Strengthening collaborative learning in secondary school: Development and evaluation of a lesson-integrated training approach for transactive communication

Abstract

Background: Transactive communication in collaborative learning environments occurs when students refer to their learning partners' ideas, build on them, and transform them into more elaborate ideas. This learning activity is essential for students to be able to benefit from collaboration, but they need scaffolding to produce transactive statements. From the teachers' perspective, this intervention should take little effort.
Aims: We developed and evaluated a lesson-integrated training in transactive communication for secondary school students, including instructions, exercises, and feedback on transactive communication, materials and tasks on curriculum-related content, and working with various cooperative learning methods.
Sample: In a quasi-experimental pre-test post-test design, 594 ninth-grade students in 23 classes received training in transactive communication or presentation skills. A complete set of data exists for a parallelized sample of 82 students in each condition.
Methods: At both points of measurement, students worked in dyads, and their communication was audiotaped, transcribed, and their transactive statements and non-transactive, content-related externalizations were coded. Furthermore, students completed knowledge tests about the topic of partner work and reported on their experiences with collaboration and motivation for group work (trait).
Results: Analyses revealed positive effects of the training in transactivity on transactive statements and experiences with collaboration. Students with the training in transactive communication also produced more exter-nalizations. However, no differences were found for students' knowledge acquisition and motivation for group work.
Conclusion: The training in transactive communication was effective for the collaborative working process, but transfer effects need to be further investigated.

Full text

Learning and Instruction 92 (2024) 101934
Available online 7 May 2024
0959-4752/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Strengthening collaborative learning in secondary school: Development and
evaluation of a lesson-integrated training approach for
transactive communication
Susanne Jurkowski
a
,
*
, Lukas Mundelsee
b
, Martin H¨
anze
c
a
University of Erfurt, Educational Science, Nordhaeuser Strasse 63, 99089 Erfurt, Germany
b
University of Heidelberg, Department of Educational Science, Akademiestraße 3, 69117 Heidelberg, Germany
c
University of Kassel, Department of Psychology, Hollaendische Strasse 36-38, 34109 Kassel, Germany
ARTICLE INFO
Keywords:
Transactive communication
Training
Collaborative learning
Dialogue coding
Knowledge co-construction
ABSTRACT
Background: Transactive communication in collaborative learning environments occurs when students refer to
their learning partners’ ideas, build on them, and transform them into more elaborate ideas. This learning ac-
tivity is essential for students to be able to benet from collaboration, but they need scaffolding to produce
transactive statements. From the teachers’ perspective, this intervention should take little effort.
Aims: We developed and evaluated a lesson-integrated training in transactive communication for secondary
school students, including instructions, exercises, and feedback on transactive communication, materials and
tasks on curriculum-related content, and working with various cooperative learning methods.
Sample: In a quasi-experimental pre-test post-test design, 594 ninth-grade students in 23 classes received training
in transactive communication or presentation skills. A complete set of data exists for a parallelized sample of 82
students in each condition.
Methods: At both points of measurement, students worked in dyads, and their communication was audiotaped,
transcribed, and their transactive statements and non-transactive, content-related externalizations were coded.
Furthermore, students completed knowledge tests about the topic of partner work and reported on their expe-
riences with collaboration and motivation for group work (trait).
Results: Analyses revealed positive effects of the training in transactivity on transactive statements and experi-
ences with collaboration. Students with the training in transactive communication also produced more exter-
nalizations. However, no differences were found for students’ knowledge acquisition and motivation for group
work.
Conclusion: The training in transactive communication was effective for the collaborative working process, but
transfer effects need to be further investigated.
1. Introduction
When students interact with their peers about learning content, they
have the opportunity to refer to their learning partners’ contributions,
elaborate, and build on their partners’ ideas, and jointly construct a
deeper or novel understanding of the material (Chi & Wylie, 2014). The
incorporation of the partners’ ideas into the construction of knowledge
represents the unique potential of collaboration (Chi & Wylie, 2014),
which is not only assumed to promote student learning in the classroom
(Webb, 2009) but is also seen as essential for the development of society
and people living together (Organisation for Economic Co-operation and
Development, 2019). Within this cognitive perspective on peer collabo-
ration, student learning is also supported by learners cueing each other’s
knowledge when they present and discuss their ideas (Nokes-Malach
et al., 2015) and by enhancing the working memory capacity in the group
(Kirschner et al., 2011).
Meta-analyses support the benet of peer collaboration for students’
knowledge acquisition compared to learners working individually or
being instructed by their teachers (Hattie, 2023; Rohrbeck et al., 2003).
The impact of cooperative learning has been investigated for a long time
with positive yet varying effects on student learning (Kyndt et al., 2013;
Roseth et al., 2008). Cooperative learning organizes peer interaction
* Corresponding author.
E-mail addresses: susanne.jurkowski@uni-erfurt.de (S. Jurkowski), mundelsee@ibw.uni-heidelberg.de (L. Mundelsee), haenze@uni-kassel.de (M. H¨
anze).
Contents lists available at ScienceDirect
Learning and Instruction
journal homepage: www.elsevier.com/locate/learninstruc
https://doi.org/10.1016/j.learninstruc.2024.101934
Received 18 May 2023; Received in revised form 15 March 2024; Accepted 23 April 2024

Learning and Instruction 92 (2024) 101934
2
following the principles of positive interdependence and individual
accountability, thereby stimulating students to share information,
co-construct knowledge, and assist in each other’s learning processes
(Johnson & Johnson, 1994). Cooperative learning includes individual
tasks and then sharing the outcomes of these tasks, whereas collaborative
learning focuses solely on learners’ joint elaboration activity (Tolmie
et al., 2010). However, the boundaries between cooperative learning and
collaborative learning are uid (Davidson & Major, 2014).
Variations in the effectiveness of peer collaboration can be explained
by characteristics of the instructional design, including the distribution
of information and materials between the learning partners (Deiglmayr
& Schalk, 2015), the structuring of the working process (Supanc et al.,
2017), and the difculty of the task (Hijzen et al., 2007). These studies
indicate that collaborative learning is more benecial for student
learning when learners share a common ground on core concepts, their
interaction is highly structured, and they face a challenging task.
In addition to these design characteristics, students’ communication
and their underlying social skills play a vital role in the effectiveness of
collaboration (Johnson & Johnson, 1994). Research has shown the
importance of communicative activities such as questioning (King,
1999), explaining (Webb, 1989) and transactive communication
(Jurkowski & H¨
anze, 2015). Transactive communication means that
learners refer to their partners’ contributions, for example, by asking
partners for further or more detailed information about their ideas,
adding additional thoughts to complete or rene the partners’ contri-
butions, or testing the partners’ ideas by questioning or opposing
alternative ideas (Berkowitz et al., 2008; Berkowitz & Gibbs, 1983).
Thus, transactive communication comprises learning activities that
represent the unique potential of peer collaboration according to Chi
and Wylie (2014).
Training learners in transactive communication has been shown to
support university students’ collaboration and individual knowledge
acquisition (Jurkowski & H¨
anze, 2015). Such a training addresses
teachers’ concerns that students lack the skills for collaboration
(Abramczyk & Jurkowski, 2020; Buchs et al., 2017) and researchers’
recommendation that students need to be prepared for collaborative
learning (Webb, 2009). Most importantly, students also report that a
lack of skills to effectively communicate and work together poses a large
obstacle (Le et al., 2018). Although training in transactive communi-
cation provides the necessary preparation for students for collaborative
learning, it is time-consuming, which teachers are also concerned about
when using collaborative learning in class. In particular, teachers nd
that implementing cooperative learning is difcult because of limited
time resources and curriculum requirements (Abramczyk & Jurkowski,
2020; Buchs et al., 2017). Consequently, despite teachers’ positive be-
liefs about the effectiveness of cooperative learning, they use this
teaching strategy in class rather infrequently (Abramczyk & Jurkowski,
2020; Buchs et al., 2017).
