Experiences with team-based learning in an introductory bachelor course on sustainability

ABSTRACT Team-based learning (TBL) is a structured form of collaborative learning that is particularly beneficial in courses where students are expected to understand a significant amount of information to answer complex questions. Here we evaluate the implementation of TBL in a second-year undergraduate sustainability course. The course introduces structured and quantitative approaches for analysing human impacts on the natural environment. It consists of four learning units each focusing on a specific environmental issue, such as climate change and habitat loss. Teams of 5–6 students are formed at the beginning of the course. Each learning unit starts with individual pre-class preparation followed by a readiness assurance process. The remainder of the learning unit consists of assignments that require students to apply what they learned to environmental problems. In the peer evaluation, the students assess team members on their contribution to the team activities. The exam pass rates have been consistently high (> 82%) since we implemented TBL in 2016. The students’ appreciation of TBL increased over time, with 90% of respondents rating the added educational value of TBL as satisfactory or better in 2019. Teachers value the active participation of the students. Students repeatedly mentioned that they highly appreciate the collaboration in a team, increased both engagement and motivation. TBL’s biggest challenge is the facilitation of the discussions during the application sessions and making sure that the discussions remain concise while maintaining sufficient depth. TBL is now also implemented in other courses with structural attention to the development of collaborative skills.


Introduction
Human activities have an unprecedented influence on the Earth system, including changes in climate, land cover, nutrient flows and biodiversity. The global anthropogenic changes are so pervasive that some authors argue that we have entered a new geological epoch, named the "Anthropocene" (Crutzen 2002;Waters et al. 2016). It has even been argued that the environmental changes brought about by humans may lead to abrupt or irreversible transitions in the state of the planet, ultimately compromising the persistence of humanity itself (e.g. Rockström et al. 2009;Steffen et al. 2015). In 1987, the World Commission on Environment and Development published the report "Our Common Future", also known as the Brundtland report. They argued that the key to solve environmental problems is sustainable development, defined as "the development that meets the needs of the present without compromising the ability of future generations to meet their own needs" (World Commission on Environment and Development 1987). In 2015, the UN General Assembly adopted the 2030 Development Agenda with 17 sustainable development goals, including climate action, and saving life on land and below water (https://sdgs.un.org/goals). While the need to address sustainability problems is increasingly recognized, it is also acknowledged that environmental challenges are highly complex to understand and combat in their full breadth.
Higher education institutions play a crucial role in addressing these sustainability challenges, as they are responsible for educating the new generation of scientists, academic professionals and policy makers that will be involved in the transition to a more sustainable world (Leal Filho 2000). Team-work and working on real-life cases has been recognized as key elements of effective training in addressing "wicked" sustainability problems (Brundiers and Wiek 2013). This is why many sustainability courses and programmes work with active learning methods, such as problem-based and project-based learning, addressing real-world sustainability challenges ( Van de Laar et al. 2000;Cörvers et al. 2016). To our knowledge, however, no sustainability courses and programmes apply a formal learning concept that explicitly combines individual responsibility for pre-learning with learning from peers in a problemsolving context.
Other disciplines dealing with complex questions, including medicine, pharmacy, life sciences, and economics, readily implemented team-based learning (TBL) as an educational method in their curricula (see e.g. Jarjoura et al. 2015;Reimschisel et al. 2017;Swanson et al. 2019;Arcila Hernández et al. 2021). TBL is a structured form of collaborative learning that provides students with opportunities to apply conceptual knowledge through a sequence of activities including individual work, teamwork, and immediate feedback. TBL works with small groups, which is applicable to large classes. Individual student accountability within the teams is achieved by including pre-class preparation combined with readiness assurance testing and immediate feedback from teachers and students (Burgess et al. 2020). Larry Michaelsen originally designed TBL as a solution to the lack of effectiveness to learn from lectures by large groups of students (Michaelsen et al. 1982). TBL is considered to be particularly beneficial in courses where (1) students are expected to understand a significant amount of information and (2) a primary goal of the course is to apply or use this information for answering complex questions (Michaelsen and Sweet 2008;Swanson et al. 2019). TBL is known to improve content knowledge, end-of-course grades, student engagement and allows for deeper understanding of content compared to passive learning methods, such as lectures (Sisk 2011;Allen et al. 2013;Altintas et al. 2014;Swanson et al. 2019).
Here we explain and evaluate the implementation of TBL in a second-year undergraduate sustainability course. The course provides an introduction on how to analyse the impact of human society on its environment, as a crucial component towards identifying solutions for sustainability challenges. Below, we first explain the TBL concept in more detail. After that, we show how TBL has been implemented in the course Man and Environment. Finally, we discuss the performance of the students, experiences of students and teachers and provide conclusions on recommendations for the implementation of TBL in sustainability-related courses.

