Collaboration in in-service teacher training - the missing link between empirical evidence and practice?

ABSTRACT

Cooperative learning is an evidence-based teaching strategy that has positive effects on students’ academic and social learning, but teachers use the strategy in class infrequently. Teacher collaboration has the potential to support teachers in the transformation of knowledge into the implementation of new teaching methods. In an experimental pre-test-post-test design (N = 53), an in-service teacher training with teacher collaboration for the implementation of cooperative learning in class was compared to a training without teacher collaboration. At both points of measurement, teachers’ knowledge about cooperative learning was tested and they reported on their self-efficacy beliefs for the implementation of cooperative learning and their use of cooperative learning in class, including the frequency, quality, and challenges of cooperative learning implementation. In both conditions, teachers’ knowledge and self-efficacy beliefs, and the quality of their use of cooperative learning increased strongly, whereas perceived challenges decreased. A significant interaction effect occurred only for the frequency of teachers’ use of cooperative learning in class. These results are discussed against the background of the efficacy of collaboration in in-service teacher training.

KEYWORDS:

• Teacher collaboration
• cooperative learning
• teacher education
• evidence-based practice
• beliefs

Introduction

The development of teacher competence is assumed to be a career-long learning process, as teachers are continuously faced with innovative tasks and challenges (European Commission 2013). This learning process includes teachers expanding and deepening their knowledge about effective teaching based on reflective practice and current research findings (European Commission 2013). However, teachers, researchers, and politicians perceive a gap between empirical results and teaching practices (cf. Hemsley-Brown and Sharp 2003; van Schaik et al. 2018). Reasons for this gap include from the teachers’ perspective a lack of credibility and relevance of research findings on teaching practices as well as barriers in understanding the results due to the complex use of scientific language (Vanderlinde and van Braak 2010). Therefore, the question arises how teacher education can stimulate and support teachers in integrating evidence-based teaching strategies into their practice. One component for supporting teachers’ learning is to offer not only opportunities for knowledge acquisition about evidence-based teaching strategies in a top-down approach but also to address teachers as experts and provide them the opportunity to take an active role to transform this knowledge and develop innovative practices (European Commission 2013; Vanderlinde and van Braak 2010).

Cooperative learning is a teaching strategy that has been well studied over the past decades regarding its effects on student outcomes (Ginsburg-Block, Rohrbeck, and Fantuzzo 2006; Hattie, 2009; Kyndt et al. 2013; Rohrbeck et al. 2003) and specifics for its implementation in class (for an overview, see Nokes-Malach, Richey, and Gadgil 2015). Positive research findings suggest to regularly integrate cooperative learning into teaching practice. Teachers also have positive beliefs about the effectiveness of cooperative learning (Abramczyk and Jurkowski 2020; Völlinger, Supanc, and Brunstein 2018). However, they also perceive several challenges in implementing the strategy and consequently use cooperative learning in class infrequently (Abramczyk and Jurkowski 2020; Buchs et al. 2017; Völlinger, Supanc, and Brunstein 2018). We developed and evaluated an in-service teacher training to support teachers’ use of cooperative learning. This teacher training is based on teacher collaboration as an approach to stimulate teachers’ transformation of theoretical knowledge into reflective practice. In the following sections, we provide an overview on theories and empirical findings on cooperative learning and teacher collaboration that form the framework for the research question addressed in this study.

Cooperative learning

Cooperative learning is based on the principles of positive interdependence between group members and individual accountability of each student (Johnson and Johnson 2009). With this learning approach, students can only achieve their goals when their group members also achieve their respective goals and students are accountable for their own contribution to the group goal. Meta-analyses reveal that cooperative learning, compared to teacher-centred, individualistic, or competitive learning situations, has positive effects on students’ academic achievements, self-concept, social-emotional skills, and peer relationships (Ginsburg-Block, Rohrbeck, and Fantuzzo 2006; Kyndt et al. 2013; Rohrbeck et al. 2003). Furthermore, cooperative learning has shown to enhance students’ experiences of relatedness with their peers and to reduce their bullying involvements, perceived stress, and emotional problems (Van Ryzin and Roseth 2018, 2019). However, the effects of cooperative learning on students’ academic achievements vary (Kyndt et al. 2013) and working together is sometimes not advantageous to some learners (Nokes-Malach, Richey, and Gadgil 2015). Research indicates that important moderators for the effectiveness of students’ cooperation include the problem-solving demands of the learning task (Kirschner et al. 2011), the distribution of information among group members (Deiglmayr and Schalk 2015), the structuring of the learning process (Supanc, Völlinger, and Brunstein 2017), the expertise of the group members (Buchs and Butera 2009), and students’ communication skills (Jurkowski and Hänze 2015). Therefore, teachers need to have knowledge about the principles of cooperative learning and the specifics for the implementation of cooperative learning in class (Abramczyk and Jurkowski 2020).