In the present study, we developed a training approach for trans-
active communication that integrates students’ learning of curriculum-
related content and evaluated this approach in secondary school stu-
dents against the background of a control training in presentation skills.
This study thus investigates the trainability of transactive communica-
tion in adolescent students and contributes to the use of collaborative
learning in class as an evidence-based teaching strategy. Within the
social and emotional learning framework, children’s and adolescents’
skills for collaboration represent a core aspect of their relationship
development (Durlak et al., 2022). Therefore, training in transactive
communication also contributes to students’ social and emotional
learning.
1.1. Transactive communication: the unique potential of collaborative
learning
In the 1980s, Berkowitz and Gibbs (1983) introduced the concept of
transactive communication when investigating the impact of peer
discussion about moral dilemmas on the development of students’ moral
reasoning. In their framework, they described several forms of trans-
active statements by which students refer to their learning partners’
ideas and verbally express that they cognitively elaborate their partners’
ideas and incorporate them into their thinking further. The authors
differentiate between statements that represent the partners’ ideas,
ensuring that the learning partners share a mutual understanding of
their ideas (representational transacts), and statements that operate on
the partners’ ideas and transform them into more elaborate ideas, sup-
porting learners’ integration of ideas (operational transacts).
Following the early work of Berkowitz and Gibbs (1983), several
authors, including Kruger and Tomasello (1986; see also Kruger, 1993),
Azmitia and Montgomery (1993), and Jurkowski and H¨
anze (2015)
have addressed transactive communication and developed and adapted
coding schemas for transactivity that are suitable for their respective
study’s learning environment. Berkowitz and Gibbs (1983), Kruger and
Tomasello (1986), and also Teasley (1997) included statements by
which the speaker refers to a previous contribution of their own. How-
ever, these so-called self-referential transacts do not rely on the presence
of a learning partner. Rather, students can engage in self-questioning
(Ebersbach et al., 2020; King, 1989) or self-explanations (Bisra et al.,
2018), which are constructive and support student learning but are not
interactive and unique for peer collaboration (Chi & Wylie, 2014).
Synthesizing the literature, seven forms of transactive statements
with reference to the learning partners’ contributions can be described:
(1) explaining in response to the partners’ questions (responding), (2)
formulating the partners’ ideas in one’s own words (paraphrasing), (3)
asking partners for further or more detailed explanations of their ideas
(questioning), (4) adding additional thoughts that complete or rene the
partners’ ideas (extending), (5) asking critical questions about the
partners’ ideas that stimulate partners’ reasoning (testing), (6)
comparing and contrasting the partners’ ideas with alternative ideas
(opposing), and (7) integrating different ideas in favor of an elaborate
and shared solution (integrating). Responding, paraphrasing, and
questioning serve to clarify ideas and are thus representational trans-
acts. Extending, testing, opposing, and integrating are operational
transacts because they transform the partners’ ideas.
The concept of referring to, elaborating, and building on the learning
partners’ contributions is included in other research without explicitly
framing this type of communication as transactive. For example,
Asterhan and Schwarz (2016) describe two types of interaction between
learners in argumentation: (1) consensual co-construction in which
students explain or expand ideas for the purpose of a better under-
standing and (2) argumentation in which learners engage in both critical
reasoning and joint knowledge construction resulting in learning part-
ners’ either accepting or refuting the thesis. Compared to the approach
of Berkowitz and Gibbs (1983), representational transacts seem similar
to consensual co-construction, and operational transacts overlap with
argumentation in Asterhan and Schwarz (2016). Another example is the
concept of collaborative argumentation in Noroozi et al. (2013).
Collaborative argumentation relates to students’ discourse about their
perspectives and describes that learners build arguments, support posi-
tions, consider and weigh arguments as well as counterarguments.
According to Larraín (2017), collaborative argumentation promotes
students’ knowledge acquisition because as they discuss contradictory
ideas, their implicit beliefs become explicit, students can examine their
own theories and the reasons behind them, and as learners negotiate
their divergence, they evaluate their ideas together. These benets all
rely on contradictory ideas between learners (Larraín, 2017). Therefore,
the learning situation must be designed to stimulate students’ devel-
opment of contradictory ideas. In the present study, we follow the
broader concept of transactive communication described above, which
includes argumentation (testing, opposing) but also goes beyond this
specic form of referring to the learning partners’ ideas.
Research has yielded positive relations between students’ transactive
communication and their learning outcomes for various learning
S. Jurkowski et al.

Learning and Instruction 92 (2024) 101934
3
objectives, tasks, and performance measures. The frequency of trans-
active statements during the discussion of moral dilemmas correlated
positively with students’ post-tested moral development (Berkowitz &
Gibbs, 1983; Kruger, 1993) and their skills for perspective coordination
(Mischo, 2005). The more students produced transactive statements
during peer collaboration, the better they performed individually on
knowledge tests about scientic problem solving (Azmitia & Mont-
gomery, 1993), biology (Russell, 2005), or educational psychology
(Jurkowski & H¨
anze, 2015). Students’ knowledge acquisition was not
only predicted by their own transactive communication but also by their
partners’ transactivity (Jurkowski & H¨
anze, 2015). Furthermore, stu-
dents’ transactive statements are also related to the group’s perfor-
mance, as the study of MacDonald et al. (2000) revealed for the
composition of a piece of music in learning dyads. These positive asso-
ciations between learners’ transactive communication and their per-
formances were also present in computer-supported peer learning on
reasoning in mathematics with students’ argumentation skills as the
outcome of interest (Vogel et al., 2016).
However, research shows that students produce transactive state-
ments far less frequently than non-transactive statements (Berkowitz &
Gibbs, 1983; Jurkowski & H¨
anze, 2015). For example, in Jurkowski and
H¨
anze (2015), higher education students produced approximately twice
as many externalizations (that is, articulation of knowledge and ideas
without reference to the partners’ ideas; Weinberger & Fischer, 2006)
than transactive statements. Although externalizations are typically the
starting point for the collaborative working process and thus central to
the social situation of peer collaboration (Weinberger & Fischer, 2006),
these empirical results indicate that students’ transactive communica-
tion is an important yet expandable aspect of their benet from
collaboration. This leads to the question of how teachers could provide
this support.
1.2. Teachers’ support for students’ transactive communication
One approach for teachers to promote students’ transactive
communication is the instructional design they prepare, including the
task and the structuring of students’ collaborative working process. This
kind of support is also known as scripting learners’ social and cognitive
activities during collaboration (Kollar et al., 2006; Vogel et al., 2017).
Research shows, for example, that when the task was formulated in such
a way that it stimulated students to question their learning partners’
ideas, oppose ideas, and search for differences that supported students in
developing critical thinking (Thiebach et al., 2016) and gaining deeper
knowledge about the learning topic (Wagner et al., 2018). Furthermore,
a study by V¨
ollinger et al. (2022) revealed that a high structure of the
working process characterized by the assignment of roles, task special-
ization, and reection prompts—all principles of cooperative learning
according to Johnson and Johnson (1994) or Slavin (1995)—leads stu-
dents to a greater level of knowledge use and more transactive state-
ments during the collaborative working process as well as a higher
performance in the knowledge posttest than a low structure of the
working process (cf., e.g., Moreno (2009) for hindering effects of a high
structure and V¨
ollinger et al. (2022) for a discussion of possible reasons
for varying results).
Despite the promising results of a well-prepared instructional design,
success depends on learners’ basic skills for transactive communication.
Training students in transactive communication is an approach for
teachers’ support that addresses students’ socio-communicative skills in
producing transactive statements (Jurkowski & H¨
anze, 2015). For
example, Deiglmayr and Spada (2011) found positive effects of training
for college students’ collaborative inference-drawing of pooled un-
shared knowledge during partner work in a murder-mystery case
following the training session. The authors dened collaborative
inference-drawing as learners combining different information they
possess uniquely, which is similar to the transactive form of integrating.
