Is group work beneficial for producing creative designs in STEM design education?

Creativity in design, or design creativity, is a significant element for new product development, which occurs between the problem and solution space (Demirkan & Afacan, 2012). It is widely considered an essential part of the conceptual design stage where successful designs are often originated from (Chiu & Shu, 2012; Crilly & Moroşanu Firth, 2019; Thompson & Lordan, 1999; Toh & Miller, 2015). However, there is a long-term debate surrounding the definition of creativity in classical research fields (such as psychology and cognitive science) and design research. Han et al. (2021) reviewed over 50 studies in various fields, with a particular focus on design, to explore how creativity is defined and assessed. They revealed that novelty and usefulness are consistently used as the core elements for defining and assessing creativity across the board. Novelty refers to both newness and originality, while usefulness refers to quality, feasibility and value (Chiu & Shu, 2012; Girotra et al., 2010; Han et al., 2021; Sarkar & Chakrabarti, 2011; Shah et al., 2003). Although Gero et al. (2019) argued that surprise should be included as the third element of design creativity for measuring the unexpectedness of a design, many other researchers, including Chiu and Shu (2012), Zheng and Miller (2020) and Koronis et al. (2019), indicated surprise as a nuance of novelty. Therefore, novelty and usefulness are considered the two key elements of design creativity. In other words, design creativity refers to the generation of novel and useful products, which is in line with the common design creativity definition by Sarkar and Chakrabarti (2008). Few studies have explored the links between novelty and usefulness, of which He and Luo (2017) found that an inventive design’s novelty profile can influence its usefulness value, while Han et al. (2021) showed that there are no significant relations between a product’s novelty and usefulness.

Innovation is a crucial factor for commercial success and viability in a competitive market (Bruton, 2011; Guo et al., 2017), which is highly related to product development. Creativity and innovation are often discussed simultaneously in design research, while creativity is considered central to innovative product development (Crilly & Moroşanu Firth, 2019; Shai et al., 2009). Moreover, Sarkar and Chakrabarti (2011) showed that creativity is considered the prerequisite for product innovation, Cropley et al. (2011) indicated that creative product is the initial point for product innovation, and Starkey et al. (2016) claimed that successful innovation depends on creative design ideas. This has implied that design creativity is a key driver of product innovation, ultimately leading to business success and commercial benefits. Cultivating design creativity to prepare students to be innovative designers and design engineers is thereby significant for design education.

Group work is a popular teaching strategy in higher education, and it is commonly asserted that students need collaborative group working experience prior to employment (Almond, 2009). Benefits frequently highlighted include that it can foster students’ abilities and skills in oral communication, written communication, collective decision making, group management, interpersonal dynamics issues, critical reflection, self-directed learning, and social interactions (Almond, 2009; Boud et al., 1999; Burke, 2011; Gupta, 2004). These skills and abilities are not discipline-specific but are transferable that could benefit lifelong learning. Although students generate one final ‘product’ in group work, it still provides opportunities for individuals to learn related concepts and knowledge (Pfaff & Huddleston, 2003). In addition, students are also able to learn from each other in several ways, such as giving and receiving help, sharing knowledge, observing others, and building on each other’s ideas, which will increase their knowledge and understanding (Webb & Mastergeorge, 2003). Mbalamula (2018) indicated that students perform better in groups than individuals, and the students’ scores increase while the number of students in a group increases. Group work also has the advantage of using limited resources effectively. For instance, Almond (2009) and Pfaff and Huddleston (2003) indicated that higher education institutions may reduce the material resources and marking effort by employing group work.

However, many researchers hold a sceptical view of the value of group work, despite its claimed benefits in higher education. Beebe and Masterson (2003) indicated that group work might cause issues such as peer pressure to conform to majority opinions, individuals dominating discussions, some members relying heavily on others, and time-consuming compared with individual work. Bacon (2005) indicated that students learned less in group projects than in individual projects. Pauli et al. (2008) claimed that motivational, interactional, and logistical difficulties are the main source of causing problems in group work. Lin et al., (2019) revealed that the two main flaws in group work projects are ‘social loafing’ and ‘unfair mark’. ‘Social loafing’ refers to the phenomenon that people spend less effort in group work than individual work for achieving a goal. ‘Unfair mark’ refers to the marks, which cannot truly reflect each individual’s contribution, given to the group members. One of the common methods to address the issues arisen in group work is to perform self and peer assessment among students (Elliott & Higgins, 2005). However, there are doubts about the usefulness of peer assessment, as indicated in the research conducted by Kennedy (2005) and King and Behnke (2005).

