University of Southern California Dissertations and Theses

Bridging gaps, building futures: a meta-analysis of collaborative learning and achievement for Black and Latinx students

(USC Thesis Other)

Bridging gaps, building futures: a meta-analysis of collaborative learning and achievement for Black and Latinx students

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Bridging Gaps, Building Futures: A Meta-Analysis of Collaborative Learning and
Achievement for Black and Latinx Students
Sheree Cheng Mooney
Rossier School of Education
University of Southern California
A dissertation submitted to the faculty
in partial fulfillment of the requirements for the degree of
Doctor of Education
August 2024

© Copyright by Sheree Cheng Mooney 2024
All Rights Reserved

The Committee for Sheree Cheng Mooney certifies the approval of this Dissertation.
Akilah Lyons-Moore
Adam Kho, Committee Co-chair
Erika A. Patall, Committee Co-chair
Rossier School of Education
University of Southern California
2024

iv
Abstract
This research synthesis examines the impact of collaborative learning on academic achievement
among Black and Latinx students in the United States, addressing a notable gap in existing
literature. Grounded in John Hattie’s (2023) meta-analysis, which shows a significant positive
effect of collaborative learning on student achievement globally, this study focuses on whether
these benefits extend to historically marginalized groups in the United States. By limiting the
analysis to racially diverse samples, the research aims to provide nuanced insights into the
effectiveness of collaborative learning for Black and Latinx students, with additional exploration
of technology’s role as a moderating factor. The literature review explores collaborative
learning’s theoretical foundations, defining characteristics, and outcomes, emphasizing
constructivism and culturally responsive pedagogy. It highlights the complexity of defining
collaborative learning and underscores the importance of understanding its impact on Black and
Latinx students’ academic achievement. Building on Hattie’s work, this meta-analysis uses a
rigorous search strategy and systematic screening to synthesize research literature
comprehensively. The study employs specific inclusion criteria and a meticulous data extraction
process, coding various characteristics to ensure the meta-analysis’s validity and reliability.
Random-effect modeling and meta-regression are used to explore effect sizes and heterogeneity.
Seven studies meeting the inclusion criteria were incorporated into the final sample, yielding 10
samples and thirty-four effect sizes. The analysis revealed a statistically non-significant negative
effect size for achievement (g = –0.0267), with substantial heterogeneity among studies (I² =
74.73%), suggesting significant variation in effect sizes across different contexts. Moderator
analysis indicated no significant relationship between the percentage of Black or Latinx students
and the effectiveness of collaborative learning, nor did the presence of technology significantly

v
alter the impact of collaborative learning on academic achievement. These findings highlight the
variability in the effectiveness of collaborative learning and suggest the need for further research
to explore the impact of culturally responsive teaching practices and the nuanced role of
technology in enhancing the effect of collaborative learning on educational outcomes for diverse
student populations.

vi
Dedication
To my husband, Brandon Lee Mooney; The support you have given me over the 20 years we
have been together has brought me to this point. Throughout my doctoral journey, you have been
my rock, and together, we embarked on the incredible adventure of parenthood with our two
exceptional children. With your unwavering support, I discovered my passion for my career and
had the strength to follow it through. This work is as much yours as it is mine. Thank you for
being my partner in every sense.
To my children, Sebastian and Bernadette, you have given me the strength and power to
complete this degree for you. Despite the challenges, I wanted you to be a part of this journey.
Your presence has been my inspiration and motivation to persevere. This is for you.

vii
Acknowledgments
I would like to express my deepest gratitude to my dissertation co-chairs, Dr. Erika Patall
and Dr. Adam Kho. Your motivation, empathy, support, guidance, and expertise have inspired
me to embark on this meta”–analysis dissertation journey. I am truly fortunate to have had the
opportunity to work with such exceptional scholars.
I am deeply thankful to my family. To my husband, Brandon Lee Mooney, your
unwavering love and belief in me have been the bedrock of my success. To our children,
Sebastion and Bernadette, your presence has been my greatest source of inspiration and strength.
This journey would not have been possible without the three of you.
I am also grateful to my parents, who have been extremely supportive and helpful,
especially with having two little ones during this journey. Your assistance has been invaluable,
and the joys of grandparenting have been a blessing.
To my extended family: Dr. Chesser, for being my mentor in every way—educationally,
culinarily, and professionally; BJ, for always being there through every phone call and stressful
moment and for being my forever cheerleader; and Ken, for advocating for me and helping with
the kids—a significant contribution.
To my friends: Katarina, who always had the answers to my questions and pushed me to
be prompt on assignments; Daniela, whose virtual meetings provided accountability, dedicated
writing time, and the reassurance that we weren’t alone; and Tim, whose humor kept classes
interesting and motivated us with the countdown to graduation. I love that these people I met
have become more than just classmates—they have become friends.
Reflecting on this journey, I realize it often felt isolating, but the support of so many
loved ones made it possible. I appreciate each and every one of you.