The training included instructions, practice, feedback, and prompts for
producing collaborative inference statements. A variation of the training
only with instructions and practice but without feedback and prompts
revealed no positive effects, indicating that feedback and prompts are
important elements of effective training. Mayweg-Paus et al. (2016)
developed a training in critical questioning (a specic form of trans-
active statements) for college students that includes instructions and
practice. The instructions focused on the benets of critical questioning,
which was supposed to foster students’ willingness to produce critical
questions during discussions. This training resulted in more frequent
critical questions during subsequent online partner work on a contro-
versial topic and a higher intrinsic motivation for the task and the
collaboration.
A training specically developed for transactive communication that
consisted of instructions, practice, and feedback on the different forms of
transactive statements revealed positive effects on college students’
transactive communication and knowledge gains in a partner jigsaw
conducted after the training session (Jurkowski & H¨
anze, 2015). The
positive effects on students’ knowledge acquisition were mediated by
the improved transactive communication, which indicates specic
training effects.
In sum, trainings for college students’ transactive communication
have already been developed and evaluated with positive effects.
However, whether these results can be transferred to the secondary
school context and whether students can also benet from being trained
in transactive communication is still an open research question.
Furthermore, note that previous studies used untreated (passive) control
groups. Thus, the training in transactive communication could have also
led to unspecic effects, including higher student motivation due to the
special attention.
1.3. Development of a lesson-integrated training approach
Training in transactive communication addresses teachers’ and stu-
dents’ reports that peer collaboration is sometimes ineffective because
of a lack of students’ skills (Abramczyk & Jurkowski, 2020; Buchs et al.,
2017; Le et al., 2018). In their study of students’ perspectives on the
challenges of peer collaboration and the strategies groups use to over-
come these challenges, Koivuniemi et al. (2018) found that if students
are unable to resolve challenges in their communication, this can create
further challenges in terms of emotional and motivational aspects of
peer collaboration. This may mean that students who perceive dif-
culties in communication due to a lack of skills experience collaboration
less positively and are less motivated to work together.
A training in transactive communication could support students’
collaborative working process and knowledge acquisition in collabora-
tive learning. In addition, the training could lead to students having
more positive experiences with collaboration and a higher motivation
for group work as they can cope better with collaborative learning.
However, such a training can be time-consuming, and it can contradict
teachers’ concerns about limited time resources and curriculum re-
quirements (Abramczyk & Jurkowski, 2020; Buchs et al., 2017). This
dilemma calls for a lesson-integrated training approach that combines
students’ learning of curriculum-related content with the training in
transactive communication. Furthermore, based on the research on
trainings for self-regulated learning that are integrated into students’
learning of curriculum-related content (cf. Donoghue & Hattie, 2021),
we assume that this integrated training approach supports the transfer of
students’ skills to further learning situations.
We developed a training in transactive communication for secondary
school students that integrates instructions, exercises, and feedback on
transactivity with the learning of curriculum-related content. The
training consists of eight units on transactive communication that take
45–90 min each. Every unit focuses on the instruction of a specic form
of transactive statement. During the exercises, students practice several
forms of transactive statements depending on the learning material and
the task. The nal unit is a review of the seven forms of transactive
S. Jurkowski et al.

Learning and Instruction 92 (2024) 101934
4
statements, and students practice all of them. The instructions include
explainer videos, example dialogues, and linguistic sentence stems that
can assist students in verbally producing transactive statements (G¨
atje &
Jurkowski, 2021). After every exercise, students reect together and
individually on their collaborative working process and give each other
peer feedback (for an overview of peer feedback, see Topping, 2009).
The reection also refers to the benets that students perceive in
collaboration and transactivity. After the reection and peer feedback,
student groups present their curriculum-related results from the
collaborative working process and share their experiences with trans-
active communication.
The exercises are based on students reading texts about curriculum-
related content and include tasks that students are to solve collabora-
tively. These tasks require students to deeply elaborate on the learning
material and transfer and integrate their knowledge to a further problem
(cf. Hijzen et al., 2007). During the exercises, students’ collaborative
working process is supported by the prompting of transactive state-
ments. Text reading and the collaborative working process are
embedded into cooperative learning methods, including jigsaw puzzle
or constructive controversy, focusing on students’ subsequent joint
elaboration. The training manual for teachers illustrates the kind of texts
and tasks that need to be integrated with the units of transactive
communication (Sawatzki et al., 2022a). Therefore, the training can be
transferred to a variety of curriculum-related content.
1.4. This study
The present study investigates whether the lesson-integrated training
in transactive communication strengthens secondary school students’
skills in producing transactive statements and supports them in
benetting from the collaboration. The training in transactive commu-
nication was compared to a control training in presentation skills to
consider unspecic training effects. Like the training in transactive
communication, the training in presentation skills refers to students’
communication but not directly to students’ collaborative working
process.
We assumed that compared to the training in presentation skills, the
training in transactive communication leads to a greater frequency of
students’ transactive statements during collaboration (hypothesis 1).
Based on the empirical ndings about positive correlations between
students’ transactive communication and their learning results (Azmitia
& Montgomery, 1993; Berkowitz & Gibbs, 1983; Mischo, 2005; Vogel
et al., 2016) and the positive effects of training in transactive commu-
nication on college learners’ knowledge acquisition (Jurkowski &
H¨
anze, 2015), students with the training in transactive communication
should have a greater knowledge acquisition in collaborative learning
than students with the training in presentation skills (hypothesis 2).
Furthermore, because of their newly learned skills for collaboration,
students can cope better with the learning situation. Therefore, we ex-
pected that students with the training in transactive communication
would report more positive experiences with collaboration (hypothesis
3) and a higher motivation for group work in general (hypothesis 4). In
addition, we explored whether the training in transactive communica-
tion has positive effects on students’ externalizations during collabora-
tion (research question 1).
2. Methods
All experimental procedures were in accordance with the ethical
standards of the institutional or national research committee and with
the 1964 Helsinki declaration and its later amendments or comparable
ethical standards. The study was approved by the German Federal
Ministry of Education and Research.
2.1. Sample and design
A total of N =594 ninth-grade students from 23 comprehensive
school classes in Germany participated in the study (300 females, 285
males, 9 other; M
age
=14.41, SD
age
=0.64). The classes were randomly
assigned to the experimental conditions. This resulted in 12 classes (n =
307 students) in the experimental group receiving a training in trans-
active communication and 11 classes (n =287 students) in the control
group receiving a training in presentation skills. In both conditions,
during the training, students worked together in various cooperative
learning methods with joint activity on the learning content sustainable
use of resources (see Fig. 1 for an overview of the design).
Pre and post to the training interventions, students worked in pairs
with unshared knowledge (partner jigsaw), and their transactive
communication and externalizations were measured via the coding of
transcribed audio recordings. Student pairs were randomly assigned for
the pretest and remained in their tandem for the posttest. Learning
topics were climatic, economic, and political inuences on smallholder
farming in Africa (pretest) and use of resources in vertical farms (post-
test). For each of the two learning situations or points of measurement,
students’ prior knowledge before the partner work and knowledge after
the partner work was tested, and students reported on their experiences
with collaboration. In addition, pre and post to the training in-
terventions, students responded to questions about their motivation for
group work.