Several researchers have investigated the use of group work and individual work in different education settings. For example, Rickheim et al. (2002) showed that group work and individual work are equally effective at delivering key elements in medical health education, while group work is more cost-effective and efficient. Fernández Dobao (2012) revealed that groups perform better than individuals in terms of accuracy while conducting writing tasks. Cronin and McCabe (2017) showed that students working individually in laboratory-based science classes have significantly improved laboratory techniques in comparison with students working in groups. Lin (2019) indicated that students perform better in group work than individual work in terms of vocabulary learning, despite students prefer individual work. These examples show that the effects of group work vary in different educational contexts.

In design education, group work is employed to support students in experiencing and understanding collaborative design processes (Qu et al., 2019). The ability to work in a group is highly demanded by design practice (Thompson, 2002). Many design activities involve group work, especially when solving complex design problems (Daly et al., 2019). Early studies revealed that group work could lead to better clarifications of tasks, the generation of a large variety of ideas, and a more intensive analysis of the ideas generated, but dominant group members often influence group discussions, which might even interrupt group cooperation (Ehrlenspiel et al., 1997). In recent studies, Lee et al. (2017) investigated the cognitive challenges for design teams and their educators. Tucker (2017) proposed a framework for understanding the effectiveness of group work in design assignments, which supports educators to better integrate group work into their modules and evaluate learning outcomes. Qu et al. (2019) indicated that design educators are positive about the students’ results and effort in group work. However, the learning outcomes of each individual in group work varies greatly, depending on the course set up, cognitive levels, commitments, and group role-related knowledge. Mattioli (2019) raised the attention of the extensive use of cross-culture design teams, including disciplinarily diverse, demographically diverse and cognitively diverse, in nowadays education. Group work tends to be a widely used approach in design education (Thompson, 2002) with an increasing research interest, but whether group work is a beneficial approach for educating design creativity needs further investigation.

Creativity plays a significant role in practical design and is presented in all stages of the design process, including detailed design (Court, 1998). Several studies have explored the impacts of group work and individual work on creative design activities. Court (1998) indicated that group methods are favoured, while it is important to create a suitable environment and construct multidisciplinary groups. Taggar (2002) showed that the combinations of contributions from the individual group members at group level could lead to ideas that are more creative. Kohn et al. (2011) indicated that groups outperform individuals on idea combination, generating more novel and feasible outputs. Bissola and Imperatori (2011) showed that designers with poor individual creative skills can perform well in groups if organised properly. McMahon et al. (2016) also showed that it is beneficial to develop ideas in collaborative groups compared with individuals. By conducting a real-life setting case study, Petersson et al. (2017) revealed that group work is beneficial for the analysis of design topics and the generation of ideas to address the topics. However, researchers, such as Stroebe et al. (2010) and Putman and Paulus (2009), held the view that individuals are more creative and perform better in brainstorming compared with groups. Nevertheless, Linsey et al. (2011) indicated that individual and group work are both important in idea generation.

Design education is considered the base of knowledge about design creativity (Rauth et al., 2010). Therefore, fostering design students with creative skills and methods is considered an important part of their education stage and future careers. In addition, creativity is also considered a significant element of the UK design course accreditation criteria of the Institution of Engineering Designers (IED) and the Chartered Society of Designers (CSD) (Dowlen, 2019).