viii
Table of Contents
Abstract.......................................................................................................................................... iv
Dedication...................................................................................................................................... vi
Acknowledgments......................................................................................................................... vii
List of Tables .................................................................................................................................. x
List of Figures................................................................................................................................ xi
Review of the Prior Literature ............................................................................................ 3
Defining Collaborative Learning ............................................................................ 3
Operationalization of Collaborative Learning ........................................................ 6
Theoretical Foundations for Collaborative Learning............................................ 10
The Role of Race and Ethnicity in the Relationship Between Collaborative
Learning and Achievement................................................................................... 12
Review of Empirical Research of Collaborative Learning and
Achievement ......................................................................................................... 15
Factors Contributing to Variation in the Relationships Between
Collaborative Learning and Achievement ............................................................ 21
The Present Synthesis........................................................................................... 26
Methods............................................................................................................................. 28
Literature Search................................................................................................... 29
Inclusion Criteria .................................................................................................. 30
Data Extraction ..................................................................................................... 32
Computing Effect Sizes and Data Analysis.......................................................... 34
Results............................................................................................................................... 35
Overall Average Effects/Correlations................................................................... 36
Publication Bias.................................................................................................... 37
Moderator: Percentage of Latinx students............................................................ 38

ix
Moderator: Percentage of Black Students............................................................. 39
Moderator: Technology ........................................................................................ 40
Discussion......................................................................................................................... 41
Summary of Key Findings.................................................................................... 41
Alignment of Key Findings With Theory and Prior Research ............................. 44
Implications for Theory ........................................................................................ 46
Implications for Practice ....................................................................................... 47
Limitations and Recommendations for Future Research...................................... 49
Conclusions........................................................................................................... 50
References..................................................................................................................................... 52
Tables............................................................................................................................................ 58
Figures........................................................................................................................................... 60
Appendix A: List of Studies Meeting Inclusion Criteria, Alphabetically by Author................... 62
Appendix B: Coding Guide Prior to Coding................................................................................. 64

x
List of Tables
Table 1: Overall Average Effect of Collaborative Learning 58
Table 2: Results of Moderator Analyses 59
Appendix A: List of Studies Meeting Inclusion Criteria, Alphabetically by Author 62
Appendix B: Coding Guide Prior to Coding 64

xi
List of Figures
Figure 1: PRISMA Chart 60
Figure 2: Publication Bias Funnel Plot 61

1
Bridging Gaps, Building Futures: A Meta-Analysis of Collaborative Learning and
Achievement for Black and Latinx Students
This research synthesis investigates the relationship between collaborative learning and
academic achievement among Black and Latinx students in the United States. Collaborative
learning approaches have garnered significant attention in educational research due to their
potential to enhance student engagement and improve learning outcomes. While studies highlight
the benefits of these approaches, a research gap exists on their impact on Black and Latinx
students in the United States.
John Hattie’s Visible Learning (2023) is an influential foundation for this synthesis,
globally recognized for its insights into instructional practices and student achievement. This
seminal work explores the effect of collaborative learning, among several hundred other
influences, on student achievement, contributing to the broader landscape of educational
research. Hattie’s (2023) meta-analysis encompasses over 350 factors influencing student
achievement, with an average effect size of 0.4. Notably, any effect size greater than 0.4
indicates a significant positive impact on student learning. Within the meta-analysis,
collaborative learning is represented by 955 studies from 12 meta-analyses, yielding a 0.45
weighted mean effect size, indicating that collaborative learning is an approach that educators
should consider applying to enhance student learning.
For educators interested in closing the opportunity gap among historically marginalized
groups in the United States, a limitation in Hattie’s approach is that his synthesis of collaborative
learning included samples from across the globe, which prevents a nuanced understanding of the
strength and direction of the influence on minoritized groups in the United States. This meta-

2
analysis aims to understand better if the effect size holds for Black and Latinx students. These
two minoritized groups have been historically marginalized in the national education system.
Understanding whether collaborative learning yields the same impact among Black and
Latinx students is highly relevant for American educators committed to reducing the opportunity
gap by adopting effective instructional practices for all subgroups. Hattie’s (2023) synthesis
suggests that collaborative learning enhances academic achievement by fostering a supportive
community where students deepen their understanding through mutual engagement and effective
communication. By limiting the studies on collaborative learning to those that contain racially
diverse samples in the United States, this synthesis will attempt to understand whether
collaborative learning is as effective for Black and Latinx students as it is for the aggregate
examined in Hattie’s work. In this way, this synthesis will provide a more granular
understanding of the relationship between collaborative learning and academic achievement.
To address these gaps, this study will explore the following research questions:
1. What is the effect of collaborative learning on academic achievement among samples,
including Black and Latinx students?
2. Does inclusion of technology explain variability in the effect of collaborative learning
on academic achievement among samples including Black and Latinx students?
3. To what extent does the effect of collaborative learning on academic achievement
vary depending on the percentage of the sample that is Black and/or Latinx?
By investigating these questions, this research aims to provide actionable insights for
educators and policymakers to enhance the academic outcomes of Black and Latinx students
through effective collaborative learning practices.