In line with the empirical results about the effects of cooperative learning on student outcomes, teachers report predominantly positive experiences with cooperative learning (Gillies and Boyle 2010). They also state mainly positive beliefs about the effectiveness of cooperative learning for students’ motivation, knowledge acquisition, and social-emotional development, and the cohesion in class (Abramczyk and Jurkowski 2020; Völlinger, Supanc, and Brunstein 2018). Despite these positive beliefs, teachers implement cooperative learning in class rarely (Abramczyk and Jurkowski 2020; Abrami, Poulsen, and Chambers 2004; Baines, Blatchford, and Kutnick 2003), at most sometimes (Buchs et al. 2017; Völlinger, Supanc, and Brunstein 2018). Accordingly, correlations between teachers’ beliefs about the effectiveness and the frequency of using cooperative learning are small (Abramczyk and Jurkowski 2020). Furthermore, the association between teachers’ knowledge about the principles and methods of cooperative learning and their use of cooperative learning in class is small and moderate respectively (Abramczyk and Jurkowski 2020).

The gap between empirical results and teachers’ beliefs about the effectiveness of cooperative learning compared to their teaching practice can be explained by the challenges and obstacles that teachers perceive. Several studies revealed that teachers find planning and conducting lessons with cooperative learning challenging and also integrating cooperative learning into the curriculum against the background of limited time resources and spatial conditions in class (Buchs et al. 2017; Gillies and Boyle 2010; Völlinger, Supanc, and Brunstein 2018). Furthermore, teachers report that they need to prepare students for cooperation and train their social-emotional skills (Gillies and Boyle 2010; Le, Janssen, and Wubbels 2018), consider group composition and the evaluation of individual and group results carefully (Gillies and Boyle 2010; Völlinger, Supanc, and Brunstein 2018), and find real group tasks that will motivate and engage every group member (Gillies and Boyle 2010). In sum, from the teachers’ perspectives, they must utilise much effort into the preparation of tasks, materials, and students’ working process. Thus, teachers’ perspectives on the challenges of cooperative learning are notably in line with research on moderators for student outcomes.

Teachers’ beliefs are assumed to be the gateway to implementing cooperative learning in class (Abramczyk and Jurkowski 2020). Thus, teacher education and training should fully account for teachers’ perspectives. That is, teachers should not only acquire knowledge about the theoretical background, methods, and models of effective cooperative learning, but the training should also address teachers’ concerns and provide the opportunity for practice and reflection (Baloche and Brody 2017). Collaboration between teachers might be a suitable setting for practice and reflection and, thus, might support teachers’ implementation of cooperative learning in class.

Teacher collaboration and professional development

Based on a systematic review, Vangrieken et al. (2015) defined teacher collaboration as joint interaction in activities that are necessary to fulfil job-related tasks. According to Little (1990), collaboration can take several forms that vary in the interdependence between partners, from low interdependence that includes the exchange of information and the communication of experiences, over mutual support and sharing of materials, methods, and ideas, to high interdependence that involves joint work with shared responsibility. Furthermore, a variety of collaborative activities can be initiated, including planning lessons and preparing materials together, sharing experiences about teaching, visiting other classrooms, discussing about instruction as well as working together to develop and try new ideas (Reeves, Hung Pun, and Sun Chung 2017).