Producing ideas is essential in design, which has a significant impact on the success of a product. It is revealed that there is a strong link between the raw ideas of a product and the consumers’ purchase intent of the product (Kornish & Ulrich, 2014). However, it can be challenging for designers to come up with creative ideas, particularly design students (Wong & Siu, 2012). Therefore, educating students with creative idea generation skills and methods plays a significant role in the curriculum of design education. Many studies have investigated how to prepare students with the ability to generate creative ideas, as it significantly affects the quality of the final creative outputs (Zhang et al., 2015). For example, Hernandez et al. (2014) investigated the use of TRIZ, Sketching and Smartpen in a design course, and indicated all three aids had enhanced the effectiveness of students’ idea generation while TRIZ was the best in improving novelty; and Daly et al. (2019) showed the use of the 77 Design Heuristics (Yilmaz et al., 2016) in design classrooms could support the exploration of more innovative ideas. Many creativity tools or ideation techniques are employed in design classes for supporting design students in creative idea generation, which involves:However, some of the tools are suitable for group work while some others are more suitable for individuals. For examples, six thinking hats (De Bono, 1985), C-sketch (Shah et al., 2001), and brainstorming (Osborn, 1979) including its variations, such as brain-writing, sticky note and flip chart brainstorming, are more suitable for groups than individuals. Creativity tools, such as TRIZ (Altshuller, 1984), are more suitable to be used by individuals rather than groups. Linsey et al. (2005) showed that group idea generation techniques, such as C-Sketch, could support designers in generating ideas with a larger quantity and better quality. In addition to these conventional idea generation techniques, computational creativity support tools have been developed in recent years for supporting designers in creative idea generation, such as Idea Inspire 4.0 (Siddharth & Chakrabarti, 2018), the Retriever (Han et al., 2018b), the Combinator (Han et al., 2018a), and InnoGPS (Luo et al., 2019). Although most of these computational tools are aimed for supporting individual designers, few of the tools have been employed in the current design education.

Design creativity assessment is also central to the curriculum of design education, which often occurs while students are developing their design ideas and concepts (Georgiev & Casakin, 2019). In design practice, many methods, involving Consensual Assessment Technique (CAT) (Amabile, 1982) and Creative Product Semantic Scale (CPSS) (O'Quin & Besemer, 1989), are often employed to assess the degree of creativity for product ideas and concepts. In design education, students evaluate the ideas generated and select the best or the most preferred one by employing idea evaluation and selection methods, which are not aimed at assessing creativity specifically, rather than using design creativity assessment methods. However, idea evaluation and selection is still considered a challenging task for students, which involves compromise, judgment and risk (Oman et al., 2013). Furthermore, many textbooks used by educators for teaching STEM design courses, including the ‘Engineering Design: A Systematic Approach’ (Pahl & Beitz, 2013), the ‘Delft Design Guide: Design Strategies and Methods’ (Van Boeijen et al., 2014), and the ‘Mechanical Design Engineering Handbook’ (Childs, 2018), provide useful methods for evaluating and selecting ideas, but not considering creativity as a design requirement in general. Several methods often used in design education for idea evaluation and selection are listed as follows:These idea evaluation and selection methods are often used by both groups and individuals. Faure (2004) showed that the quality or novelty of the ideas selected by a group of individuals and an interactive group do not have significant differences. However, Nijstad and De Dreu (2002) suggested that the selection of high-quality ideas performs better in groups rather than individuals. Rietzschel et al. (2006) indicated that idea selection is a judgment task, where group discussion could improve the selection process. They revealed that groups tend to select more novel ideas compared with individuals.

According to the preceding reviews, both group and individual work have their advantages and disadvantages, but group work tends to be more popular and effective in design education and creative design activities. However, few practical studies have been conducted that could provide the direct evidence. Therefore, a case study is conducted in the next section to yield more practical insights.

In order to investigate the three hypotheses proposed in this paper, a case study has been conducted to explore and compare the impacts of employing group work and individual work on design creativity. The case study involves two cohorts of year two undergraduate students studying the same product development module within an Engineering degree Industrial Design programme at a university in the UK. The two cohorts are provided with the same design project, while the first cohort employs group work and the second cohort employs individual work. The project assignments, in forms of design portfolios, produced by the two cohorts are assessed by the module teacher using the same marking rubric to ensure the two cohorts possess similar intellectual abilities for a fair comparison. Expert assessment is then introduced to evaluate the degree of creativity for the final designs produced and followed by statistical analysis for examining the three hypotheses. The participants, instruments and implementation of the case study are presented in the following sub-sections, while the results and statistical analysis are provided in “Results” section.

The first cohort (cohort A), which involves 32 students, studied the product development module in 2018/2019 academic year, while the second cohort (cohort B), which involves 30 students, studied the module in 2019/20 academic year. Both of the cohorts were taught by the same module teacher and used the same set of teaching materials. Students from both cohort A and B were admitted into the university under the same entry requirements, and studied the same set of modules in their year one studies.