3
Review of the Prior Literature
This literature review provides an overview of collaborative learning, highlighting its
defining characteristics, theoretical foundations, and outcomes. The review explores the
theoretical frameworks of constructivism and culturally responsive pedagogy, demonstrating
how these perspectives inform an understanding of the effects of collaborative learning
approaches. Current research offers context and valuable insights into potential factors
influencing the relationship between collaborative learning and academic achievement for Black
and Latinx students.
Defining Collaborative Learning
Defining collaborative learning is a complex task with no universally accepted definition
(Dillenbourg, 1999; Koschmann, 1996; Whipple, 1987). According to Bruffee (1999),
collaborative learning “creates conditions in which students can negotiate the boundaries
between the knowledge communities they belong to and the one that the professor belongs to”
(p. 144). This perspective challenges traditional notions of power and authority, suggesting that
knowledge is not merely transmitted from professors or teachers to students but is socially
constructed within a community (Bruffee, 1984, 1999).
Expanding on Bruffee’s conception, Panitz (1999) defined collaboration as “a philosophy
of interaction and personal lifestyle where individuals are responsible for their actions, including
learning and respecting the abilities and contributions of their peers” (p. 3). Similarly, Oxford
(1997) acknowledged the philosophical underpinnings of collaborative learning. Focusing on the
learning processes, Roschelle and Teasley (1995) described collaboration as “the mutual
engagement of participants in a coordinated effort to solve the problem together” (p. 70).

4
Due to its philosophical foundation, collaborative learning typically avoids imposing
rigid structures on learning activities (Bruffee, 1995, 1999). Instead, students often work together
in small groups that are usually self-selected, self-managed, and loosely structured (Davidson,
2021a, p. 12). In other words, collaborative learning is an instructional approach where students
work together to tackle challenging problems, tasks, or projects, both collectively and
independently within the group (Mayer, 2011).
This method is born from the philosophy that knowledge is a social construct,
emphasizing student collaboration’s importance in enhancing learning experiences (Borokhovski
et al., 2012). Proponents of collaborative settings argue that collaborative learning is particularly
beneficial as it distributes cognitive processing demands among individuals, effectively
alleviating the cognitive load on learners and enhancing learning efficiency, especially in the
face of heightened task complexity (Kirschner et al., 2009; as cited in Schunk, 2020). This
approach fosters student discussions, assistance, knowledge evaluation, and compensation for
individual weaknesses, promoting a deeper understanding of subjects (Capar & Tarim, 2015).
Grounded in constructivism theories, collaborative learning facilitates knowledge coconstruction through social interaction. It encourages active student participation in group
activities, with teachers acting as facilitators to guide the process (Chen et al., 2018). The role of
social interaction in knowledge construction highlights scholars such as Bruffee (1993) and
Stahl, Koschmann, and Suthers (2006), who underscore its significance in collaborative learning
environments.
Hattie’s (2023) synthesis of research, echoing findings from Schroeder et al. (2007),
points to collaborative learning as a highly effective teaching strategy, with one of the highest
effect sizes (d = 0.67) observed in promoting student achievement. This underscores the

5
effectiveness of collaborative learning strategies in facilitating active engagement and interaction
among students and significantly contributing to academic achievement through mutual support
and shared knowledge construction. Collaborative learning represents a dynamic and interactive
instructional strategy prioritizing social collaboration and active participation for knowledge
construction.
According to Hattie, collaborative learning, including strategies like self-questioning,
peer tutoring, and critical thinking, is pivotal in students’ ability to consolidate deeper thinking
and become strategic learners. This approach supports the automation of learning specific
concepts or subtopics and integrates these into a broader repertoire of skills and strategies
essential for continuous learning and achievement. Studies indicate that collaborative learning
strategies consistently positively affect student achievement, particularly in enhancing
knowledge comprehension process skills and encouraging innovative problem-solving
approaches (Johnson et al., 2010; Kwon & Cifuentes, 2009; Wendt & Rockinson-Szapkiw,
2014).
Hattie’s (2023) findings also reveal a significant overlap between collaborative and
cooperative learning methods, which have been found to positively impact student achievement
to a lesser degree than collaborative learning but still above the hinge point Hattie suggests is
necessary to increase student learning. However, Borokhovski et al. (2012) differentiate between
the two, noting that while they share a common goal of fostering student interaction, they
originate from different theoretical foundations. Cooperative learning is characterized by
structured strategies that optimize student learning through specific techniques such as
consensus-reaching (Johnson and Johnson, 2008, as cited in Borokhovski et al., 2012). In
contrast, collaborative learning is less structured, focusing on group dynamics, task relevance,