Overview
In a TBL-based course, the content is typically divided into distinct learning units. While each learning unit has specific learning objectives, every unit follows the team-based instructional activity sequence (Figure 1; Michaelsen and Sweet 2008). Before any in-class activities take place, students must prepare by studying assigned materials. After the preclass preparation, each learning unit starts with the readiness assurance process (RAP). This process consists of an individual readiness assurance test (iRAT), a team readiness assurance test (tRAT), a written appeal and corrective instruction by the teacher. The RAT is a short multiple choice test on the key elements of the reading material. Each student first individually completes the RAT. Without knowing the results, the students then do the same test again, but now jointly as a team of 5-7 individuals, requiring the team to reach consensus on the answers. The team receives immediate feedback on the outcome of each test question (right or wrong) and the team gets the opportunity to answer the question again, until the right answer is given. When the team disagrees with an answer, it has the opportunity to formally send in a motivated appeal that the teacher needs to respond to. The final step in the readiness assurance process is a corrective instruction session led by the teacher, with a specific focus on clarifying any issues that turned out challenging during the team test. Once the readiness assurance activities are completed, the rest of the learning unit is spent on activities and assignments that require students to practice applying what they learned to complex, relevant problems. The main role of the teacher is to properly design and facilitate the overall learning process, while the students take responsibility for the pre-class preparation and the in-class teamwork.

Essential elements
Successful implementation of TBL depends on (i) the proper formation of teams, (ii) frequent and timely feedback to students, (iii) the design of group assignments promoting both learning and working in teams, and (iv) accountability of students for their individual and group work, (Michaelsen and Sweet 2008;Burgess et al. 2020). These four essential elements are further explained below.

Teams
Teams are composed by the teacher with the intention to have a balanced mix of student characteristics, such as nationality and scientific background. The idea of high diversity in the teams is to have students who can bring in different perspectives. The teacher is in charge of forming the teams to minimize the risk of having subgroups within a team that may jeopardize the team-spirit. As groups need time to develop into effective learning teams, students stay in the same group for the entire course (Michaelsen et al. 1982).

Feedback
Feedback is immediately provided to students during the team test, with further clarification provided by the teachers in a corrective instruction session. This direct feedback is important for the TBL process, identifying and addressing potential gaps in understanding of the students and stimulating critical thinking (Burgess et al. 2020).

Applications
During the applications, teams are required to use their knowledge to address complex problems representative of real-life situations. Michaelsen and Sweet (2008) recommend that the "four S's" of complex problem solving should be applied in assignments: • significant problem: working with significant real-life problems is key in TBL.
Backward design is typically used to first identify what will be significant and relevant topics for the students to deal with: what are the learning objectives and what case is worth selecting to study? The challenge is to develop applications that stimulate discussion with no clear-cut answers. • same problem: group assignments should promote discussion both within and between teams. When teams are working on the same problem at the same time, they will have a greater involvement in discussions. When each team develops thorough knowledge of the problem, this allows for a more informed and engaged discussion not only within but certainly also between the teams on possible solutions.
• specific choice: assignments include a list with possible, well-founded solutions to the problem of which each team has to independently select the best option. Team members all come to agreement on a single, clearly-defined answer for a complex problem. This can be done via multiple-choice questions, provided the question is not trivial, or via an open question where the teams formulate short answers. • simultaneous reporting: responses from all teams are reported to the class at the same time. Simultaneous reporting motivates the teams to seriously engage in the activity, since the response will be available for all the other teams. Another advantage of simultaneous reporting is that there is no possibility for teams to modify their answer based on the responses of other teams. Finally, simultaneous reporting typically creates excitement and engagement among the teams in the classroom.

Accountability
Students must be accountable for their individual preparation as well as their contribution to the team (Sibley and Ostafichuk 2014). The grades of the iRATs provide a measure of individual accountability for pre-class preparation, while the team performance is measured by the outcomes of the tRATs. Finally, peer evaluation is considered an important element of TBL concept. Peer evaluations hold students accountable for their level of participation. Students contribute to the grades of other students via quantitative and qualitative feedback to their respective team members (Burgess et al. 2020).

