure to those experiences and complex problems improved their ability to engage in clinical reasoning. While the settings in these studies were educational, the researchers’ intent was to simulate real-world situations. These studies illustrate how educational contexts functioned as structured environments for rehearsing discipline-specific reasoning, where students were encouraged to articulate assumptions, justify interpretations, and coordinate perspectives as they worked through the case. In this way, the setting and discipline jointly shape the form and enactment of reasoning, emphasizing explanation, evidence evaluation, and collaborative sensemaking.

In the workplace setting, we observed a narrower focus on specific forms of reasoning and changes in the number of disciplines involved. For instance, this was evident in studies examining analogical reasoning as key to solving interdisciplinary problems (n = 3; Chan & Schunn, 2015; Christensen & Ball, 2016; Paletz et al., 2013). In these workplace contexts, analogical reasoning functioned as a way for team members to map concepts across disciplinary boundaries, making their interpretations visible and mutually interpretable. Interdisciplinary teams require members to synthesize knowledge from different fields and apply it to specific problems. Therefore, having the ability to recognize patterns of similarity across these various disciplines becomes increasingly critical.

Problem Structuredness by Task Directives by Thinking Processes

Further, we identified interactions among problem structuredness, task directives, and the thinking processes utilized in these problems. The most common combination of problem structure and task directives was ill-structured problems with directives that prompted the team to solve the problem in creative ways or to devise novel solutions (n = 12). This combination was found in studies interested in design thinking (n = 5). Given the flexibility and the degree of autonomy that teams had in these spaces, design thinking was presented to these teams as a method they could use to navigate ill-structured spaces requiring creative solutions (Allinson & Mahon, 2022; Avery et al., 2024; Baaki et al., 2017; Brockman et al., 2015; H. Lee, 2019). Design thinking was described as though, if an individual were engaging in it, they would follow through with an investigation phase, find design opportunities when interpreting data, generate ideas, experiment with prototypes, and engage in iterative cycles through these processes (H. Lee, 2019). Thus, the structuredness of the thinking process was not synonymous with the structuredness of the task.

Several teams operating in an ill-structured problem space were tasked with gathering information that could potentially provide insight into a problem (n = 4). The purpose of bringing together disciplinarily diverse teams for these design tasks was not, per se, to solve problems. Rather, it was more to discuss ways in which a problem could be solved (Brown et al., 2023; Collins, 2021). Similarly, other researchers were interested in how teams engaged in the first stage of problem-solving, understanding the problem. In this early stage of problem-solving, team members worked to develop a shared framing of the problem, where the reasoning centered on jointly interpreting incomplete information, questioning disciplinary assumptions, and negotiating how the problem itself should be defined (Biello et al., 2022; McMahon & Bhamra, 2017).

In contrast to ill-structured problem spaces, well-structured problems did not appear within this review as frequently when researchers were interested in the thinking processes of teams (n = 7). We found that teams working on well-structured problems were presented with tasks that were very procedural in nature (n = 3). These more structured problems were the focus for researchers interested in creating experiences for teams where they learn how to work together and develop critical thinking skills in the process (Avery et al., 2024; Spelt et al., 2017; Wu & Huang, 2017). The findings of this research were often meant to inform pedagogical frameworks that promote interdisciplinary learning and professional development (Abueg, 2020; Spelt et al., 2017).

Expertise and Subdomains by Knowledge Demands by Thinking and Reasoning Processes

We further found that various combinations of team members and problem characteristics occurred for studies addressing thinking and reasoning processes. We noticed that the pool of studies addressing thinking and reasoning processes involved participants with a range of expertise. For instance, as seen in online Table S5, we have empirical investigations on the nature of critical, design, and creative thinking, as well as analogical and clinical reasoning, that assessed undergraduates, graduates, and professionals (citations for these can be found in Table S3 online).

When assessing the knowledge demands of these problems, we observed that knowledge-rich problems (n = 28) were more frequent than knowledge-lean problems (n = 12). When tasks demanded a base of prior knowledge and were specifically situated within a particular subject, roughly a third of these studies targeted professionals (n = 9). Other researchers considered teams with a mixture of professionals and graduate students (n = 5). Further, teams with higher levels of expertise also represented multiple specializations with a range of 2 to 10 subdomains (M = 5.14). For example, Soukup Ascencao (2017) followed a group of medical professionals from five different domain specializations who were analyzing patient cases to deduce patterns or observations of their critical thinking processes within these contexts.