The transcription and coding of audio material are very resource
intensive. In order to calculate the necessary sample size, we used power
analysis for nested ANCOVA modelling via GLIMMPSE (Kreidler et al.,
2013). Since a previous study with a 100-min training in transactive
communication with college students showed a medium effect size
(Jurkowski & H¨
anze, 2015) and our training for secondary school stu-
dents was considerably longer (2.5 days), our power analysis was based
on a medium effect size. The analysis revealed a necessary sample size of
168 students or 84 dyads respectively. Therefore, our data analysis is
based on a parallelized subsample of n =164 students with a complete
set of data with 82 students or 41 dyads respectively in each condition
(85 females, 76 males, 3 other; M
age
=14.42, SD
age
=0.66). The stu-
dents were distributed among the 23 school classes, with a range of two
up to 24 students per classroom. These students were selected with the
following procedure: For the pretest, audio recordings were transcribed
and coded until dyads were identied that matched pairwise in the
number of transactive statements across conditions. Only for the
selected students, the posttest audio recordings were transcribed and
coded. In this selected sample, students in the experimental and control
groups did not differ in the pretest measurements (see Supplement A).
2.2. Procedure
The trainings were held in blocks and took two and a half school
days. They were conducted by four professional trainers who were
randomly assigned to the classes and experimental conditions. To check
for treatment delity (cf. O’Donnell, 2008), the trainers participated in a
workshop on the training in transactive communication and the training
in presentation skills as well as on the learning content sustainable use of
resources and cooperative learning methods. Furthermore, they received
a manual for conducting the trainings, and each trainer completed a trial
run in a school class and received feedback from the research team. The
trial runs also served to revise the learning materials. During the de-
livery of the training in each of the 23 classes, one person from the
research team was present and lled in a protocol on the training
implementation and students’ engagement. In the protocol, the re-
searchers noted for each module whether the trainers had adhered to the
manual and whether there had been any classroom disruptions. No
events relevant to the treatment delity were noted. In addition, after
each training unit, students reported on difculty (5-point scale; very
easy – very difcult) and interest (5-point scale; boring – interesting)
S. Jurkowski et al.

Learning and Instruction 92 (2024) 101934
5
using one item each from H¨
anze and Berger (2007). Across the units,
difculty was appropriate (M =2.82) and students were rather inter-
ested (M =3.19). No differences were found between the conditions
(difculty: t (161) =0.62, p =0.54; interest: t (161) = − 1.13, p =0.26).
Partner work at the pretest and the posttest consisted of four steps. In
the rst step, the trainer introduced the learning topic at hand. For the
second step, the learning material was divided into two segments, and
each student of a learning tandem prepared one segment by reading a
text and completing a written task on that text. In the third step, learning
partners instructed each other on their particular learning segment.
Following the reciprocal instruction, student pairs worked together on
further tasks that required learners to integrate and apply both segments
of the learning topic. The measurement of transactive communication
and externalizations refers to this joint elaboration phase.
2.3. Measures
Scale statistics include the maximum of available data.
2.3.1. Transactive communication and externalizations
In line with Berkowitz and Gibbs (1983), we coded representational
transacts that the speakers produced to ensure a mutual understanding
of their learning partner’s ideas and operational transacts by which the
speakers transform the learning partner’s ideas into more elaborate
ideas. Furthermore, we coded externalizations (i.e., non-transactive
content-related statements) and residual statements. We developed a
codebook with descriptions and examples of these codes (see Supple-
ment B for an excerpt). Two research assistants trained in using the code
book and the content of the two learning situations with partner work
coded representational and operational transactive statements, exter-
nalizations, and residual statements (for further information about the
coding procedure, see Beltz.Jurkowski et al., 2022b).
A total of 30% of the transcripts were coded by both research as-
sistants. Interrater agreement was determined via Krippendorf’s
cu
alpha
for non-segmented, non-unitized units (Krippendorff et al., 2016). The
overall agreement was
cu
alpha =0.55 (representational transacts: 0.48;
operational transacts: 0.59; externalizations: 0.45; residual: 0.60). Data
analyses include the sum of coded representational transacts, opera-
tional transacts, and externalizations, respectively. Therefore, we used
the intraclass correlation coefcient as another indicator of coding
reliability. Absolute interrater agreement for representational transacts
was ICC =0.90, for operational transacts ICC =0.60, for externaliza-
tions ICC =0.74, and for residual ICC =0.95. Following the guidelines
in Cicchetti (1994), the reliability is good to excellent.
2.3.2. Knowledge acquisition
Following the reciprocal instruction of the partner jigsaw, students
answered an open-ended question about the two expert topics. This
question measured students’ basic knowledge about the expert topics
with which learners entered the joint elaboration phase of the jigsaw
puzzle. For this prior knowledge, students could attain up to three
credits. After the elaboration phase, students answered an open-ended
question requiring them to integrate their knowledge about the two
expert topics to solve a further problem. Here, students could attain up
to six credits. Questions and sample answers with credits are presented
in Supplement C.
A trained research assistant coded students’ answers using a
correction key developed by two experienced teachers. A second trained
rater coded the answers of 38% of the students (prior knowledge: ICC
t1
=0.64, ICC
t2
=0.65; knowledge: ICC
t1
=0.63, ICC
t2
=0.76).
2.3.3. Experiences with collaboration
Isoh¨
at¨
al¨
a et al. (2018) developed 11 items that refer to
socio-emotional processes during group collaboration, including group
cohesion, psychological safety, and group satisfaction. We adapted 10
items for collaboration with a partner (sample items are: “We listened to
each other”, “I felt safe to make mistakes”, and “I am satised with the
performance of our partner work”). Students answered these questions
on a 5-point scale (not true at all – exactly true). Cronbach’s
α
for this
scale was 0.89 in the pretest and 0.91 in the posttest.
2.3.4. Motivation for group work
We formulated two items with a 5-point scale (not true at all –
exactly true) that refer to how much a student likes collaborating with
Fig. 1. Study design with independent and dependent variables.
S. Jurkowski et al.

Learning and Instruction 92 (2024) 101934
6
peers. The two items are: “In general, I like partner work and group
work” and “In general, I am looking forward to working with a partner
or in a group”. The correlation of these items was r =0.78, p <0.001 in
the pretest and r =0.86, p <0.001 in the posttest.
2.4. Data analyses
All analyses were conducted using R (version 3.6.3) and the packages
“psych” (Revelle, 2020) and “lme4” (Bates et al., 2015). We ensured that
the entered data fullled their assumptions, that is, no collinearity or
overdispersion existed, and random effects were distributed approxi-
mately normally (Tabachnick & Fidell, 2013).
Given the multilevel structure of the data, we ran mixed models with
random slopes to account for variability across the levels. We checked
for the intraclass correlation (ICC) of the potential cluster variables in
our data, i.e., participants, dyads, and classes. Following the guidelines
in Hox (2010), we included levels with ICC >0.10. Therefore, in two
models (i.e., externalizations and motivation for group work), pretest
and posttest measures were nested in participants, which were nested in
dyads, whereas in all other models, pretest and posttest measures were
nested only in dyads.
All models were specied as mixed effect ANCOVA models with the
factor ‘time’ (coded as 0 =pretest, 1 =posttest) and its interaction with
the factor ‘condition’ (coded as 0 =control group, 1 =experimental
group) as independent variables and the posttest measurement as the
dependent variable. In the model on knowledge acquisition, prior
knowledge was entered as a covariate for predicting knowledge.
Consequently, for both measurement points (i.e., pretest and posttest),
knowledge was controlled for the respective prior knowledge (small-
holder farming or vertical farming). All variables were grand-mean
centered before being entered into the models. The ANCOVA strategy
was preferred over a gain score analysis, since it brings about a reduction
in the unaccounted-for variance, and, therefore, greater power for the
detection of treatment effects (Maxwell et al., 2018; van Breukelen,
2013).
Likelihood ratio tests revealed the best model t, which were a xed-
slope model for two outcomes (i.e., externalizations and motivation for
group work) and random-slope models, where the time slopes were
allowed to vary across the dyads, for all other outcomes (Bates et al.,
2015).
3. Results
Table 1 presents the descriptive statistics for the measurement points
and conditions and the correlations between the study variables (see
Supplement D for the correlations between study variables separately by
experimental condition). Note that transactive communication and
externalizations were not correlated with knowledge.