Motivation for implementing TBL
In the period 2013-2015, the Man and Environment course was based on lectures and (computer) assignments. The course evaluations indicated that a significant share of the students was not highly motivated to actively participate in the course.
In 2016, we decided to change the teaching format of the course to an active learning strategy, inspired by the implementation of TBL in the curricula at the Faculty of Medical Sciences at the Radboud University.
To better understand the basics of the didactic concept of TBL, the teachers of the Man and Environment course first followed a one-day TBL training at the Radboud university medical centre in August 2016. From September to November 2016, the course was changed into a TBL concept with help from two TBL experts from the Center for Evidencebased Education, located at Amsterdam UMC. The experts first organized workshops for elucidating how the course should contribute to active problem-solving capacities of the students in the sustainability domain. After it was clear what the students were expected to learn, the experts helped the teachers to develop (i) applications to stimulate teamdiscussion for better understanding how real-life sustainability problems can be addressed, and (ii) self tuition materials and related RAT questions to provide the necessary knowledge for an informed discussion during the applications. The Man and Environment course in the TBL concept was given for the first time in December 2016 and has now been running for five years.

Context
The course Man and the Environment is a 3 EC course scheduled for two days a week over a period of four weeks for all BSc Biology students (100-200) at the Faculty of Science, Radboud University in Nijmegen. The teachers involved in this course consist of a fixed core of four teachers, including the coordinator, supported by assistants who help with the computer assignments.

Aims
The main learning aims of the course are to: • Report on the cause-impact pathways for four main environmental themes, i.e. plastic pollution, climate change, habitat loss & fragmentation, and chemical pollution. This is done by using the so-called DPSIR-framework (European Environment Agency 1999), which stands for Driver -Pressure -State -Impact -Response ( Figure 2). • Explain the underlying scientific mechanisms that relate human pressures to environmental impact for plastic pollution, climate change, habitat loss & fragmentation, and chemical pollution. • Identify management strategies to reduce the impact of plastic pollution, climate change, habitat loss & fragmentation, and chemical pollution.

Team formation
We compose teams of five to six students at the start of the course ensuring a mix of gender, background (medical biology vs. general biology) and secondary school location. The students stay for the full course in the same team.

Pre-class preparation
On the first day of each learning unit, students independently study the corresponding theme by going through the learning materials available in the digital learning environment Brightspace. The materials include articles, videos and links to websites, with an estimated required time investment of eight hours of self-study for each theme.

Readiness assurance process
For the second day of each unit, we developed an intensive in-class programme with several activities. Following the established TBL format (see section 2), this day starts with an iRAT of 20 multiple choice questions, followed by the tRAT with the same questions immediately afterwards. Each team receives a scratch card for the tRAT, as a form of immediate feedback. Once the team agrees on an answer, they scratch open the box of their choice. If the answer is correct, they move on to the next question. If the answer is wrong, they have to scratch open a new box, until they find the correct answer. The team earns full points when the answer is correct on the first go, and points are lost for each incorrect attempt ( Figure 3). After the RATs, teams get the opportunity to write an appeal. Such an appeal may relate to ambiguity in the wording of the question or disagreement with the "correct" answer. Appeals are considered by the teacher, and if granted the points of the question are awarded to the team(s) that made Figure 3. Immediate Feedback Assessment Technique. Teams of students fill in their answers on a scratch card that provides immediate feedback: when a star is unveiled, the correct answer was scratched open. The number of attempts to get to the right answer can be used in the scoring process.
In the above example the maximum of score of four is halved with every additional attempt, and no points are given when four attempts were required. the appeal. Only the subjects that the students have not quite well understood, even after the team discussions, are selected for the plenary feedback session. In preparation, the teacher has slides available on all topics, but only the slides that relate to the identified knowledge and skills gaps are used.