Knowledge-lean problems often focused on undergraduates (n = 7) and graduate students (n = 2) or a mixture of these groups (n = 2). The focus was often on the way in which disciplinarily diverse teams engaged in solving problems that were more general in nature and did not require highly specialized knowledge (Noroozi, 2013). The use of these knowledge-lean problems may have allowed researchers to focus on the manipulation of certain team characteristics, such as personality (Barré et al., 2017; Wu & Huang, 2017), and the influence of instructional approaches (Noroozi et al., 2013; Tan et al., 2021), or to test particular theories (Spelt et al., 2017; Zhou et al., 2022) or frameworks (Avery et al., 2024; Zyfers, 2019). For example, Tan et al. (2021) revealed how providing preparation time can shape how teams enter into the collaborative space, influencing how they articulated prior knowledge, proposed initial interpretations, and coordinated their reasoning once the task began.

Methods by Data Sources by Findings

The prior findings regarding team members and problem characteristics highlight the importance of understanding the design and methodological approaches undertaken by researchers. As discussed, qualitative methods were most commonly employed, with the focus primarily on observations of team processes unfolding in natural contexts (Brockman et al., 2015; Chen & Qian, 2014).

The data collection for these qualitative studies centered around evidence from interviews, transcripts, field notes, and artifacts, with codes generated from thematic analyses (n = 8). Triangulation of evidence is known to be an essential component of qualitative research that bolsters the integrity and reliability of findings (Denzin, 2009; Merriam & Tisdell, 2016). Half of the qualitative studies in this review incorporated more than three data sources (n = 6), while the others examined only one or two pieces of evidence (n = 7). Further, while some researchers used specific qualitative methods such as ethnography (Chen & Qian, 2014), case study (Brockman et al., 2015; Kiernan et al., 2020b; Kiernan et al., 2022; H. Lee, 2019; McMahon & Bhamra, 2017; Miller et al., 2019), or phenomenology (Hanum et al., 2023), few identified no specific approach (Allinson & Mahon, 2022; Goel et al., 2014; Spelt et al., 2017; Zyfers, 2019).

For quantitative studies, the analysis of codes generated from coding schemes was common (n = 5), as was the deployment of both researcher-developed (n = 8) and standardized measures (n = 3). Studies that employed coding schemes were often focused on capturing naturalistic team processes similar to the aims of qualitative researchers. However, the analysis focused primarily on numerically identifying connections and associations between reasoning processes and problem characteristics (Abueg, 2020; Shen et al., 2015) guided by a priori hypotheses to guide analysis (Christensen & Ball, 2016; Paletz et al., 2013). Researcher-developed measures often included questionnaires (Schöfer et al., 2018; Sorici et al., 2023) or surveys (Yong et al., 2014) that were systematically administered during the collaborative teamwork to capture details of members’ perceptions of the ongoing project and the collaborative experience. In contrast, standard measures were typically used to capture specific data on teams’ reasoning processes (Seif et al., 2014; Ward et al., 2016) or creative thinking (Wu & Huang, 2017).

Whether the primary methodology was qualitative or quantitative in nature, there was evidence that interdisciplinary teams were generally capable of engaging in problem-solving in both educational and workplace contexts. In fact, every study in this review described the usefulness of such teams, especially when problems were knowledge-rich, ill-structured, and creative.

The Meaning of Interdisciplinarity Remains Vaguely and Variably Defined

Even though our primary aim in this review was to delve into the nature of reasoning and interdisciplinary team problem-solving, we first had to deal with the very notion of interdisciplinary represented in this body of work. We did not expect there to be a singular or consistent definition of this or related constructs. However, we had anticipated some common characteristics as to what constitutes interdisciplinarity. Yet, beyond the fact that more than one area of study, profession, or domain was presumably involved, there was little else that established that groups or teams in many studies were actually interdisciplinary in nature. This was made evident in the abstract and title review, where we removed over 400 documents that merely mentioned “interdisciplinary” or “multidisciplinary” without further conceptual explication, identification of the domains or disciplines involved, or other delineating characteristics. Given the high level of diversification we encountered, there was a clear need for more detailed descriptions of the members constituting these interdisciplinary teams. In effect, there needed to be researcher-provided explanations for why the collections of individuals they studied could justifiably be labeled interdisciplinary.