6
and developing group skills and confidence. Scaffolding strategies, such as assigned roles and
metacognitive prompts, are crucial for enhancing collaborative learning, especially when
tackling complex problems.
Operationalization of Collaborative Learning
Over the past 5 decades, research on collaborative learning has proliferated, and
Dillenbourg et al. (1996) have categorized this research into three paradigms: the effect
paradigm, the conditions paradigm, and the interaction paradigm. Each paradigm stems from
different theoretical perspectives of collaborative learning. Building upon this taxonomy, a
fourth paradigm—the design paradigm—has emerged to describe design-based research in
Computer Supported Collaborative Learning (CSCL) over the last 20 years. Collectively, these
four paradigms encapsulate the research on collaborative learning and, in turn, the
operationalization of the constructs, which, as shown in the literature, is difficult to do.
The Effect Paradigm
The effect paradigm seeks to determine whether collaborative learning is more effective
than individual learning. Researchers typically conduct experiments with control groups
(working alone) and experimental groups (working collaboratively) in classroom or laboratory
settings to test their hypotheses. In this meta-analysis, studies investigating face-to-face
collaboration versus independent work (Johnson et al., 2010; Samaha & De Lisi, 2000; Serpell et
al., 2006) consistently found that face-to-face collaboration generally leads to better learning
outcomes, with students demonstrating higher levels of understanding, improved reasoning
skills, and better academic performance compared to those who work independently. Although
many studies report positive outcomes, there are also studies that have shown no significant
difference or even negative effects. For example, some research has found that low achievers

7
may become passive when collaborating with high achievers, leading to decreased engagement
and learning outcomes for these students. Additionally, there have been instances where group
dynamics and conflicts have hindered the overall effectiveness of collaborative learning.
Dillenbourg et al. (1996) argued that negative or null results should not be dismissed, as some
stable negative effects, such as low achievers becoming passive in collaboration with high
achievers, are well-documented (p. 8). Meta-analytic studies generally indicate a positive effect
of collaborative learning (Johnson et al., 2000; Slavin, 1980). Research indicates that
collaborative learning should not be seen as a mere group formation that inherently enhances
learning achievement (Slavin, 1983). The important question is what conditions make
collaborative learning more effective than individual work, which leads to the next paradigm.
The Conditions Paradigm
The conditions paradigm investigates the specific conditions facilitating effective
collaborative learning. Research methods are similar to those in the effect paradigm, but this
paradigm focuses on systematically exploring variables such as group formation, task types,
communication mediums, and collaborative contexts (Dillenbourg et al., 1996). For instance,
heterogeneous groups with varied expertise levels are typically more productive than
homogeneous groups. However, these groups can impact high and low achievers differently;
high achievers often take on leadership roles and benefit from teaching others, which can
reinforce their own understanding and skills. Conversely, low achievers may become passive or
overly reliant on their higher-achieving peers, potentially diminishing their active engagement
and learning opportunities.
Slavin’s (1983) meta-analysis on incentive and task structures showed that in K–12
settings, group rewards and individual accountability are critical for improving achievement. In

8
this meta-analysis, studies that fall under the conditions paradigm include Schacter (2000), who
examined the effect of the use of teacher support in collaborative learning, and Samaha (2000),
who investigated how same-gender and mixed-gender groups affected collaborative learning.
Several studies tested communication conditions in collaborative learning by creating
experimental designs that test the effect of online or traditional face-to-face, web-assisted, and
online in collaborative learning (Caldwell, 2006; Kwon & Cifuentes, 2007; Wendt & RockinsonSzapkiw, 2014).
Although the conditions paradigm offers insights into collaborative learning mechanisms,
real classroom environments involve interactions among numerous variables, leading to
inconsistent findings (Dillenbourg et al., 2009) and pointing to the need to look at moderators to
explain variability. Key conditions that have been theorized and found to be important for
effective collaborative learning include the composition of groups (heterogeneous versus
homogeneous), the types of tasks assigned, and the communication mediums used (online, faceto-face, web-assisted). Additionally, the broader collaborative context, including teacher support
and the structure of incentives and accountability within groups, significantly influences the
outcomes of collaborative learning. Effective collaborative learning hinges on productive group
interactions, prompting a shift towards the “interaction” paradigm.
The Interaction Paradigm
This paradigm explores a more nuanced conception of collaborative learning by looking
into what conditions trigger specific interactions and the effects of these interactions
(Dillenbourg et al., 1996). Research in this paradigm becomes more process-oriented, often
employing qualitative methods such as discourse analysis and conversation analysis to identify
moments of collaboration within groups (Stahl, 2006). Key interaction variables studied include

9
explanation, argumentation, negotiation, and regulation (Dillenbourg et al., 2009). For example,
Webb (1982) found that giving and receiving elaborate explanations correlated positively with
learning gains, whereas off-task behavior negatively impacted outcomes. Due to the constraints
of this meta-analysis, which only included experimental designs, no studies aligned with the
interaction paradigm were included.
The Design Paradigm
The design paradigm describes a strand of CSCL research focused on creating conditions
for effective group interactions (Dillenbourg et al., 2009). This paradigm is prevalent in CSCL
literature (Chen et al., 2018; Radkowitsch et al., 2020) and is characterized by design-based
research (DBR). DBR involves iterative collaboration between researchers and practitioners to
study educational phenomena in authentic contexts, refining design principles through iterative
design (Stahl & Hakkarainen, 2020).
Studies included in this meta-analysis predate the development of DBR. However, most
of the included studies fall under the umbrella of CSCL, a precursor to DBR. Researchers have
increasingly become interested in comparing the benefits of traditional, face-to-face
collaborative learning opportunities and those mediated through technology. For example,
Johnson et al. (2010), Schacter (2000), Wendt and Rockinson-Szapkiw (2014), and Caldwell et
al. (2006) compared the benefits of leveraging digital tools for collaborative learning. Some of
these tools include the HyLighter tool within the Social Annotation Model-Learning System
(SAM-LS) to facilitate student engagement through web-based social annotation (Johnson et al.,
2010) and the Edmodo platform on middle school students’ science misconceptions (Wendt and
Rockinson-Szapkiw, 2014).