The two cohorts were given the same design project ‘a new manual juicer design’ as their module assignment. Cohort A was required to conduct their design projects in groups of four (8 groups in total), while students in cohort B conducted their projects individually without any collaborations. However, students in cohort B were also provided with the additional opportunity to work on projects of their own choices, such as ‘a new garlic press design’ or ‘a new portable camera handle design’. 15 students in cohort B, who selected the ‘a new manual juicer design’ project, are included in the case study for comparison with the 8 groups of students in cohort A. The 15 students are named as cohort B* to distinguish from cohort B in the following research. The remaining students, who worked on their own project choices, are excluded from the case study. This is to deliver a fair comparison between the final designs produced by students employing group work and individual work.

The participant demographic information of cohort A and B, including B*, is provided in Table 1. The age of a student is calculated based on the student’s date of birth and the date that the module started. Please note that a mature student was involved in cohort B but not cohort B*. All the students signed up with standard case study protocols giving permission to use the data. However, the intellectual properties of the designs created by the students are owned by themselves according to the University’s Intellectual Property Policy.

In the product development module, marking rubrics were distributed to both cohorts at the beginning of the module to provide students with the information regarding the marking criteria, which were used to assess the students’ portfolios produced. The rubrics are also employed in the case study as the instrument for measuring the students’ intellectual abilities.

Comparing with the marking rubric used to assess cohort A in 2018/2019 academic year, as shown in Fig. 1a, a slightly different version was used to assess cohort B in 2019/2020 academic year, as shown in Fig. 1b. Both of the rubrics contain the same set of main marking criteria, ‘portfolio presentation’, ‘design phase and outcome’, and ‘presentations of design’, of which each criterion contains a sub-set of marking items. Two marking items in the criterion ‘design phase and outcome’, ‘concept generation and selection’ (10% marking weighting) and ‘demonstration of concepts’ (10% marking weighting), in the marking rubric for cohort A were combined into ‘concept generation, selection and demonstration’ in the one used for cohort B, but maintained the same overall weighting of 20%. Besides, a 5% marking weighting was deducted from the marking item ‘sketches, drawings and prototypes’ in the ‘presentations of design’ criterion and added to the item ‘demonstration of engineering knowledge’ in the ‘design phase and outcome’ criterion regarding the marking rubric for cohort B. The two marking rubrics used to assess cohort A and B were almost identical, despite the two slight changes. Therefore, it is considered that the changes made in the marking rubrics have zero to minimal impacts on students’ learning. In addition, creativity, as well as other related terms, such as idea generation, idea evaluation and selection, were only mentioned in marking items ‘demonstration of concepts’ and ‘concept generation and selection’ in the marking rubric for cohort A and ‘concept generation, selection and demonstration’ for cohort B. Therefore, creativity was not a direct driving force for learning perceived by students from both cohorts.

Creative Product Semantic Scale (CPSS), proposed by O'Quin and Besemer (1989), is a creativity assessment method for measuring products. This assessment method has been validated several times (Chulvi et al., 2012). Chulvi et al. (2012) defended the use of an adapted version of CPSS focusing on novelty and resolution (usefulness). These two criteria, novelty and usefulness, are also claimed as the key elements of design creativity (Sarkar & Chakrabarti, 2008). Thereby, this adapted version of CPSS, as shown in Fig. 2, is employed in this study to measure the creativity of the designs produced by the students.

As shown in the figure, the adapted version of CPSS involves 18 bipolar pairs of items, of which 8 pairs refer to novelty (the rows in grey shadings) and 10 pairs indicate usefulness. Please note that the grey shadings are only shown here to provide the readers with a better understanding of the creativity assessment method used, while no shadings are used in the case study. Each bipolar pair of items is assessed using a seven-point Likert scale ranging from ‘1’ to ‘7’. The novelty score of a design is calculated based on the 8 pairs of novelty items, and the usefulness score is calculated according to the ten pairs of usefulness items. The overall creativity is produced by adding up the 18 pairs of novelty and usefulness items. Thereby the maximum scores of novelty, usefulness and overall creativity are 56, 70 and 126, respectively.

Assessing design concepts is often led by experts, as it is a complex multi-criteria decision-making process (Zhai et al., 2009). Expert assessment has been used in a number of design studies for evaluating creativity, such as the ones conducted by Han et al. (2018c) and Sarkar and Chakrabarti (2011). Expert assessment is thereby employed in this study to measure the creativity of the final designs, by employing the adapted version of Creative Product Semantic Scale (CPSS).