Figures for the interaction effects are presented in Supplement E. In
support of hypothesis 1, analyses revealed a signicant interaction effect
between time and condition in favor of the students with the training in
transactive communication for the frequency of representational trans-
active statements and operational transactive statements (see Table 2,
for the statistics). Furthermore, there was a signicant interaction effect
for the frequency of externalizations. Because of this effect and because
of the positive correlations between transactive communication and
externalizations, we tested the robustness of the training effects on
transactive communication by additionally entering externalizations as
a covariate in the ANCOVA models. In these extended models, the
interaction effect between time and condition for the frequency of
representational transactive statements and operational transactive
statements remained signicant (see Table 2). This indicates that the
effects of the training in transactive communication were specic for
transactivity.
In contrast to hypothesis 2, we found no signicant interaction effect
on students’ knowledge acquisition. In line with hypothesis 3, analyses
revealed a signicant interaction effect between time and condition in
favor of the experimental group for students’ reports on their experi-
ences with collaboration. In contrast to our assumption in hypothesis 4,
no signicant interaction effect was found for students’ reports on their
motivation for group work. These results indicate that the training in
transactive communication supported students in coping with the
collaborative learning situation directly but had no effect on the moti-
vational dispositions for collaboration in general.
4. Discussion
This study aimed to strengthen collaborative learning in secondary
school students through a lesson-integrated training approach for
transactive communication and to evaluate this training against the
background of a control training in presentation skills. We found posi-
tive effects of the training in transactive communication for the coding
of students’ transactivity. These training effects occurred for both
representational and operational transactive statements, indicating that
trained students more often produced statements that represented the
partner’s ideas and ensured that the learning partners shared a common
ground as well as statements that operated on the partner’s ideas and
transformed them into more elaborate ideas.
The descriptive statistics show that students produced representa-
tional transacts more often than operational transacts. These results are
in line with Berkowitz and Gibbs (1983; Berkowitz et al., 2008) who
assumed that operational transactive statements are more difcult for
learners to produce than representational transactive statements.
Following the ICAP framework and the idea of cognitive processes that
Table 1
Descriptive statistics and correlations between study variables.
Study variable 1 2 3 4 5 6 7
1 Representational transacts – 0.57
b
0.53
b
0.02 0.31
b
0.22
b
0.19
a
2 Operational transacts 0.47
b
– 0.40
b
0.02 0.17
a
0.10 0.17
a
3 Externalizations 0.44
b
0.31
b
– 0.07 0.31
b
0.09 0.13
4 Knowledge −0.02 0.09 0.06 – 0.15 −0.05 0.14
5 Experiences with collaboration 0.07 0.13 −0.02 0.23
b
– 0.45
b
0.23
b
6 Motivation for group work −0.01 0.01 0.04 0.13 0.27
b
– 0.15
7 Prior knowledge 0.06 0.03 0.11 0.34
b
0.08 0.02 –
Min, Max 0.00,
69.00
0.00, 20.70 0.00, 57.30 0.00,
5.00
1.80,
5.00
1.00,
5.00
0.00,
3.00
M (SD) of pretest in experimental group 20.20 (11.70) 3.55 (3.06) 13.80 (9.36) 2.40 (1.26) 4.30 (0.69) 3.73 (0.92) 1.79 (0.58)
M (SD) of pretest in control group 21.20 (13.70) 3.05 (2.83) 13.60 (8.63) 2.16 (1.24) 4.12 (0.64) 3.68 (0.76) 1.72 (0.57)
M (SD) of posttest in experimental group 33.60 (15.60) 5.93 (4.89) 24.50 (11.80) 2.08 (1.47) 4.24 (0.75) 4.05 (0.85) 1.10 (1.04)
M (SD) of posttest in control group 20.90 (13.30) 3.17 (3.23) 19.00 (11.50) 2.28 (1.33) 3.94 (0.75) 3.75 (1.14) 0.96 (0.94)
Note. Entries below the diagonal representing correlations of the pretest, entries above the diagonal representing correlations of the posttest.
a
p <.05.
b
p <.01.
S. Jurkowski et al.

Learning and Instruction 92 (2024) 101934
7
underlie overt learning activities (Chi & Wylie, 2014), we speculate that
more sophisticated cognitive processing of the partners’ ideas are
needed to add further and sometimes alternative or contradictory
thoughts than to reformulate the partners’ ideas or to question for
missing or additional information. Moreover, the learning material and
the tasks in this study might have provided more opportunities for stu-
dents to produce representational transactive statements than opera-
tional transactive statements. The characteristics of the instructional
design and their inuence on students’ transactive communication still
represent a gap in research. They should be addressed in further studies
because these investigations could result in deeper insights about types
of curriculum-related material and tasks that are suitable for the training
of specic forms of transactive statements.
Training in transactive communication also positively affected stu-
dents’ externalizations during the collaborative working process, which
might suggest unspecic training effects because of the special attention
students received in the experimental group and an increase in their
motivation. However, the effects of the training in transactive
communication remained stable when entering externalizations as a
covariate. Furthermore, we implemented an active control group with a
training in presentation skills. Given that both the control training in
presentation skills and the training in transactive communication refer
to students’ communication, we suppose that students had similar ex-
pectations on their improvement in learning skills, although we have not
checked for these expectations empirically. A match in expectations
between experimental and control groups is important for interpreting
intervention effects as specic (Boot et al., 2013). Therefore, we assume
that the training in transactivity had specic effects on transactive
communication as well as transfer effects on externalizations in the
collaborative working process.
In contrast to our expectations, the training in transactive commu-
nication lead to no greater knowledge acquisition. These missing effects
could be explained by the control training in presentation skills and its
possible impact on students’ cognitive processing of the material and
tasks. The training in presentation skills could have supported students
in structuring the material, which helped them organize their knowl-
edge. Hence, the training could have been a constructive learning ac-
tivity (Chi & Wylie, 2014). In this case, the active control group would
have made it difcult for the effects of the training in transactive
communication on students’ knowledge acquisition to occur. This
interpretation is supported by Jurkowski and H¨
anze (2015) who found
strong positive effects of the training in transactive communication on
college students’ knowledge acquisition compared to a passive control
group. However, this interpretation is contradicted by the design of the
knowledge tests in which learning partners should have integrated and
applied their ideas. The tests should have posed a challenging task that
requires students’ collaboration (cf. Hijzen et al., 2007) far beyond
structuring the material and organizing knowledge. In addition, stu-
dents’ transactive communication were not correlated with their
knowledge acquisition. Also, students’ externalizations were not asso-
ciated with their performances in the knowledge tests. These results
indicate that students’ cognitive load required to produce transactive
statements left too little capacity for storing and memorizing the
learning material and content-related ideas. The relatively low perfor-
mances in the knowledge tests (see Table 1) support this assumption.
Given the training in transactive communication was implemented
within 2.5 school days, further studies should analyze long-term effects
with more time spacing between the training units and distributed
practice (for the effects of distributed practice, see Cepeda et al., 2006).
The present results show that the lesson-integrated training in trans-
active communication is suitable for strengthening students’ transactive
communication. However, the effects of this training on learners’
knowledge acquisition might rely on a routine in transactive commu-
nication resulting from more and distributed practice.
The training in transactive communication positively affected stu-
dents’ experiences with collaboration, including a higher sense of
belonging with the learning partner, a more intense feeling of being safe
when contributing ideas or making mistakes, and a greater satisfaction
with the collaborative working process and its results. This effect cannot
be attributed to students’ collaborative learning practice because students
of both experimental conditions worked in various cooperative learning
methods. Le et al. (2018) found that students feel challenged by collab-
orative learning because they perceive a lack of skills to communicate and
work together. Unresolved challenges in communication can create
challenges in emotional and motivational aspects of peer collaboration
(Koivuniemi et al., 2018). Thus, we assume that the training in transactive
communication improved students’ coping with the learning situation,
resulting in more positive experiences with collaboration. For example,
students were instructed, practiced, and received feedback on asking
critical questions about their partner’s ideas and comparing and con-
trasting their partner’s contributions with alternative ideas. Hence, stu-
dents were encouraged to test and oppose ideas which is generally
difcult for them (Mayweg-Paus et al., 2016). This positive experience
might also be associated with group cohesion, psychological safety, and
Table 2
Effects of the training in transactive communication.