Applications
After the Readiness assurance process, the students apply their knowledge in a 2-hour session of application assignments. The application sessions include eight to ten teams with a maximum of 60 students. The sessions are held in a collaborative learning room specifically designed for group activities (Figure 4). The collaborative learning room was furnished in 2019 and equipped with furniture that facilitates collaboration and technical support that allows information sharing. We use two ways of reporting in the applications. The first type of reporting in the application sessions are open questions for the environmental theme of that week. For example, each team has a large screen in the collaborative learning room that can be used to electronically draw the DPSIR with concrete examples for the environmental theme of that week in 15-20 minutes. After that, each team can see how other teams constructed their DPSIR. The teams are asked to select two other teams with in their view the best DPSIR. This gallery walk takes about 10-15 minutes. In a final discussion among the teams (10-15 minutes), teams are asked to give the arguments for their choice of the best examples. The teacher moderates the final discussion.
The second way of reporting is via cards marked A, B, C and D as answers to multiple choice questions. An example of a multiple-choice question is "Select which of the following mitigation efforts in the Netherlands is best": A. Closing a coal-fired power plant. B. Subsidizing bioenergy. C. Subsidizing electrification of the transport system. D. Installing a tax on meat consumption. Students first discuss within their team the arguments to arrive at the preferred solution and then simultaneously report by raising one of their cards (A-D). The teams seldom all hold up the same card. The teacher then acts as a discussion leader and prompts the groups to exchange the arguments for their answer. This allows for exploring the possible solutions together with the teams.

Peer evaluation
In the peer evaluation, the students critically (and anonymously) evaluate team members on their contribution to and participation in the team activities (tRAT and applications). The students learn to give and receive feedback so as to improve their skills as a team player. There are two peer evaluations: mid-term and end-term. Each of these peer evaluations consists of two parts: 1) giving qualitative feedback to each team member via answering eight questions (not graded); and 2) quantitative grading of their team members. An example of a qualitative feedback question is whether Team member x: A. Did not contribute enough knowledge, strong improvement required. B. Only contributed some knowledge, this should be improved somewhat. C. Contributed knowledge at the group average. D. Contributed a lot of knowledge (more than average). For the quantitative grading, the students give each of their team members a grade between 1 and 10, such that the sum of the grades does not exceed the size of the team minus one times 8. The rationale behind this cap is that it creates scarcity of points and forces students to be more critical of every team member's contribution. As an example, for a team of 6 students, 5 × 8 = 40 points can be divided among the other group members, for instance, 5, 7, 8, 10, 10.
Since 2020, after each peer evaluation, the students automatically receive the anonymous feedback from their team members. Developing our own peer evaluation software greatly facilitated the peer evaluation process for the teachers. Up to 2019, students filled out an online form and teachers had to bundle and anonymize this information per team member, and then distribute it among team members. This approach was very time-consuming and also prone to errors. With the new peer evaluation software, teams are imported from our digital learning environment, teachers set up an event with opening and due date, using a reusable evaluation format. After students filled out the evaluation, teachers can release the results anonymously by the touch of a button.

Computer assignments
We included computer assignments in the course on top of the team-based instructional activity sequence to further strengthen the quantitative skills of the students. During the computer assignments students are expected to apply their knowledge in a number of exercises. For instance, they calculate how much CO 2 a gas-fired power station emits over its full life cycle, or how much of a chemical substance accumulates in the human body.

Final exam
Students round off the course with an exam. The exam is 2 hours and consists of 45-50 questions to test the knowledge, comprehension and quantitative application skills of the students. The majority of the questions are multiple choice, providing four alternative answers of which one is correct and the others are distractors. Alternative formats include questions where students have to connect statements or terms with each other, where they have to rank or order given elements, or where they have to fill blanks based on a predefined list. Students must pass the test (grade ≥ 5.5 on scale from 1 to 10) to complete the course.

Grading
The grade of the Readiness assurance process (RAP) and the grade of the exam count equally for the final grade of the course (each 50%). The RAP grade is a weighted average of the iRAT (50% weight), tRAT (30% weight) and peer evaluation (20% weight). This means that the importance of the individual component (iRAT) and group components (tRAT and peer evaluation) of the RAP are weighted equally. Note that the first iRAT, tRAT and peer evaluation are formative, so the students can get used to the grading. The grades of the second and final peer evaluation are summative. The grade of the iRAT and tRAT is the average of the two highest scores of the three summative tests.

Covid-19 adjustments
In the years before the Covid-19 pandemic, all TBL activities took place on campus. Students performed the iRAT on paper and teams used a paper scratch card to do the tRAT (Figure 3). For the team discussions, multiple teams were in the same room using voting cards to report their answer for the application exercises (Figure 4). During the Covid-19 pandemic in 2020, we included a number of adjustments to still be able to run the course with the TBL format.