The case study has been performed in the project-based product development module led and taught by the principal author. Project-based learning is commonly used in design education that engages students in learning, and has a positive impact on enhancing students’ creativity (Hanif et al., 2019; Lee, 2009). The module aims to teach intermediate elements of new product design and development, reinforce the role designers play in the development of new or existing products, and further develop students’ drawing, graphical, presentation and design communication skills. The module requires students to design and develop a new consumer product through comprehending a given design problem, conducting user research, generating and selecting ideas, and developing the ideas into final design products. As a STEM design module, it also requires students to conduct engineering analysis, such as material and manufacturing considerations.

The two cohorts of students, cohort A and cohort B (including B*), were given the same amount of time, 12 weeks, to accomplish their design projects and submit their assignments in the format of design portfolios. Before comparing the manual juicer designs produced by cohort A and B* regarding creativity, it is necessary to ensure that both cohorts are at a similar intellectual level (knowledge and skills). The module teacher re-marked the 8 group work portfolios produced by cohort A (N = 8), simultaneously while marking the individual portfolios of cohort B, including cohort B* (N = 15), using the same marking rubric (the one for cohort B) and employing the same marking standard. This could also eliminate the effects on marking caused by the time elapsed and the slight changes in the marking rubrics.

Cohort A achieved an average grade of 66.34 out of 100, with a median (Mdn) value of 67.50 and a median absolute deviation (MAD) value of 1.50, while cohort B* achieved an average grade of 65.67 (Mdn = 6.00, MAD = 2.00). A Mann–Whitney U test is conducted to measure the statistical differences between the assignment grades of cohort A and B*. It indicates that cohort A and cohort B* are not significantly different in terms of their assignment grades, U = 55.50, Z = −0.29, p = 0.77, r = 0.06. This shows that cohort A and B* possess similar intellectual abilities, including knowledge and skills, for a fair comparison.

The degree of creativity of the final designs produced by cohort A and cohort B* are then assessed to yield practical results for examining the three hypotheses proposed. Six design experts with an average of 9.17 years (SD = 4.41 years) of design experience participated in this case study voluntarily, and signed up with standard case study protocols to not disclose the data used in the case study. The six design experts were asked to assess the 23 final design individually using the adapted version of CPSS. Before the design experts started the formal creativity assessment, short briefing and training sessions were provided. In order to avoid distracting the six design experts from focusing on the 23 final designs in the portfolios produced by the 8 groups in cohort A and 15 individuals in cohort B*, the final designs, involving the presentation of design, the context of use and engineering considerations, were extracted from the portfolios. The 23 final designs were collected and then mixed for creativity assessment to avoid evaluation bias and enhance reliability. Examples of the extracted final designs, one from cohort A and one from cohort B*, are shown in Fig. 3. The two examples attracted assignment grades of 68 and 69, respectively.

A Cronbach’s alpha test is then conducted to measure the internal consistency of the assessment results among the six design experts. The test is conducted based on the CPSS bipolar pairs of items of the 8 group work designs produced by cohort A, and the 15 individual work designs produced by cohort B*. Each expert assessed 23 designs in total, with 18 CPSS bipolar pairs of items per design. The Cronbach’s alpha value α = 0.66, which indicates an acceptable level of internal consistency as a convention (Ursachi et al., 2015). This shows that the reliability of the expert evaluation is acceptable.

As the module teacher and leader of the product development module, the principal author was required to change the module from employing group work in 2018/2019 academic year to individual work in 2019/2020 academic year according to the recommendations of an accreditation visit. This provided the opportunity for the authors to study how group work and individual work affect students in learning design knowledge, especially creativity.

Although many existing studies have emphasised the significance of employing group work in design education, the case study conducted has shown that group work and individual work have no significant differences in the final designs produced, in terms of novelty, usefulness, and overall creativity. Therefore, the empirical results of this study do not support that group work is more beneficial for supporting students in producing creative design outputs in STEM design education in comparison with individual work. Further correlation analysis suggests that the level of creativity (including novelty, usefulness and overall creativity) of a final design is not related to its corresponding assignment’s grade. In other words, there tends to be no relationships between a student’s academic performance and creativity performance in the product development module. As a result, whether a student could produce a highly creative design is not significantly associated with the learning strategy used (group work or individual work) and even the student’s academic performance, for the case study concerned.