Teacher collaboration can have positive effects at the student, teacher, and organisational levels (for an overview, see Vangrieken et al. 2015). For example, Reeves et al. (2017) found in a longitudinal study that the frequency of teacher collaboration predicted students’ academic achievements and teachers’ job satisfaction. Teacher collaboration can also enhance the quality of teaching (Gore et al. 2017) and support the implementation of new teaching methods (Meirink 2007). In this sense, Akiba and Liang (2016) described collaboration as teacher professional learning activity. In their study, they found that teacher-centred professional learning activities, including collaboration and informal communication with peers, were stronger predictors for student maths growth than teachers’ individual learning activities or their participation in professional development programmes. The influence of collaboration on teachers’ professional development is manifold (for an overview, see Reeves, Hung Pun, and Sun Chung 2017). For example, teachers in collaboration can learn instructional techniques from each other. They have the opportunity to reflect upon their teaching and receive feedback from their peers, which in turn can give teachers more confidence to try new instructional techniques. Teachers’ self-efficacy beliefs have shown to be essential for the implementation of cooperative learning in class (Abrami, Poulsen, and Chambers 2004). Specifically for teacher training in cooperative learning, teachers’ experiences with collaboration might initiate new ideas about how to handle the perceived challenges of cooperative learning and increase awareness among teachers about student benefits from cooperation.

Already 25 years ago, Shachar and Shmuelevitz (1997) developed a teacher training in cooperative learning based on teacher collaboration. At a single point of measurement, the frequency of teachers’ use of cooperative learning, their collaboration, and their self-efficacy beliefs were measured. Analyses revealed positive associations between these measures. In Miquel and Duran (2017), teachers were interviewed on their experiences with teacher training in cooperative learning. This training included teacher collaboration within and between schools. Teachers reported to feel support through collaboration because they had the opportunity to discuss the implementation of cooperative learning with their peers. These two studies hint on the benefits of teacher collaboration for the implementation of cooperative learning in class. However, to our knowledge, experimental evidence to date on the effects of collaboration-based teacher training on teachers’ use of cooperative learning is missing.

This study

Against the background of the gap between empirical results and teachers’ beliefs compared to their use of cooperative learning in class as well as the potential of teacher collaboration for professional learning and the implementation of teaching methods, we developed and tested a collaboration-based teacher training for cooperative learning. The collaboration-based teacher training was compared to a teacher training for cooperative learning without collaboration between participants. We assumed that the teacher training would result in an increase in teachers’ knowledge about cooperative learning (hypothesis 1). This hypothesis on the main effect controls for the basic effectiveness of the training in both conditions. Teachers’ self-efficacy beliefs and their use of cooperative learning in class should especially benefit from collaboration. Therefore, we expected that the teacher training would result in an increase in teachers’ self-efficacy beliefs for implementing cooperative learning (hypothesis 2a) and that this increase would be greater in the condition with teacher collaboration (hypothesis 2b). We also expected an increase in teachers’ use of cooperative learning in class (hypothesis 3a), with a stronger increase for teachers in the condition with the collaboration-based teacher training (hypothesis 3b).

Methods

All experimental procedures were in accordance with the 1964 Helsinki declaration and its later amendments. The study was approved by the ethics committee of our university.

Sample

A total of 53 teachers in the Language Arts (n = 22 English, n = 29 German, n = 2 English and German) participated in the study (n = 49 female, n = 4 male). Most of the teachers (90.6%) had from 11 to 30 years of experience in teaching, and 77.4% of the teachers were between 31 and 50 years old.

The study was conducted in Poland where teachers voluntarily take courses in teacher training. Teachers enrolled in a course about cooperative learning, which was provided by the teacher training institute in the city of Wrocław. The course was given by two trainers who are teachers and have an additional qualification as teacher trainers with 16 years of experience in teacher training at the time of this study. Teachers were informed that the course would be evaluated and were asked for their participation in this evaluation. All teachers gave their written informed consent.

Design

In a pre-test-post-test design, teachers were randomly assigned to either the experimental group (n = 26) or the control group (n = 27). The conditions were implemented in two separate courses taking place at different dates and given by the same two teacher trainers. With the assignment to the condition, teachers were also given the date for the course. In the experimental condition, teachers had a tandem partner for the planning and the reflection of cooperative learning in their class. Dyads were composed randomly. In the control condition, teachers implemented cooperative learning individually.