Certain Problem Characteristics May Necessitate Disciplinarily Diverse Teamwork

Researchers most often studied disciplinarily diverse teamwork in relation to ill-structured, knowledge-rich, and creative problems, suggesting that such contexts are viewed as especially relevant for interdisciplinary collaboration. Even in samples that observed more novice teams, researchers were often concerned with how these teams were functioning so that they could better inform educational practices to ensure the next generation of students entering the workforce can solve such problems. Additionally, it was very clear from the descriptive findings that professions within the applied sciences dominated the focus of participants sampled within this pool of studies. This makes sense, given that fields within the applied sciences are highly interdisciplinary by nature, drawing on knowledge and skills in formal, natural, and social sciences to achieve workable and relevant solutions to existing problems.

Scope of Understanding Thinking and Reasoning Processes Varied

It was further evident from this review that the scope of investigation into thinking and reasoning processes varied. Some studies focused on thinking and reasoning processes that subsumed an entire system or sequence of sub-processes, whereas other studies investigated more granular forms of thinking and reasoning (Table S6, online). For instance, the studies into design thinking often evaluated multicomponent thinking and reasoning processes within the entire scope or cycle of design thinking (Baaki et al., 2017; H. Lee, 2019; Sorici et al., 2023). In contrast, we had several studies that investigated specific forms of thinking and reasoning, such as the studies into analogical reasoning (Chan & Schunn, 2015; Christensen & Ball, 2016; Goel et al., 2014; Paletz et al., 2013). We found the definitions of these processes to be clear and targeted within the studies in which they were operationalized, which allowed the outcomes to be interpreted more precisely, highlighting potential relationships between reasoning processes and collaborative outcomes reported by the sampled authors.

Education Research

Instructional Practice

What seemed evident from the studies we analyzed is that carrying out interdisciplinary problem-solving studies effectively is challenging even for researchers with a background in this area. The question arises, therefore, what would educational practitioners need to know or be able to do to incorporate interdisciplinary teams in their classrooms in a manner to facilitate reasoning and thinking within interdisciplinary team problem-solving?

One suggestion would be to draw on existing instructional programs that have interdisciplinary reasoning and thinking as core components, such as problem-based or project-based learning (Bates et al., 2022; Razak et al., 2022). By relying on these established curricular programs, educators would have access to guidelines and resources that could support their efforts. Even then, however, teachers need to be sensitive to the backgrounds and experiences of their students. As the work of those carrying out training in clinical reasoning with health professionals illustrates, it is critical to try to match the complexity of the problem characteristics within the task to the knowledge and skills of students (Avery et al., 2024; Spelt et al., 2017; Wu & Huang, 2017). Therefore, in order to provide appropriate challenges for students in the initial stages of learning to collaborate in these contexts, educators should be mindful of the types of tasks they ask teams to perform, particularly if their goal is to develop the reasoning and thinking abilities of these individuals.

For example, starting with less complex and more structured problems and modeling ways of thinking and sharing ideas would be helpful when working with younger or less-experienced individuals (D. H. Jonassen, 1997; Reed, 2016). With younger and less-experienced learners, it might also be advisable to discuss the value of interdisciplinarity and provide procedural guidelines on how teams should function (Murphy et al., 2018). What are the expectations for each team member, for instance, and how does the team ensure that all members are positively engaged? Further, educators could introduce specific reasoning and thinking strategies that students can use when tackling different types of problems (Biello et al., 2022; Chen & Qian, 2014; Goel et al., 2014; Noroozi et al., 2013). Learning to synthesize information from differing sources does not come naturally for many students, as the literature on multiple document use has repeatedly demonstrated (List & Sun, 2023). Consequently, it would seem helpful for educators to take time to explain how they synthesize ideas that are expressed in writing or orally (Nelson & King, 2023). In that way, students may benefit more from exploring problems from different disciplinary angles and be better equipped to propose alternative solutions to the problem at hand.

Finally, when incorporating interdisciplinary problem-solving into instructional contexts, educators need to consider the composition of the teams. That consideration should not only weigh the background knowledge and skills members would bring to the task, but also social, cultural, gender, and ethnic factors (Borgonovi et al., 2023; Hardt et al., 2025). Care should be taken so that certain students do not dominate the process and that other students do not feel isolated and even dismissed by team members. These observations parallel foundational principles identified in the cooperative learning literature, particularly positive interdependence and individual accountability (Johnson & Johnson, 1989; Slavin, 1995). Instructors designing interdisciplinary team experiences may draw from these models to structure tasks that emphasize mutual reliance and clear expectations for participation.