Predictors Fixed Part ICC R
2
m
β SE Individual/
Dyad
Model for representational transacts
b
–/0.97 0.15
Intercept −0.67
e
0.17
Time −0.18 0.19
Time ×Condition 0.42
e
0.11
Model for operational transacts
b
–/0.74 0.10
Intercept −0.50
e
0.13
Time −0.17 0.19
Time ×Condition 0.34
e
0.11
Model for externalizations
a
0.12/0.22 0.16
Intercept −1.07
e
0.14
Time −0.36
d
0.15
Time ×Condition 0.24
e
0.08
Extended model for representational
transacts
b
–/0.98 0.14
Intercept −0.83
e
0.19
Time −0.11 0.21
Time ×Condition 0.44
e
0.12
Externalizations −0.16
e
0.02
Extended model for operational
transacts
b
–/0.72 0.10
Intercept −0.44
e
0.14
Time −0.19 0.19
Time ×Condition 0.32
e
0.11
Externalizations 0.05 0.04
Model for knowledge acquisition
b
–/0.73 0.05
Intercept −0.19 0.21
Time 0.19 0.19
Time ×Condition −0.04 0.09
Prior Knowledge 0.25
e
0.06
Model for experiences with
collaboration
b
–/0.71 0.04
Intercept 0.24 0.14
Time −0.49 0.18
Time ×Condition 0.22
d
0.11
Model for motivation for group work
a
0.32/0.20 0.03
Intercept −0.36
d
0.14
Time −0.01 0.15
Time ×Condition 0.17
c
0.09
Note. β =standardized regression coefcients; ICC =Intra-class correlation;
R
2
m =marginal pseudo-R
2
.
a
Fixed-slope model.
b
Random-slope model.
c
p <.10.
d
p <.05.
e
p <.01.
S. Jurkowski et al.

Learning and Instruction 92 (2024) 101934
8
group satisfaction. This assumption is supported by the positive correla-
tions in the posttest between transactive statements and students’ reports
on their experiences with collaboration.
Despite the positive experiences with the training in transactive
communication, it had no effect on students’ motivation for group work,
which could also have been the result of an improved coping with the
learning situation. A possible reason for this is that students might need a
longer period of time with positive experiences of collaboration and self-
efcacy until their trait motivation is positively affected. This again calls
for further studies on the long-term effects of the training in transactive
communication. Although there is a wide variety of different challenges
in peer collaboration (cf. Abramczyk & Jurkowski, 2020; Koivuniemi
et al., 2018), future studies could also investigate students’ perceptions of
difculties in communication and how these develop over time.
Working constructively and productively with others is a core skill of
people in the 21st century (Organisation for Economic Co-operation and
Development, 2019) and is important for children’s social and academic
development (Durlak et al., 2022). In sum, the positive effects of the
training in transactive communication on students’ skills to produce
transactive statements and on their experiences with collaboration show
that the training in transactivity strengthened an important skill of stu-
dents and contributed to their social and emotional learning. In contrast
to programs that solely support children’s development of social and
emotional skills (Durlak et al., 2022), the training in transactive
communication follows a lesson-integrated training approach and com-
bines the training in transactive communication with students’ learning of
curriculum-related content. This approach addresses teachers’ concerns
about limited time resources and curriculum requirements (Abramczyk &
Jurkowski, 2020; Buchs et al., 2017) and has the potential to encourage
teachers to use collaborative learning more frequently as an
evidence-based teaching strategy that can further foster students’ social
and emotional skills and peer relationships (Ginsburg-Block et al., 2006).
Finally, some limitations and further directions for research need to
be mentioned. Professional trainers conducted the training in trans-
active communication and presentation skills, respectively. The train-
ings were held in two and a half school days as blocks. These
standardized conditions limit the generalizability of our ndings to
everyday school life. In future studies, based on the training manual
(Authors, 2022a), teachers could integrate the instructions and exercises
on transactive communication into their specic curriculum-related
content and train their students themselves over a longer period of
time. Investigating long-term effects on students and teachers could be
another aim of future research. Focusing on the teachers, an interesting
study would be to investigate whether they use collaborative learning
more often in class because of students’ strengthened collaboration
skills. Given the primary concern of teachers that students need to be
prepared for collaboration but also the concern about the lack of time
and pressure to conform to curriculum requirements, which are associ-
ated with their infrequent use of collaborative learning in class
(Abramczyk & Jurkowski, 2020; Buchs et al., 2017), we expect that the
training in transactive communication would result in teachers’ more
frequent use of this teaching strategy.
5. Conclusion
In sum, the training in transactive communication positively affected
students’ collaborative working process, as indicated by the more
frequent use of representational and operational transactive statements
and externalizations as well as students’ more positive experiences with
collaboration. Thus, the transactivity training resulted in specic and
more generalized effects for collaborative learning. Long-term effects
that rely on distributed practice need further research. The lesson-
integrated training in transactive communication is a promising
approach for supporting students’ social-emotional and academic
learning simultaneously.
CRediT authorship contribution statement
Susanne Jurkowski: Conceptualization, Formal analysis, Funding
acquisition, Investigation, Methodology, Project administration, Super-
vision, Validation, Writing – original draft. Lukas Mundelsee:
Conceptualization, Data curation, Formal analysis, Investigation,
Methodology, Validation, Visualization, Writing – review & editing.
Martin H¨
anze: Conceptualization, Formal analysis, Funding acquisi-
tion, Methodology, Validation, Writing – review & editing.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.learninstruc.2024.101934.
References
Abramczyk, A., & Jurkowski, S. (2020). Cooperative learning as an evidence-based
teaching strategy: What practitioners know, believe, and how they use it. Journal of
Education for Teaching, 46, 296–308. https://doi.org/10.1080/
02607476.2020.1733402
Asterhan, C. S. C., & Schwarz, B. B. (2016). Argumentation for learning: Well-trodden
paths and unexplored territories. Educational Psychologist, 51, 164–187. https://doi.
org/10.1080/00461520.2016.1155458
Sawatzki, D., Mundelsee, L., H¨
anze, M., & Jurkowski, S. (2022a). Partner- und
Gruppenarbeit lernwirksam gestalten: Ein Training der transaktiven Kommunikation
macht’s m¨
oglich [Making partner and group work effective for learning: Transactive
communication training makes it possible].
Beltz.Jurkowski, S., Mundelsee, L., Jüngst, C., & H¨
anze, M. (2022b). Messung
gemeinsamer Wissenskonstruktion: Ein Vergleich von hoch-inferenter Beobachtung,
niedrig-inferenter Codierung und Selbsteinsch¨
atzung der transaktiven
Kommunikation [Measuring knowledge co-construction: A comparison of high-
inference observation, low-inference coding and self-report of transactive
communication]. Zeitschrift für Erziehungswissenschaft. https://doi.org/10.1007
/s11618-022-01124-w.
Azmitia, M., & Montgomery, R. (1993). Friendship, transactive dialogues, and the
development of scientic reasoning. Social Development, 2, 202–221. https://doi.
org/10.1111/j.1467-9507.1993.tb00014.x
Bates, D., Maechler, M., Bolker, B., & Walker, S. (2015). Fitting linear mixed-effects
models using lme4. Journal of Statistical Software, 67, 1–48. https://doi.org/
10.18637/jss.v067.i01
Berkowitz, M. W., Althof, W., Turner, V. D., & Bloch, D. (2008). Discourse,
developments, and education. In F. Oser, & W. Veugelers (Eds.), Getting involved:
Global citizenship development and sources of moral values (pp. 189–201). https://doi.
org/10.1163/9789087906368_013. Sense.