Several factors could have caused these results. One of them could be the disadvantages of group work, such as peer pressure and the existence of dominant members, revealed by many researchers. These disadvantages might have affected the students, which employed group work, in learning and applying creativity knowledge and skills. For example, Wiltermuth (2009) showed that a dominant group member may influence other submissive members in contributing fewer ideas to group discussion. This leads to a reduction of the number and diversity of ideas produced by the group, which impairs the capability of the group in generating creative ideas. As a consequence, designs generated by employing group work and individual work could be at a similar creativity level. Therefore, if design educators are equipped with better understandings and management skills of group work (Davies, 2009; Woolard, 2018), there is a potential to better support groups of students in producing creative designs in the context of STEM design education.

Starkey et al. (2016) revealed that there is a reduction of creativity throughout the design process, as students often abandon novel concepts. It is suggested that individuals’ definition of a ‘good’ idea affects the ideas they select (Johnson & D’Lauro, 2018). For example, if a ‘good’ idea is defined as a novel and useful idea, then individuals might select ideas that are creative. Rietzschel et al. (2010) indicated that students prefer to select feasible and desirable ideas rather than novel ones after idea generation. Similarly, Toh and Miller (2015, 2016) demonstrated that groups that produce highly creative ideas do not necessarily select creative concepts, as students focus on feasibility and neglect creativity during the selection process, despite creativity being set as a design requirement. Thereby, students involved in this study might have selected ideas, which they believed would meet the requirements of the assignment, but not creative ones for further development. This could be another main reason underpinning the fact that the creativity of the final designs produced by group work (cohort A) and individual work (cohort B*) are at a similar level, not relating to the assignment grades. Design creativity assessment methods should thereby be employed in design education, rather than merely using the existing idea evaluation and selection tools. This could support students in selecting creative ideas that are novel and useful, rather than desirable ideas perceived by the students.

Another factor might be that creativity was not considered a significant item in the marking rubrics used in the case study, as shown in Fig. 1. The rubrics were provided to the students from both cohorts during the ‘week one—introduction’ class of the module. Therefore, students might have spent more time on marking items which have higher weightings, such as ‘sketches, drawings and prototypes’, rather than creative activities, such as idea generation and evaluation, to achieve a better overall assignment grade. Thus, increasing the weightings of creativity related marking items has the potential to drive students to focus on creative design activities. This might result in highly creative outputs for both group and individual work.

However, the results of the case study might also be determined by the design project ‘a new manual juicer design’ employed in the module. A manual juicer is a classic product, with numerous existing ones on the market, which has rather stable and maturely defined functions in general. It might be challenging for students, even design experts, to come up with radically new and path-breaking manual juicers that are considered highly creative. Therefore, innovative design projects, such as ‘a portable blender’, could be employed in future studies to provide more insights.

In addition, design students (novices) think and work differently from expert practitioners. For example, design novices tend to use ‘trial and error’ to approach design tasks, while experts employ particular design strategies (Ahmed et al., 2003); design experts spend significantly more time on problem scoping and information gathering than design students during the design process (Atman et al., 2007); design experts prefer to use analogical reasoning for problem solving and identification, while novices lack such reasoning processes (Ahmed & Christensen, 2009). It is also shown that the creativity of designs produced is closely related to the designer’s knowledge of managing and monitoring the solution generating process (Christiaans & Venselaar, 2005). In addition, design experts are often involved in designing complex systems and products, which requires cooperation and group work. Hence a driver for design experts to master group work skills. Therefore, group work could be a useful and effective approach to support the generation of creative design outputs for design experts, but might not be effective for design students.

Group work offers a series of potential benefits and attractions in higher education, especially in delivering STEM design programmes. It can help students understand the nature of collaborative design processes, and fosters sought-after employability skills for future professional design activities (Qu et al., 2019; Thompson, 2002). Furthermore, the use of group work could reduce the resources needed, such as staffing and marking effort (Almond, 2009; Pfaff & Huddleston, 2003), as well as meet the requirements of design course accreditations (Dowlen, 2019). However, Kim et al. (2008) indicated that design team creativity is related to the personal creativity of team members. Individual work can help students cultivate independent thinking and decision-making abilities, as well as conducting work without depending on others with accompanying ease of assessment. For the case study concerned, both group work and individual work have their advantages and disadvantages, and do not show significant differences in generating creative design outputs. We thereby suggest a mixed-use of individual work and group work in design education to foster students with an extended range of skills and abilities, as well as provide different learning experience.