Pre and post to this intervention, teachers’ knowledge about cooperative learning was measured, and they reported on their self-efficacy beliefs for implementing cooperative learning and their use of cooperative learning in class. For the use of cooperative learning, we assessed the following variables: frequency of the use, quality of the implementation, and challenges for the implementation of cooperative learning in class.

As control variables, teachers reported on their beliefs about the effectiveness of cooperative learning at the pre-test, and at the post-test, we measured whether teachers had received support from colleagues for the implementation of cooperative learning and specifically in the experimental group whether teachers had used the opportunity to collaborate with their tandem partner.

Procedure

Each course consisted of four sessions and an implementation period of six weeks between the second and the fourth session. The first session of 2 hr took place online. Participants got to know each other and the teacher trainers and received information about the content and the organisation of the training. Furthermore, the session included an input and small group work on the basic elements of cooperative learning according to Johnson and Johnson (1999) and Slavin (1995). For the group work, teachers had access to a Padlet with various texts containing information and examples of the elements of cooperative learning, and groups took notes in a Google Docs document. Participants had access to this Padlet throughout the training, and materials for the following weeks were provided on this platform. The materials included a text that teachers should read in preparation of the second session.

The second session of 8 hr took place in presence one week later. The teacher trainers gave information on research results and various methods of cooperative learning, including think-pair-share, jigsaw classroom, and student teams achievement division (for an overview, see Topping et al. 2017). Participants worked in groups using these methods and had the opportunity to transfer their knowledge and experiences into their own lesson planning for specific learning topics. The teacher trainers reflected together with the participants on possible challenges and the demands on students and the learning content when using these methods in class. The information and experiences with cooperative learning methods formed the basis for the teachers to implement cooperative learning in class and to deepen their expertise.

During the following six weeks, teachers were encouraged to implement cooperative learning into their teaching. To support this transfer, teachers received a worksheet for lesson planning and reflection with cooperative learning. In addition, during the six weeks of implementation the third session of 2 hr took place online. In this session, the trainers provided input on tools for online cooperative learning. For example, these tools included Google Docs, ZUM-Pad, Classroomscreen, and Oncoo. Teacher trainers used these tools during the session, which allowed participants to gain experience with the tools for online cooperative learning.

After the six weeks, in the fourth session that took 8 hr and was held in presence, teachers reflected on their experiences with cooperative learning and exchanged ideas and materials on the implementation of cooperative learning. The reflection was based on the completed worksheets. Teachers had been notified on this use of the worksheets in the second session.

In the experimental condition, in the second session, tandem partners had the opportunity to get to know each other and organise their collaboration for the following weeks. For their collaboration, teachers were offered various options (phone, email, online platform) and were encouraged to use the online tools addressed in the third session. However, tandem partners self-organised their communication. Moreover, the worksheet for lesson planning in this group included prompts for the feedback from the tandem partner (see Supplement A).

Measures

Teachers completed a questionnaire online during the week before the first session and the week after the last session.

Knowledge about cooperative learning

Teachers answered 18 single-choice questions (right/wrong) on the implementation of positive interdependence, individual accountability, support for students’ communication and their social-emotional skills as well as reflection of group processes when using cooperative learning in class. Two example items are: ‘Positive interdependence can be implemented by teachers through dividing the task into substeps’ and ‘Individual accountability can be implemented by teachers through group members working on a collective outcome’. For each correct answer, teachers were given 1 credit point. The sum of the correct answers served as teachers’ knowledge about cooperative learning.

Self-efficacy beliefs

We used eight items from the 10-item scale in Völlinger et al. (2018), which is an adaptation of a scale to measure teachers’ general self-efficacy beliefs from Schwarzer and Schmitz (1999) to the specific issue of cooperative learning. Teachers reported on their beliefs on a 5-point scale (not true at all – exactly true). Two example items are: ‘I believe that I can guide my students during cooperative learning so that they achieve the learning goals’ and ‘Possible difficulties in students’ cooperation I face calmy because I can always trust my own abilities’. Internal consistency of the eight items was good at both points of measurement (Cronbach’s α_t1 = .88, Cronbach’s α_t2 = .85).