Readiness assurance process
The Readiness assurance process was conducted via video conferencing in Zoom with breakout rooms. For the iRAT, we switched from tests on paper to tests in our digital learning platform, i.e. via quizzes in Brightspace. Students had to switch on their cameras in Zoom, so that lecturers were able to supervise the iRAT. For the tRAT, we developed an online scratchcard in which students can digitally scratch and lecturers can immediately view the results. Teams were assigned to breakout rooms, so students could together work together on their scratchcards. After all the teams finished the tRAT, students came back to the main virtual room for the instructor feedback.

Applications
In addition to the usual pre-class preparation, we asked students to prepare the application questions individually, as we expected that discussions online are typically less efficient than face-to-face. During the applications, the lecturer explained the assignment in the main room, then teams went into the breakout rooms to discuss and decide on their answer. Coming back into the main session, teams voted via the chat, and the lecturer facilitated the discussions between the teams. This procedure was then repeated for the next applications.

Final exam
The students did their final exam via the on-line examination software Cirrus with proctoring software.

Readiness assurance process
Grades for both the iRAT and tRAT were consistently high over the five years (2016-2020) that the Man and Environment course has been taught in the TBL format ( Figure 5). Grades below 5.5 out of 10 (the threshold to pass in the Dutch educational system) were rare, which indicates that the vast majority of students studied the learning materials well and were therefore well prepared for the remaining TBL activities. This corresponds with teacher observations of high student motivation and knowledge (see section 4.3). Students scored 30% [3-84%; 95%-confidence interval] higher on the team test (average Figure 5. Grades for the individual and team readiness assurance tests (iRAT and tRAT, respectively) in the Man and Environment course. Each blue dot represents the course-averaged grade of one student for the iRAT and tRAT within the 2016-2020 time period (N = 687 students). The Dutch scoring system of 1-10 is used.
tRAT grades of 8.9-9.4 over 2016-2020), as compared to the individual test (average iRAT grades of 7.1-7.3 over 2016-2020). These numbers indicate that interaction within the teams lead to an enhanced understanding of complex global problems, which is in line with findings reported in scientific literature (Michaelsen et al. 1989;Chung et al. 2009;Ngoc et al. 2020).

Peer evaluation
The grades that students anonymously received from their team members during the course, i.e. the peer evaluation grades, diverged relatively little ( Figure 6). This may indicate that students consider their peers to contribute similarly and fairly to the team, but also that students may be hesitant to give each other low grades. We hypothesize that both effects occur. Most teams seem to work well, as has been observed by the teachers in the classroom during tRAT and application sessions, and as supported by high tRAT grades and the predominantly positive qualitative peer feedback. However, students are new to TBL when they start this course and giving feedback is not a skill they have typically practiced much. More diverging grades in the peer evaluation feedback might be expected when students become more used to providing feedback to peers and when they work for a longer period as a team. For instance, Michaelsen et al. (2004) argued that students should stay in the same group for at least a semester to be able to develop their group into a cohesive, self-managed and effective learning team. Only then, students are able to engage in interactions freely without having to worry about making bad impressions, which has a positive influence on team performance. Note that the peer evaluation grades in the first year (2016) were substantially higher than in following years, because only in 2017 we introduced a limit on the total number of points to be distributed over the team members (corresponding with a maximum team-averaged grade of 8.0).

Exam
The pass rates for the Man and Environment course have been consistently high (> 82%) since TBL was introduced in 2016 (Figure 7). This may reflect that TBL forces students to start studying learning materials well in advance to prepare for the iRAT/tRAT and allows students to immediately apply their knowledge in application sessions. On average, grades and passing percentage of the first exam attempt increased after the introduction The percentages show the share of students that passed the first exam attempt. The Dutch scoring system of 1-10 is used, in which a grade ≥ 5.5 is required to pass (indicated by the red line) and 10 is the maximum score. TBL was introduced in 2016. Note that these grades do not include iRAT/tRAT or peer evaluation grades. of TBL. The average exam grade increased from 6.05 (± 0.12; 95%-confidence interval) before introduction of TBL (2013-2015) to 6.59 (± 0.08) after introduction of TBL (2016-2020) with strong statistical evidence (P < 0.001; two-sample difference of means t-test). The passing percentage of the first exam attempt after the introduction of TBL improved from 66.5% (± 4.3%) to 85.0% (± 2.7%) also with strong statistical evidence (P < 0.001; two sample difference of proportions Z-test).