Berkowitz, M. W., & Gibbs, J. C. (1983). Measuring the developmental features of moral
discussion. Merrill-Palmer Quarterly, 29, 399–410.
Bisra, K., Liu, Q., Nesbit, J. C., Salimi, F., & Winne, P. H. (2018). Inducing self-
explanation: A meta-analysis. Educational Psychology Review, 30, 703–725. https://
doi.org/10.1007/s10648-018-9434-x
Boot, W. R., Simons, D. J., Stothart, C., & Stutts, C. (2013). The pervasive problem with
placebos in psychology: Why active control groups are not sufcient to rule out
placebo effects. Perspectives on Psychological Science, 8, 445–454. https://doi.org/
10.1177/1745691613491271
Buchs, C., Filippou, D., Pulfrey, C., & Volp´
e, Y. (2017). Challenges for cooperative
learning implementation: Reports from elementary school teachers. Journal of
Education for Teaching, 43, 296–306. https://doi.org/10.1080/
02607476.2017.1321673
Chi, M. T. H., & Wylie, R. (2014). The ICAP framework: Linking cognitive engagement to
active learning outcomes. Educational Psychologist, 49, 219–243. https://doi.org/
10.1080/00461520.2014.965823
Cicchetti, D. V. (1994). Guidelines, criteria, and rules of thumb for evaluating normed
and standardized assessment instruments in psychology. Psychological Assessment, 6,
284–290. https://doi.org/10.1037/1040-3590.6.4.284
Davidson, N., & Major, C. H. (2014). Boundary crossings: Cooperative learning,
collaborative learning, and problem-based learning. Journal on Excellence in College
Teaching, 25, 7–55.
Deiglmayr, A., & Schalk, L. (2015). Weak versus strong knowledge interdependence: A
comparison of two rationales for distributing information among learners in
collaborative learning settings. Learning and Instruction, 40, 69–78. https://doi.org/
10.1016/j.learninstruc.2015.08.003
Deiglmayr, A., & Spada, H. (2011). Training for fostering knowledge co-construction
from collaborative inference-drawing. Learning and Instruction, 21, 441–451. https://
doi.org/10.1016/j.learninstruc.2010.06.004
Donoghue, G., & Hattie, J. (2021). A meta-analysis of ten learning techniques. Frontiers in
Education, 6. https://doi.org/10.3389/feduc.2021.581216
Durlak, J. A., Mahoney, J. L., & Boyle, A. E. (2022). What we know, and what we need to
nd out about universal, school-based social and emotional learning programs for
children and adolescents: A review of meta-analyses and directions for future
research. Psychological Bulletin, 148, 765–782. https://doi.org/10.1037/bul0000383
S. Jurkowski et al.

Learning and Instruction 92 (2024) 101934
9
Ebersbach, M., Feierabend, M., & Barzagar Nazari, K. (2020). Comparing the effects of
self-generating questions, testing, and restudying on students’ long-term recall in
university learning. Applied Cognitive Psychology, 34, 724–736. https://doi.org/
10.1002/acp.3639
G¨
atje, O., & Jurkowski, S. (2021). When students interlink ideas in peer learning:
Linguistic characteristics of transactivity in argumentative discourse. International
Journal of Educational Research Open, 2. https://doi.org/10.1016/j.
ijedro.2021.100065
Ginsburg-Block, M. D., Rohrbeck, C. A., & Fantuzzo, J. W. (2006). A meta-analytic review
of social, self-concept, and behavioral outcomes of peer-assisted learning. Journal of
Educational Psychology, 98, 732–749. https://doi.org/10.1037/0022-0663.98.4.732
H¨
anze, M., & Berger, R. (2007). Cooperative learning, motivational effects and student
characteristics: An experimental study comparing cooperative learning and direct
instruction in 12th grade physics classes. Learning and Instruction, 17, 29–41. https://
doi.org/10.1016/j.learninstruc.2006.11.004
Hattie, J. (2023). Visible learning: The sequel. A synthesis of over 2,100 meta-analyses
relating to achievement. Routledge.
Hijzen, D., Boekaerts, M., & Vedder, P. (2007). Exploring the links between students’
engagement in cooperative learning, their goal preferences and appraisals of
instructional conditions in the classroom. Learning and Instruction, 17, 673–687.
https://doi.org/10.1016/j.learninstruc.2007.09.020
Hox, J. J. (2010). Multilevel analysis: Techniques and applications (2 ed.). Routledge.
https://doi.org/10.4324/9780203852279
Isoh¨
at¨
al¨
a, J., N¨
aykki, P., J¨
arvel¨
a, S., & Baker, M. (2018). Striking a balance: Socio-
emotional processes during argumentation in collaborative learning interaction.
Learning, Culture and Social Interaction, 16, 1–19. https://doi.org/10.1016/j.
lcsi.2017.09.003
Johnson, D. W., & Johnson, R. T. (1994). Learning together and alone: Cooperative,
competitive, and individualistic learning. Allyn and Bacon.
Jurkowski, S., & H¨
anze, M. (2015). How to increase the benets of cooperation: Effects of
training in transactive communication on cooperative learning. British Journal of
Educational Psychology, 85, 357–371. https://doi.org/10.1111/bjep.12077
King, A. (1989). Verbal interaction and problem-solving within computer-assisted
cooperative learning groups. Journal of Educational Computing Research, 5, 1–15.
https://doi.org/10.2190/ynv2-qrb2-hucn-dgjk
King, A. (1999). Discourse patterns for mediating peer learning. In A. M. O´Donnell, &
A. King (Eds.), Cognitive perspectives on peer learning (pp. 87–115). Erlbaum. https://
doi.org/10.4324/9781410603715.
Kirschner, F., Paas, F., Kirschner, P. A., & Janssen, J. (2011). Differential effects of
problem-solving demands on individual and collaborative learning outcomes.
Learning and Instruction, 21, 587–599. https://doi.org/10.1016/j.
learninstruc.2011.01.001
Koivuniemi, M., J¨
arvenoja, H., & J¨
arvel¨
a, S. (2018). Teacher education students’
strategic activities in challenging collaborative learning situations. Learning, Culture
and Social Interaction, 19, 109–123. https://doi.org/10.1016/j.lcsi.2018.05.002
Kollar, I., Fischer, F., & Hesse, F. W. (2006). Collaboration scripts – a conceptual analysis.
Educational Psychology Review, 18, 159–185. https://doi.org/10.1007/s10648-006-
9007-2
Kreidler, S. M., Muller, K. E., Grunwald, G. K., Ringham, B. M., Coker-Dukowitz, Z. T.,
Sakhadeo, U. R., Baron, A. E., & Glueck, D. H. (2013). Glimmpse: Online power
Computation for linear models with and without a baseline covariate. Journal of
Statistical Software, 54(10), i10. https://doi.org/10.18637/jss.v054.i10
Kruger, A. C. (1993). Peer collaboration: Conict, cooperation or both? Social
Development, 2, 165–182. https://doi.org/10.1111/j.1467-9507.1993.tb00012.x
Kruger, A. C., & Tomasello, M. (1986). Transactive discussions with peers and adults.