Use of cooperative learning

Teachers reported on the frequency of using cooperative learning in class on a single item with a 6-point scale (never – almost every lesson). This item is an adaptation of the question used in Abramczyk and Jurkowski (2020). We changed the verbal anchoring of the response alternatives to be more sensitive for the effects of the training.

The quality of teachers’ implementation of cooperative learning was measured with the 14-item scale in Völlinger et al. (2018). Teachers reported on the quality on a 5-point scale (not true – true). Sample items are: ‘During typical cooperative learning in my class, all students are engaged’ and ‘During typical cooperative learning in my class, I give students time to reflect upon their group work’. Internal consistency was excellent at the pre-test and good at the post-test (Cronbach’s α_t1 = .93, Cronbach’s α_t2 = .82).

For the measurement of the challenges that teachers perceive, we used 13 items from the 24 questions in Abramczyk and Jurkowski (2020) with a 5-point scale (not true at all – exactly true). Starting with the phrase ‘What is or might be from your perspective a challenge when using cooperative learning?’ two example items are: ‘Dealing with disciplinary problems’ and ‘Assessment of individual student achievements’. Internal consistency of the 13 items was good at both points of measurement (Cronbach’s α_t1 = .84, Cronbach’s α_t2 = .84).

Control variables

Teachers’ beliefs about the effectiveness of cooperative learning were measured at the pre-test with the 4-item scale in Völlinger et al. (2018). Teachers reported on a 5-point scale (not true at all – exactly true). Internal consistency was good (Cronbach’s α = .86; sample item: ‘Students have a great learning gain when they explain things to each other’).

At the post-test, teachers in the experimental condition were asked whether they had used the opportunity to collaborate with their tandem partner (yes/no). In addition, teachers in both conditions were asked whether they had support from their colleagues at school during the six weeks of implementation of cooperative learning (yes/no).

Data analyses

SPSS v. 28 was used for all analyses. First, we tested differences between the conditions at the pre-test for all relevant outcome variables as well as teachers’ beliefs about the effectiveness of cooperative learning using independent-samples t tests. Moreover, we described teachers’ use of the opportunity to collaborate with their tandem partner or with their colleagues respectively. To test the hypotheses, a series of repeated-measures ANOVAs were calculated with the point of measurement and the condition as independent variables and teachers’ knowledge and self-efficacy beliefs and the frequency, quality, and challenges of implementing cooperative learning as dependent variables. For all analyses, significance level was set at p = .05.

Results

Preliminary analyses

We found no difference between the two conditions of teachers attending the training with collaboration and teachers participating in the training without collaboration in their beliefs about the effectiveness of cooperative learning, t(51) = −0.42, p = .676.

At the pre-test, no differences were found between the two conditions in teachers’ knowledge about cooperative learning, t(51) = 0.29, p = .776, and teachers’ self-efficacy beliefs, t(51) = 0.06, p = .541. Furthermore, we found no differences between the two conditions in the frequency of the use of cooperative learning, t(51) = 1.71, p = .093, the quality of the implementation of cooperative learning, t(51) = 1.03, p = .307, and the challenges that teachers perceive, t(51) = 1.40, p = .168. Thus, in all outcome variables, teachers in the two conditions were comparable at the pre-test.

At the post-test, 12 teachers from the experimental group (46%) reported that they had used the opportunity to collaborate with their tandem partner. From these 12 teachers, six had additionally used the support from colleagues at their school. From the 14 teachers in the experimental group that had not used the opportunity to collaborate with their tandem partner (54%), six had used support from their colleagues. That is, eight teachers reported to have neither collaborated with their tandem partner nor with colleagues from their staff. In the control condition, three teachers reported that they had used support from their colleagues (11%). Teachers in the experimental condition reported more often receiving support from their colleagues than teachers in the control condition, χ²(1) = 8.02, p = .005.

Main analyses

Table 1 presents the descriptive results for the conditions and points of measurement. In line with hypothesis 1, teachers’ knowledge about cooperative learning increased significantly from pre-test to post-test, F(1, 51) = 5.83, p = .019, η² = .10. Neither the interaction effect, F(1, 51) = 0.81, p = .371, nor the main effect of the condition, F(1, 51) = 0.06, p = .811, was significant. The analysis revealed a significant main effect of time on teachers’ self-efficacy beliefs, F(1, 51) = 11.92, p = .001, η² = .19, which supports hypothesis 2a. However, against our expectation in hypothesis 2b, the interaction effect between condition and time was not significant, F(1, 51) = 0.07, p = .796. The main effect of the condition was also not significant, F(1, 51) = 0.33, p = .570.