Student appreciation
In the period 2016-2019, the students were given a (mandatory) questionnaire at the end of the course to give their opinion about: (i) the added educational value of TBL as an education method in the course Man and Environment, and (ii) the potential use of TBL in other courses within the Bioscience curriculum. Note that we did not send out the questionnaire to the students in 2020 due to time constraints caused by the Covid-19 pandemic. The added educational value of TBL was rated on a scale from 1-10. The share of satisfactory or better ratings (i.e. a score equal or higher than 6) increased from 73% in 2016 to 90% in 2019 ( Figure 8). Based on the qualitative feedback of the students, it is likely that the increasing student appreciation for TBL over time relates to a better organization of the course, including a newly available collaborative learning room for the applications, and more experience of the teachers with the TBL format. We also found that the percentage of students that were (strongly) against the idea to include the TBL format in more courses of the BSc biology curriculum decreased from 45% in 2016 to 20% in 2019. This increase in appreciation may be explained by the fact that indeed more courses in the BSc biology curriculum implemented the TBL learning strategy. Student perceptions of TBL also improve when it is used for a longer period. For example, after one semester of TBL, students are more likely to agree that solving problems in teams is an effective way to learn and practice (Espey 2010).
In the same questionnaire we also asked the students what they considered as positive aspects of TBL as implemented in the course. The majority of the students indicated that they highly appreciated the collaboration in a team. It was often mentioned that the team-based activities help them to understand the study materials, particularly via the tRAT, and that one can always rely on team-members in case of lacking knowledge. Students also mentioned that their teams made them appreciate input from others and that in some cases new friendships were made. The discussions during the application sessions were also often appreciated. Students mentioned that they learned a lot during the applications from their fellow peers by hearing different perspectives, as well as arguments they had overlooked. Students also indicated that the applications forced them to think in an active and functional way, which helped them to understand the complexity of environmental problems, and that exchanging different opinions made them much more aware of other perspectives. Students further argued that the applications resulted in a lot of interaction, provided a positive learning atmosphere, and stimulated them to get more used to speaking out loud in a bigger group. Some of the students stressed that the TBL format motivated them to keep track of the study material. Our findings are in line with a study by Remington et al. (2017) who compared TBL with traditional lecturing in a pharmacotherapeutics course and found that TBL enhances learning of course content, teamwork skills, and lifelong learning skills.
We also asked the students which aspects of TBL they did not appreciate. A negatively perceived aspect of the team work was that some students did not put in as much effort in the pre-class preparation as others, but still got good grades for the tRAT. The most prominent weakness mentioned about the application sessions was that the discussions were sometimes perceived to be too long without adding new insights. It was also mentioned that some teams were finished before the others with their internal discussion and had to wait too long (in their perception) before the group discussion started. Although teachers become more and more experienced in moderating the application sessions, it is indeed not always easy to keep a healthy speed in the discussions. Some of the students indicated that they would appreciate lectures to explain and prioritize the learning material for them. They miss clear guidance in prioritizing what is important to learn for the exam and indicated that TBL was in their perception an inefficient way of teaching without a clear idea what they have learnt during the course. This is in line with the findings of Deslauriers et al. (2019), who found that students in an active classroom learn more, but they feel they are learning less compared to listening to lectures. This phenomenon can be explained by the increased cognitive effort required for active learning. In the introductory lecture of the Man and Environment course we now explicitly explain this misperception to the students. Finally, students mentioned that the course schedule was too full with the RAP, applications and computer assignments all on one day. Although the RAP and the application are connected and should be scheduled on one day, we are indeed considering moving the computer assignment to a different day.