10
Kwon and Cifuentes’ work (2007, 2009) adds nuance to the comparison between digital
and face-to-face collaborative environments. Their research investigates the differences between
collaborative work done in person with and without technological support. In particular, Kwon
and Cifuentes (2009) assess collaborative learning through computer-based concept mapping
activities conducted individually and in pairs among middle school students. Students use
computers to construct concept maps in pairs, enabling them to discuss, share ideas, and
negotiate meanings together. This approach is measured against individual efforts in terms of
comprehension test scores and concept map quality.
Future meta-analyses on collaborative learning will likely include studies within the
design paradigm. In this meta-analysis, we will use the inclusion of technology as a moderator.
Our study spans the effect, conditions, and design paradigms. It aligns with the effect paradigm
by investigating whether collaborative learning enhances academic achievement compared to
individual learning. It fits within the conditions paradigm by examining factors such as group
composition, task types, and communication mediums. By exploring how digital tools facilitate
effective group interactions, our study aligns with the design paradigm. This paradigm
organization helps explain the operationalization of collaborative learning by categorizing the
factors and approaches contributing to its effectiveness. This comprehensive approach deepens
the understanding of collaborative learning mechanisms and refines design principles for future
educational interventions.
Theoretical Foundations for Collaborative Learning
Constructivism, an educational philosophy, asserts that learners actively construct
knowledge through interaction with their environment and experiences, a process significantly
enhanced by collaborative learning. This philosophy is grounded in the belief that knowledge is

11
socially constructed, necessitating learners to engage in critical thinking and collaborative
activities to deepen their understanding. Collaborative learning, a key component of
constructivist approaches, involves students working in groups to solve problems or achieve
shared goals, enhancing learning outcomes.
The connection between constructivism and student achievement is evident through the
emphasis on active engagement and the social construction of knowledge. Research indicates
that constructivist methods, such as collaborative learning environments and problem-based
learning, effectively improve learning outcomes. These methods encourage learners to connect
new information with existing knowledge in a dynamic and engaging learning process facilitated
by teachers.
Jean Piaget’s radical constructivism underscores the importance of peer interaction in
learning. It suggests that children learn best when interacting, questioning, and exploring
concepts with their peers rather than in asymmetric adult-child interactions. This perspective
aligns with the principles of collaborative learning, where peer-directed interactions and social
negotiation play crucial roles in constructing knowledge (Tenenbaum et al., 2020; Caldwell,
2006; Miller & Miller, 1999).
Social constructivism, extending beyond a singular theory to a broader paradigm, views
reality as constructed through social interactions (Burr & Dick, 2017). In collaborative learning,
this approach posits that knowledge is co-constructed as students work together towards common
goals, with teachers acting as facilitators to guide the learning process (Chen et al., 2018). This
co-construction of knowledge in social settings is fundamental to learning, as supported by
Vygotsky’s (1986) and Wendt and Rockinson-Szapkiw’s (2014) emphasis on shared and cocreated meaning in social activities.

12
Kwon and Cifuentes (2009) explain that social constructivist learning theory emphasizes
the importance of culture and context in understanding what occurs in society and constructing
knowledge based on learners’ understandings (Derry, 1999). This perspective highlights the
intricate relationship between the learning environment and the learner, suggesting that cultural
and contextual factors deeply influence understanding and knowledge construction. Thus,
through its integration with collaborative learning, constructivism provides a robust framework
for enhancing student achievement. Constructivism offers a pathway to deeper understanding
and improved educational outcomes by fostering an environment where learners construct
knowledge through social interactions. This sets the stage for discussing how cultural and
contextual influences further shape the constructivist learning process, enhancing its relevance
and applicability across diverse educational settings.
The Role of Race and Ethnicity in the Relationship Between Collaborative Learning and
Achievement
In “Toward a Critical Race Theory of Education,” Ladson-Billings and Tate (2006)
argued that “critical race theory in education, like its antecedent in legal scholarship, is a radical
critique of both the status quo and the purported reforms (p. 25). This meta-analysis aims to
interrogate whether the influences included in Hattie’s syntheses, which drive policy and
instructional decisions around the globe, impact Black and Latinx students in the same way they
do other students. CRT would ask whether this is merely upholding the status quo, given that
many of these practices are steeped in the systematic racism that has led to the
disenfranchisement of students of color.
Understanding the structural and systemic nature of racism through CRT lays the
groundwork for implementing educational practices that recognize and actively counteract these