Developmental Psychology, 22, 681–685. https://doi.org/10.1037/0012-
1649.22.5.681
Kyndt, E., Raes, E., Lismont, B., Timmers, F., Cascallar, E., & Dochy, F. (2013). A meta-
analysis of the effects of face-to-face cooperative learning. Do recent studies falsify or
verify earlier ndings? Educational Research Review, 10, 133–149. https://doi.org/
10.1016/j.edurev.2013.02.002
Larraín, A. (2017). Group-work discussions and content knowledge gains: Argumentative
inner speech as the missing link? Learning, Culture and Social Interaction, 14, 67–78.
https://doi.org/10.1016/j.lcsi.2017.04.002
Le, H., Janssen, J., & Wubbels, T. (2018). Collaborative learning practices: Teacher and
student perceived obstacles to effective student collaboration. Cambridge Journal of
Education, 48, 103–122. https://doi.org/10.1080/0305764X.2016.1259389
MacDonald, R., Miell, D., & Morgan, L. (2000). Social processes and creative
collaboration in children. European Journal of Psychology of Education, 15, 405–415.
https://doi.org/10.1007/bf03172984
Maxwell, S. E., Delaney, H. D., & Kelley, K. (2018). Designing experiments and analyzing
data. A model comparison perspective. Routledge.
Mayweg-Paus, E., Thiebach, M., & Jucks, R. (2016). Let me critically question this! –
Insights from a training study on the role of questioning on argumentative discourse.
International Journal of Educational Research, 79, 195–210. https://doi.org/10.1016/
j.ijer.2016.05.017
Mischo, C. (2005). Promoting perspective coordination by dilemma discussion: The
effectiveness of classroom group discussion on interpersonal negotiation strategies of
12-year-old students. Social Psychology of Education, 8, 41–63. https://doi.org/
10.1007/s11218-004-1884-y
Moreno, R. (2009). Constructing knowledge with an agent-based instructional program:
A comparison of cooperative and individual meaning making. Learning and
Instruction, 19, 433–444. https://doi.org/10.1016/j.learninstruc.2009.02.018
Nokes-Malach, T. J., Richey, J. E., & Gadgil, S. (2015). When is it better to learn
together? Insights from research on collaborative learning. Educational Psychology
Review, 27, 645–656. https://doi.org/10.1007/s10648-015-9312-8
Noroozi, O., Teasley, S. D., Biemans, H. J. A., Weinberger, A., & Mulder, M. (2013).
Facilitating learning in multidisciplinary groups with transactive CSCL scripts.
Computer Supported Learning, 8, 189–223. https://doi.org/10.1007/s11412-012-
9162-z
O’Donnell, C. L. (2008). Dening, conceptualizing, and measuring delity of
implementation and its relationship to outcomes in K–12 curriculum intervention
research. Review of Educational Research, 78, 33–84. https://doi.org/10.3102/
0034654307313793
Organisation for Economic Co-operation and Development. (2019). OECD future of
education and skills 2030. Retrieved from https://www.oecd.org/education/2030
-project/contact/OECD_Learning_Compass_2030_Concept_Note_Series.pdf. (Accessed
6 April 2023).
Revelle, W. (2020). psych: Procedures for personality and psychological research.
Retreived from. https://CRAN.R-project.org/package=psych. (Accessed 17 December
2021).
Rohrbeck, C. A., Ginsburg-Block, M. D., Fantuzzo, J. W., & Miller, T. R. (2003). Peer-
assisted learning interventions with elementary school students: A meta-analytic
review. Journal of Educational Psychology, 95, 240–257. https://doi.org/10.1037/
0022-0663.95.2.240
Roseth, C. J., Johnson, D. W., & Johnson, R. T. (2008). Promoting early adolescents’
achievement and peer relationships: The effects of cooperative, competitive, and
individualistic goal structures. Psychological Bulletin, 134, 223–246. https://doi.org/
10.1037/0033-2909.134.2.223
Russell, H. A. (2005). Transactive discourse during assessment conversations on science
learning. Retrieved from https://scholarworks.gsu.edu/cgi/viewcontent.cgi?article
=1040&context=epse_diss. (Accessed 6 April 2023).
Slavin, R. E. (1995). Cooperative learning (2nd ed.). Allyn & Bacon.
Supanc, M., V¨
ollinger, V. A., & Brunstein, J. C. (2017). High-structure versus low-
structure cooperative learning in introductory psychology classes for student
teachers: Effects on conceptual knowledge, self-perceived competence, and
subjective task value. Learning and Instruction, 50, 75–84. https://doi.org/10.1016/j.
learninstruc.2017.03.006
Tabachnick, B. G., & Fidell, L. S. (2013). Using multivariate statistics. Pearson Education.
Teasley, S. (1997). Talking about reasoning: How important is the peer in peer
collaboration? In L. B. Resnick, R. S¨
alj¨
o, C. Pontecorvo, & B. Burge (Eds.), Discourse,
tools and reasoning: Essays on situated cognition (pp. 361–384). Springer.
Thiebach, M., Mayweg-Paus, E., & Jucks, R. (2016). Better to agree or disagree? The role
of critical questioning and elaboration in argumentative discourse. Zeitschrift für
Padagogische Psychologie, 30, 133–149. https://doi.org/10.1024/1010-0652/
a000174
Tolmie, A. K., Topping, K. J., Christie, D., Donaldson, C., Howe, C., Jessiman, E.,
Livingston, K., & Thurston, A. (2010). Social effects of collaborative learning in
primary school. Learning and Instruction, 20, 177–191. https://doi.org/10.1016/j.
learninstruc.2009.01.005
Topping, K. J. (2009). Peer assessment. Theory Into Practice, 48, 20–27. https://doi.org/
10.1080/00405840802577569
van Breukelen, G. J. P. (2013). ANCOVA versus CHANGE from baseline in
nonrandomized studies: The difference. Multivariate Behavioral Research, 48,
895–922. https://doi.org/10.1080/00273171.2013.831743
Vogel, F., Kollar, I., Ufer, S., Reichersdorfer, E., Reiss, K., & Fischer, F. (2016).
Developing argumentation skills in mathematics through computer-supported
collaborative learning: The role of transactivity. Instructional Science, 44, 477–500.
https://doi.org/10.1007/s11251-016-9380-2
Vogel, F., Wecker, C., Kollar, I., & Fischer, F. (2017). Socio-cognitive scaffolding with
computer-supported collaboration scripts: A meta-analysis. Educational Psychology
Review, 29, 477–511. https://doi.org/10.1007/s10648-016-9361-7
V¨
ollinger, V. A., Supanc, M., & Brunstein, J. C. (2022). A video-based study of student
teachers’ participation and content processing in cooperative group work. Learning,
Culture and Social Interaction, 32. https://doi.org/10.1016/j.lcsi.2021.100598
Wagner, K., Bergner, M., Krause, U.-M., & Stark, R. (2018). F¨
orderung wissenschaftlichen
Denkens im Lehramtsstudium: Lernen aus eigenen und fremden Fehlern in multiplen
und uniformen Kontexten [Developing scientic thinking in teacher education:
Learning from own and advocatory errors in multiple and uniform contexts].
Zeitschrift für Padagogische Psychologie, 32, 5–22. https://doi.org/10.1024/1010-
0652/a000219
Webb, N. M. (1989). Peer interaction and learning in small groups. International Journal
of Educational Research, 13, 21–39. https://doi.org/10.1016/0883-0355(89)90014-1
Webb, N. M. (2009). The teacher’s role in promoting collaborative dialogue in the
classroom. British Journal of Educational Psychology, 79, 1–28. https://doi.org/
10.1348/000709908x380772
Weinberger, A., & Fischer, F. (2006). A framework to analyze argumentative knowledge
construction in computer-supported collaborative learning. Computers & Education,
46, 71–95. https://doi.org/10.1016/j.compedu.2005.04.003
Bio: Susanne Jurkowski is professor for inclusive education with a focus on social and
emotional development. Her research interests are collaborative learning, social-
emotional skills, teacher cooperation, and attitudes towards inclusion.
Bio: Lukas Mundelsee is a PhD student and school psychologist. His research interests are
collaborative learning, shyness, and teacher-student interaction.
Bio: Martin H¨
anze is professor for educational psychology. His research interests are
collaborative learning, generative learning, guided discovery, problem-solving, and
teacher education.
S. Jurkowski et al.