Table 1. Means and Standard Deviations for the Conditions and Points of Measurement.

	Training with collaboration (experimental group)		Training without collaboration (control group)
	Pre-test M (SD)	Post-test M (SD)	Pre-test M (SD)	Post-test M (SD)
Knowledge	5.69 (1.41)	6.42 (1.21)	5.81 (1.69)	6.14 (1.26)
Self-efficacy	2.43 (.69)	2.79 (.64)	2.55 (.65)	2.86 (.67)
Frequency	1.04 (.77)	3.23 (1.31)	1.56 (1.34)	2.85 (1.38)
Quality	2.63 (.92)	3.28 (.43)	2.85 (.62)	3.26 (.46)
Challenges	2.22 (.82)	1.98 (.73)	2.48 (.45)	1.89 (.70)
Belief effective.	3.10 (.64)	-	3.01 (.85)	-

Note. Knowledge: 0–18 credits. Self-efficacy, challenges, belief effectiveness: 5-point scale, 4 = exactly true. Frequency: 6-point scale, 5 = almost every lesson. Quality: 5-point scale, 4 = true.

The analysis revealed a significant increase over time in the frequency of teachers’ use of cooperative learning, F(1, 51) = 60.60, p < .001, η² = .54, and a significant interaction effect between condition and time, F(1, 51) = 4.00, p = .051, η² = .07. Yet, the difference between the conditions at the post-test was not significant, t(51) = −1.03, p = .309. The main effect of the condition was also not significant, F(1, 51) = 0.08, p = .785. From pre-test to post-test, a significant increase in the quality of teachers’ implementation of cooperative learning occurred, F(1, 51) = 23.56, p < .001, η² = .32. Neither the interaction effect, F(1, 51) = 1.19, p = .280, nor the main effect of the condition, F(1, 51) = 0.57, p = .455, was significant. The same pattern of results show for the challenges that teachers perceive for the implementation of cooperative learning. We found a significant decrease over time, F(1, 51) = 12.64, p < .001, η² = .20, but neither the interaction effect, F(1, 51) = 2.11, p = .153, nor the main effect of the condition, F(1, 51) = 0.33, p = .571, was significant. In sum, the results support hypothesis 3a but only partially 3b for the use of cooperative learning.

As the preliminary analyses show, the teachers’ actual collaboration within the conditions that were designed to provide the opportunity was less frequent than expected, which might have led to the nonsignificant results for hypotheses 2b and 3b. Therefore, the main analyses were additionally run with a reduced sample of n = 18 teachers in the experimental condition who had collaborated and n = 24 teachers in the control condition who had not collaborated. These analyses revealed comparable results with slightly stronger main effects of time (see Supplement B).

Discussion

This study investigated the effects of a collaboration-based teacher training on closing the gap between the empirical results on the benefits of cooperative learning and teachers’ implementation of cooperative learning in class. An in-service teacher training with teacher collaboration for the implementation of cooperative learning was compared to a training without collaboration on teachers’ knowledge, self-efficacy beliefs, and use of cooperative learning. In sum, the analyses revealed a strong increase over time for these variables in both conditions. Teachers reported little benefit from collaboration, except for the frequency of cooperative learning. The reason for this strong increase might be that the basic training elements and conditions were very beneficial for teachers’ professional development. More precisely, teachers participated voluntarily and reported fairly high beliefs about the effectiveness of cooperative learning at the pre-test. Teacher trainers were also very experienced as both trainers and teachers and provided the opportunity for participants to share experiences and ideas about their teaching practice, which is already a form of teacher collaboration, even though it is a form with low interdependence (Little 1990). In sum, teachers were likely to have been very motivated in both conditions, creating on their own the opportunity for reflecting practice together with their peers so that teacher collaboration might not have been that critical for the implementation of cooperative learning.