13
inequalities. One such practice is collaborative learning, which, grounded in constructivist
theory, leverages students’ collective knowledge and experiences, including those from diverse
racial backgrounds, to foster a more inclusive, engaging, and effective learning environment.
Importantly, unlike cooperative learning, often highly facilitated by an instructor who may
perpetuate structural inequities, collaborative learning largely removes the teacher or expert from
the environment. This allows students to co-construct new learning based on their own
experiences and funds of knowledge. By incorporating students’ cultural backgrounds into the
learning process, collaborative learning aligns with CRP, which aims to create educational
practices responsive to the cultural contexts of students of color.
All of the studies in this meta-analysis include 40% or more students of color,
specifically Black and Latinx students. Engaging in collaborative learning with peers who are not
from the majoritarian class might increase the potency of the collaborative learning model on
achievement, as Black and Latinx students will be less likely to be subject to stereotype threat
and other implicit biases shared by white classmates. Collaborative learning promotes
engagement and a sense of belonging, vital for an inclusive educational environment that honors
diverse experiences and cultural perspectives.
Carlone and Johnson (2012) demonstrate how incorporating Mexican students’ everyday
life and community practices into science education connects learning to their cultural
backgrounds, fostering equitable and relevant education. Similarly, Johnson (2010) points out
that mutual support and active participation in collaborative learning environments impart
academic knowledge and vital life skills, preparing students for a diverse world. This preparation
includes strengthening teamwork through conflict resolution, noting that learning emerges from
resolving conflicts rather than the conflicts themselves. To ensure the effectiveness of these

14
environments, educators must conduct regular assessments to identify and promptly address any
issues that arise.
Building on these ideas, culturally relevant pedagogy (CRP), as described by LadsonBillings (2020), is a model that “not only addresses student achievement but also helps students
to accept and affirm their cultural identity while developing critical perspectives that challenge
inequities that schools (and other institutions) perpetuate.” She argues that there are three
components of culturally relevant pedagogy: 
1. Student learning—prioritizing students’ intellectual growth, including their ability to
problem-solve.
2. Cultural competence—creating an environment where students affirm and appreciate
their culture of origin while also developing fluency in at least one other culture.
3. Critical consciousness—teaching students how to identify, analyze, and solve realworld problems, especially those that result in societal inequities against marginalized
groups.
While collaborative learning does not necessarily address cultural competence and
critical consciousness (unless the collaborative tasks are explicitly designed to do so), it
facilitates students’ ability to problem-solve by providing them opportunities to make sense of
learning tasks without a teacher’s input. When collaborative learning is intentionally designed to
incorporate students’ cultural experiences and perspectives, it can rise to the category of CRP. As
Brown and Crippen (2016) note, such an approach values and integrates students’ cultural
experiences into learning processes. This can be particularly powerful when integrated with
technology, as digital tools can support diverse ways of thinking and expression, making
learning more accessible and relevant to students of color.

15
Review of Empirical Research of Collaborative Learning and Achievement
Extensive research has explored the effect of collaborative learning on academic success,
with numerous syntheses and meta-analyses focusing on the collaborative learning teaching
method. These comprehensive reviews have generally identified collaborative learning as a
successful educational technique. Collaborative student work on academic tasks can enhance the
learning experience by motivating students to be more active and engaged, potentially surpassing
the effectiveness of traditional teaching methods (Astin, 1993; Loes, 2022).
Several decades of empirical research have demonstrated the positive relationship
between collaborative learning and student achievement (Slavin, 1990; Webb & Palinscar, 1996;
Barron, 2000; Johnson et al., 2007; as cited in Scager et al., 2016). Theorists suggest
collaborative learning boosts student achievement by fostering engagement and deep
understanding through interactive discussions and reasoning (Visschers-Pleijers et al., 2006).
Research also shows that collaborative learning enhances academic motivation across diverse
student backgrounds throughout education levels (Loes, 2022). For example, Samaha and De
Lisi (2000) found that urban minority students who engaged in peer collaboration on nonverbal
reasoning tasks significantly improved their problem-solving skills compared to those who
worked independently. Similarly, Johnson et al. (2010) demonstrated that college students using
a collaborative social annotation tool in team-based learning settings showed better reading
comprehension and meta-cognitive skills than those who did not engage in collaborative
learning.
However, not all studies find positive effects. For example, Kyndt et al. (2013) conducted
a meta-analysis of 37 studies and found that while collaborative learning can lead to higher
achievement, the effects are often modest, with an average effect size of 0.35. The study