Teacher perception
We have asked the six teachers that have been involved in the Man and Environment course for their opinion about the strengths and weakness of TBL (as implemented in this course), the suitability of TBL for teaching environmental issues, and the consequences for the workload and job satisfaction of the teachers. In general, the teachers consider TBL effective in actively engaging students in the classroom. Working in teams, students learn from each other while practicing collaboration skills, critical thinking and scientific argumentation. Teachers see that students take shared responsibility for their learning process and that the lively discussions ensure students' engagement in the course. For example, one teacher wrote: "The most prominent active engagement are the applications where the students discuss within and between the groups complex environmental problems and possible solutions related to plastic pollution, climate change, land use and chemical pollution. It is in fact great to see both the engagement and high quality of the discussions during the applications, e.g. with underpinned arguments how to mitigate environmental problems".
When it comes to the greatest benefit of TBL, the teachers' views differ. Some mention the active and timely preparation of the students as the biggest advantage, while other teachers particularly appreciate the application of knowledge for solving complex issues or the communication skills students train during their interaction. One teacher specifically looked at the benefits from the teacher perspective, and mentioned that it is relatively easy to run a course once it is developed. When asked to consider the suitability of TBL for teaching environmental and sustainability issues, most teachers indicate that TBL is a suitable method to train students to apply their knowledge, and particularly to introduce them to the complexity of sustainability challenges.
According to most teachers, TBL's biggest challenge for the teachers is the facilitation of the discussions during the application sessions, in particular making sure that the discussion has sufficient depth and is based on sound arguments. One teacher indicated that "Moderating discussions and feedback during application activities requires much professional knowledge, improvisational skills and a great deal of experience". Multiple teachers experience the application sessions as challenging and sometimes stressful. This challenge is widely recognized within the TBL field and experienced facilitators often share their experience to support other teachers (Gullo et al. 2015). Another aspect that teachers indicate as being difficult and time-consuming, is the design of learning and teaching materials that both challenge students and support them during their learning process. Furthermore, when TBL is not widely implemented in the curriculum, it can be difficult for students to adopt an active work attitude due to their lack of experience with these types of educational activities. Some teachers are also worried about teams that do not function optimally.
When it comes to workload and work satisfaction, all teachers agree that the workload during the development and preparation of the TBL modules is higher compared to lecture-based courses. However, once the modules have been developed, teachers experience a similar or lower workload compared to traditional lecturing. If, due to large cohorts, several application sessions have to be held per module, this also requires more time investment from the teachers compared to lecturing for the entire student group. Teachers, however, appreciate the stimulating discussions between students, and the enthusiasm and creativity among students. For example, one teacher highlighted "The visible enthusiasm that the majority of students show during the tRATs and application sessions. Their enthusiasm always motivated me as a teacher. Sometimes the students actually come up with interesting viewpoints on complex problems". Similar findings have also been described in a previous study by Freeman (2012).

TBL from a curriculum perspective
One TBL-based course of 3EC is most likely too short to generate the full benefits of TBL. During one course, groups are less likely to develop into fully effective teams, and students may gain too little experience in active and collaborative learning to recognize the benefits. To further enhance the benefits of TBL as an active learning strategy for the students, a pilot was started in 2019 to implement TBL in several other courses within the Biology curriculum, including Statistics, Genomics & Big Data and Medical Embryology. In 2021, after the extended TBL-pilot had been successfully completed, it was decided to divide the first-year students at the start of the curriculum into teams that work together in all the TBL courses. These teams also follow a mandatory coaching programme that supports them in their team process.
The TBL facilities implemented for the Man and Environment course were also beneficial for other educational activities within the Faculty of Science of the Radboud University. Multiple teachers within various programmes of our faculty have discovered the benefits of the collaborative learning room for different kinds of collaborative learning. Furthermore, the peer evaluation process used in this course can also be applied in courses that do not fully implement TBL. Collaborative assignments are commonly included within our biology curriculum, and peer evaluation can ensure that students learn about and work on their collaboration skills. Our peer evaluation software has therefore been implemented more broadly than just in TBL courses and is also used in courses where other forms of collaboration take place.

Conclusions
We have shown that TBL can be successfully implemented in sustainability and environmental science education. The TBL format stimulates individual students to prepare well and to have productive group interactions on complex environmental topics, which we observed to pay off in terms of increased student performance. The application sessions in particular contribute to thinking in an active and functional way and seeing different perspectives, while providing a positive learning environment to learn to speak out in front of a group. In general, collaboration in teams is highly appreciated by the students. From a teachers' perspective, TBL is mostly perceived favourably as well, as it develops additional (transferable) skills beyond the natural science content, including working in teams, critical thinking and scientific argumentation. Teachers also observed that students take shared responsibility for their learning process and that the discussions help students to engage in the course. Switching to TBL does require significant initial time investment and leading group discussion (application) sessions remains demanding. Implementing TBL in one course of our BSc biology curriculum has proven to provide various positive spill-over effects. These include the development of other TBL-based courses that reinforce team functioning and transferable skills, as well as the uptake of TBL elements (and infrastructure) like readiness assurance testing and peer evaluation in other courses, where they are considered useful additions. With these insights we are convinced that consistently including TBL within our Biology curriculum can create a continuous path for collaborative learning to help solving complex biological and environmental challenges.