Teachers with the collaboration-based training had a stronger increase in the frequency of their use of cooperative learning in class than teachers in the control condition. Despite a lower frequency at the pre-test and a higher frequency at the post-test for teachers with the collaboration-based training, the comparisons between the two conditions at both points were not significant, most likely because of the small sample size. Given that this interaction effect only occurred for the frequency of cooperative learning but not for the quality or the challenges of cooperative learning implementation, we can speculate on the efficacy of teacher collaboration. Teacher collaboration is assumed to provide teachers with the opportunity to learn instructional techniques from each other, receive feedback from peers, and reflect upon their teaching (Reeves, Hung Pun, and Sun Chung 2017). More specifically for cooperative learning, feedback and reflection might address teachers’ perceived challenges to carefully prepare tasks, materials, and students’ working process and to consider group composition and evaluation of students’ performances (Gillies and Boyle 2010; Le, Janssen, and Wubbels 2018; Völlinger, Supanc, and Brunstein 2018). These effects focus on cognitive processes for teachers’ professional development. If these processes had exerted their effects in the present study, the interaction effects for the quality and the challenges of cooperative learning implementation would also have been significant. Another possible effect of teacher collaboration might be that teachers have a higher motivation to transform their knowledge and newly learned skills into their teaching practice when they perceive that a peer expects them to apply these skills. This motivational aspect of collaboration has already been shown for the positive effects of students’ expectancies to instruct their peers on learners’ knowledge acquisition (Lachner et al. 2022). For the present study, teachers with the collaboration-based training might have felt more obliged to use cooperative learning in class resulting in a higher frequency of cooperative learning implementation.

These assumptions about the efficacy of teacher collaboration in in-service teacher training should be investigated in further studies that might reveal insights into mediating cognitive and motivational processes of teacher collaboration. Therefore, future studies should also measure teachers’ experiences with collaboration, their actual collaborative activities, and the issues on which they and their tandem partner collaborated. If teachers collaborate online, this could provide process data relatively efficiently. Of further interest could be to assess cooperative learning implementation through observational data or from the students’ perspective. This approach would provide more in depth data on the quality of teachers’ implementation of cooperative learning, the challenges in specific cooperative learning situations, teachers’ interactions with students and reactions to challenges, and insights as to whether students perceive a benefit from their teachers’ engagement in training.

In the experimental condition, teachers were assigned to a tandem partner and the dyads self-organised their collaboration during the implementation phase. This procedure is in line with teachers’ reports on the conditions for successful collaboration. More precisely, teachers state that they need to have the scope to decide when, where, and how often they meet and how they stay in contact and which precise issues they should collaborate on (Jurkowski, Ulrich, and Müller 2020). However, this flexibility has resulted in a black box about what happens in this important stage of transferring the knowledge and skills into teaching practice and possibly in a variation of teacher collaboration. As can be seen from teachers’ reports on their use of the opportunity to collaborate with their tandem partner or with their colleagues at school, some teachers neglected to collaborate (but might have benefitted motivationally from their expectancies that a peer could ask on their implementation of cooperative learning), and some teachers switched to or additionally engaged in collaboration with their colleagues. The latter behaviour points to the importance of self-organisation and to teachers’ need to exchange experiences and ideas with their peers. However, research about cooperative learning in schools and higher education points to the importance of structuring the collaborative working process (Slavin 1995; Supanc, Völlinger, and Brunstein 2017). Consequently, tandem partners might need both opportunities for self-organisation and guidance for their collaboration in a balanced ratio. Therefore, the benefits of teacher collaboration should be addressed in further research and in the development and evaluation of concepts for in-service teacher training.

Conclusion

Evidence-based teaching practices, such as cooperative learning, have a broad empirical basis, but their implementation into daily teaching practice poses a challenge to teacher education. Our results provide some evidence for the benefits of teacher collaboration as a motivational element of in-service teacher training, in addition to teachers’ beliefs and trainers’ expertise and the opportunity they provide for participants to reflect on their practice against the background of scientific knowledge. Studies on the efficacy of teacher collaboration can lead to deeper insights about the processes involved in teachers’ professional development and could contribute to the development of concepts of in-service training with a balanced amount of structured and self-organised collaboration.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/02607476.2024.2327062