16
highlighted significant variability across different contexts and implementations, noting that in
some cases, collaborative learning had little to no impact on student achievement. Similarly,
Kirschner et al. (2006) argue that collaborative learning can sometimes lead to cognitive
overload. Their analysis suggests that when students are not provided with adequate structure
and guidance, the demands of collaborative tasks can overwhelm their cognitive resources,
resulting in decreased learning outcomes. Rienties et al. (2014) conducted a study on culturally
diverse groups and found that poorly designed collaborative tasks can lead to reduced motivation
and engagement. Their results indicated that students who preferred working individually
experienced frustration and disengagement when participating in mandatory group activities,
which negatively impacted their academic performance.
The strength of the relationship between collaborative learning and achievement varies
depending on specific conditions facilitating effective collaborative learning. Wendt and
Rockinson-Szapkiw (2014), Kwon and Cifuentes (2007, 2009), and Caldwell (2006) compare the
effectiveness of communication mediums and collaborative contexts. They examine how various
collaborative formats—online vs. face-to-face and individual vs. group activities—affect student
achievement. This research sheds light on the diverse impacts of collaborative learning on
educational outcomes, emphasizing enhancements in student learning and engagement. For
example, Wendt and Rockinson-Szapkiw (2014) found that students engaged in online
collaborative activities using platforms like Edmodo showed a greater reduction in science
misconceptions compared to those participating in traditional face-to-face collaborative learning.
Similarly, Kwon and Cifuentes (2007, 2009) explore the comparative effectiveness of individual
and collaborative concept mapping on science learning. They found that collaborative concept
mapping significantly improved comprehension and retention of scientific concepts compared to

17
individual concept mapping. These studies provide valuable insights into how different
collaborative learning modalities can enhance educational outcomes.
Meanwhile, Caldwell (2006) broadens the scope by examining various collaborative
contexts, including traditional classroom instruction and web-assisted learning, to determine their
impact on students’ performance and motivation in a computer programming course. Control
interventions in these studies often use traditional or less interactive learning methods as a
baseline to evaluate collaborative interventions. By comparing various collaborative contexts, the
research reveals the intricate relationship between collaborative learning and educational
outcomes, enhancing understanding of optimizing collaborative approaches for improved
achievement. This underscores the importance of selecting suitable modalities tailored to the
educational context and learning objectives.
The incorporation of technology-enhanced collaborative learning has been shown to
increase the strength of the positive relationship between collaborative learning and achievement
(Borokhovski et al., 2015; Chen et al., 2018; Johnson et al., 2010; Saqr et al., 2022; Schacter,
2000). These investigations delve into how digital tools, such as social annotation platforms,
computer-supported collaborative learning systems, and other technology-based interventions,
can facilitate collaborative learning experiences that are rich, interactive, and engaging. For
instance, Johnson et al. (2010) examine the use of social annotation tools in college English
courses, comparing their effect on students’ reading comprehension and critical thinking skills
against traditional reading strategies. Similarly, Borokhovski et al. (2015) conducted a metaanalysis to assess the impact of intentionally designed collaborative activities integrated into
course designs using technology versus incidental or contextual interactions that do not
specifically leverage technology for collaboration. These studies span multiple educational levels

18
and settings, comparing technology-enhanced collaborative learning with traditional teaching
methods. They aim to assess how technology affects learning outcomes, knowledge gain, and
student engagement. Detailed comparisons reveal the significant potential of incorporating
technology into collaborative learning, suggesting it can improve educational achievements and
foster deeper student engagement.
The work of Kwon and Cifuentes (2007; 2009) delves into the nuanced impact of
technological tools, like computer-based concept mapping, on collaborative learning. Their
findings suggest that while technology can aid in the organization and sharing of ideas, its
effectiveness hinges on its integration into collaborative activities. Proper technology integration
fosters dynamic interaction and deeper engagement, enabling learners to construct and negotiate
meaning collectively, achieving higher cognitive outcomes, and promoting a supportive learning
environment. For example, Caldwell (2006) found that using clickers in the classroom facilitated
increased student interaction and engagement, which led to an improved understanding of course
material. Similarly, Wendt and Rockinson-Szapkiw (2014) demonstrated that middle school
students engaged in online collaborative learning activities exhibited a deeper understanding of
scientific concepts and reduced misconceptions compared to those who participated in traditional
face-to-face collaborative activities. Multiple interconnected factors shape the impact of
collaborative learning on student achievement. At the forefront, Borokhovski et al. (2015) and
Chen et al. (2018) emphasize the importance of instructional design and intentionality in crafting
collaborative learning experiences. They advocate for intentionally designed collaborative
activities and strategic technology use, stressing the importance of clear goals and supportive
tech to enable meaningful learner interactions. This method aligns tasks with learning outcomes,
boosting engagement and understanding, thereby improving achievement.