From the students’ perspective, successful group work could be challenging, as it requires students to have self-efficacy beliefs to encourage each other to utilise resources effectively and engage in high-quality group discussions (Bandura, 2001; Chang & Brickman, 2018; Wang & Lin, 2007), and appropriate communication styles to organise group work and cooperate with one another in both monocultural and multicultural group settings (Marks et al., 2000; Popov et al., 2012). In order to achieve successful individual work, students need to have the ability to set their own objectives and monitor their own progress, the responsibility for their own studies, and the willingness to evaluate and reflect their own learnings (Hockings et al., 2018). In addition to the capabilities and skills required for group and individual work, students also have differences in their learning styles. Activist, reflector, pragmatist and theorist are the four basic learning styles (Mumford & Honey, 1992), while each style might lead to different work preferences. For instance, activists prefer small group discussions and training others but disfavour self-reading, writing and thinking (Honey & Mumford, 2000), so it might be expected that activists could perform better in group work than individual work. As a result, there is a need to consider all these differences among students in project work. Furthermore, Crilly and Moroşanu Firth (2019) indicated that the combination of group work and individual work have creative advantages. Therefore, integrated approaches that combine both individual and group work within a single project could also be employed by design educators.

Although creativity has been considered to play a significant role in STEM design education, more effort is needed in teaching design creativity to foster creative designers and design engineers. For instance, design educators could highlight the importance of creativity in their marking rubrics, by adding more weightings on creative design activities, such as idea generation and evaluation. In group work-based design projects, educators need to better manage the groups to avoid the negative impacts caused, which might impair the groups’ creative abilities. The use of design creativity assessment methods in design education is also necessary for supporting students in delivering creative design outputs. However, design educators should employ creativity assessment methods that are suitable for students, such as the Comparative Creativity Assessment (CCA) and the Multi-Point Creativity Assessment (MPCA) (Oman et al., 2013), rather than expert oriented methods, such as Consensual Assessment Technique (CAT) (Amabile, 1982). In addition, design educators could also collaborate with design creativity researchers to come up with new tools, involving computational ones, to support design students in creative design activities, especially design creativity assessment.

While this study compares group work and individual work in producing creative design outputs, there are a few limitations. First, the study involves 47 valid participants from two cohorts of students, 32 from the first cohort and 15 from the second cohort, studying the Industrial Design programme. The relatively low number of students has led to a limited number of assignments samples, with 8 group and 15 individual design portfolios from the first and second cohort, respectively. However, the programme was designed to accommodate such a small number of students in each cohort, due to the nature of the programme at the university. Therefore, a similar case study could be conducted within a STEM design programme that contains a large cohort size, or even at multiple universities around the world, to yield a more general and universal result.

Secondly, marking rubrics were used in the study as the instrument to assess the project assignments as a convention in the university, and to measure the intellectual abilities of the two cohorts for a fair comparison regarding their creativity performance in the product development module. However, rubrics have inherent weaknesses in terms of reliability and validity. Although the marking rubrics used cover a broad range of assessment criteria, the results obtained could be a function of the rubric chosen. One solution to this issue is to improve the clarity and appropriateness of the language used, which is suggested by Reddy and Andrade (2010), for enhancing the validity of the rubric. Another solution is to employ a different approach, such as comparative judgment and performance test, instead of using marking rubrics. Such approaches could provide more reliable and valid results (Pollitt, 2012) regarding the assessment of the students’ knowledge and skills through their assignments.

Thirdly, certain aspects of the students’ portfolios were removed, such as user research, idea generation and selection processes, for the creativity assessment to help the experts focus on evaluating the final designs which involve the presentation of design, the context of use and engineering considerations. However, this might have removed the insight into students’ decision makings around novelty versus usefulness, which could have potentially obscured a level of authenticity and affected the assessments made. Therefore, the whole portfolio could be provided to experts for creativity assessments in future research. This would require the experts to affirm stronger commitments to the assessments, as creativity assessment is a multi-decision making process that is often time-consuming. Alternatively, another group of experts could be employed to analyse the students’ whole design process to provide more insights on their creativity performance from the reality and complexity of designing and innovating.