19
Collaborative learning is especially effective for minoritized students. For example,
research by Serpell et al. (2006) and Samaha and De Lisi (2000) examines collaborative learning
in unique settings like culturally diverse and low-income urban schools. These studies aim to
show how tailored collaborative approaches can improve academic outcomes, problem-solving,
and knowledge transfer across various student groups, demonstrating collaborative learning’s
adaptability to diverse educational needs. Serpell et al. (2006) found the effectiveness of
communal learning environments while using different tools (e.g., computer simulations vs.
physical apparatus) among African American and White fourth graders. This study emphasizes
the cultural dimensions of collaborative learning, exploring how communal learning contexts can
leverage culturally familiar group dynamics to improve learning outcomes.
Similarly, Samaha and De Lisi (2000) focus on collaborative learning among seventh
grade minority students, examining the impact of collaborative activities on nonverbal reasoning
tasks. This research underscores the power of collaborative learning to address educational
challenges, especially in urban environments. It demonstrates how activities designed to meet
students’ unique needs and backgrounds outperform traditional, less personalized methods. The
studies show significant improvements in educational outcomes by aligning collaborative
learning with learners’ cultural and environmental contexts. This emphasizes the importance of
adaptable and effective collaborative strategies responsive to learners’ backgrounds.
Another critical aspect is the impact of group dynamics and composition on collaborative
learning and achievement. Studies by Samaha and De Lisi (2000) and Wendt and RockinsonSzapkiw (2014) demonstrate that the makeup of collaborative groups and their interactions can
significantly influence learning outcomes. For instance, Samaha and De Lisi (2000) found that
gender composition affects group performance, with mixed-gender groups often outperforming

20
single-gender groups in nonverbal reasoning tasks due to diverse perspectives and problemsolving approaches. Additionally, Wendt and Rockinson-Szapkiw (2014) showed that group
members’ ability to engage in meaningful discourse, characterized by active listening and
constructive feedback, leads to higher academic achievement. This discourse promotes critical
thinking and a deeper understanding of the material, which are vital for improved learning
outcomes. Empirical evidence supports that effective group dynamics encourage the exchange of
ideas, critical thinking, and deeper understanding. For example, Wendt and Rockinson-Szapkiw
(2014) observed that groups with high levels of meaningful discourse and mutual respect had
significantly better learning outcomes compared to groups with poor interaction quality. These
findings underscore the importance of fostering a collaborative environment where all group
members feel valued and heard.
Lastly, cultural and contextual congruence are critical to the effectiveness of
collaborative learning. Research by Serpell et al. (2006) and Caldwell (2006) indicates that
learning environments tailored to students’ cultural backgrounds and preferred instructional
modalities, whether online or face-to-face, can significantly boost engagement and achievement.
Serpell et al. (2006) found that African American students in communal learning environments,
which mirrored their cultural norms of cooperation and community, showed higher engagement
and academic performance. Similarly, Caldwell (2006) demonstrated that students’ achievement
increased when instructional methods aligned with their cultural contexts, whether through
culturally relevant examples or the use of familiar technology. These insights highlight the
importance of aligning learning environments with students’ cultural and contextual needs to
optimize collaborative learning outcomes. Therefore, considering race and ethnicity as a

21
moderator when examining the relationship between collaborative learning and achievement is
crucial.
This research highlights the interplay between cultural backgrounds, collaborative
learning, and academic achievement. While existing studies reveal the impact of cultural
diversity on collaborative learning and its connection to achievement, empirical evidence
specifically exploring how race and ethnicity influence the relationship between collaborative
learning and academic success still needs to be developed. To address this gap, the forthcoming
meta-analysis aims to investigate the effectiveness of collaborative learning on academic
achievement across diverse student samples, including Black and Latinx students. Through this
analysis, we seek to deepen our understanding of how collaborative learning interventions can
support the educational attainment of historically marginalized groups and contribute to more
equitable outcomes in education.
Factors Contributing to Variation in the Relationships Between Collaborative Learning
and Achievement
Many studies provide valuable insights into the relationship between student interactions
and instructional methods in collaborative learning, emphasizing the significance of designed
interaction treatments and explicit instructional design. These investigations further explore
variables that might account for the heterogeneity in outcomes across different studies. Key
factors include the diversity within the student sample (such as the proportion of Black and
Latinx students), the implementation strategies of collaborative learning, and the characteristics
of the comparison group. The following sections will delve into these aspects in detail,
examining how the characteristics of the sample and the specific methods of collaborative
learning influence academic achievement.

22
Characteristics of Sample
As reviewed previously by Serpell et al. (2006) and Samaha and De Lisi (2000), the
characteristics of the sample, particularly racial diversity and socioeconomic background, play
significant roles in shaping the effectiveness of collaborative learning on academic achievement.
Serpell et al. (2006) examined the impact of collaborative learning among African American and
White fourth graders, finding that collaborative learning methods improved academic outcomes
more significantly for African American students compared to their White peers. They attributed
these differences to the benefits of CRP, particularly communal learning contexts that align with
the cultural ethos of African American students. This highlights the importance of culturally
relevant pedagogies in addressing racial disparities in education by creating optimal learning
environments that resonate with students’ cultural backgrounds.
CRP, as described by Ladson-Billings (2020), emphasizes student learning, cultural
competence, and critical consciousness. In the study by Serpell et al. (2006), the communal
learning contexts directly align with these CRP components by fostering environments where
students’ cultural backgrounds are acknowledged and respected. This approach not only
enhances student learning but also promotes cultural competence by allowing students to draw
upon their cultural experiences, thereby creating a more inclusive and supportive educational
environment.
Johnson et al. (2010) assessed diverse student populations to determine the impact of
cooperative learning on achievement and interpersonal skills. Their meta-analysis revealed that
cooperative learning consistently led to higher academic achievement, better attitudes toward
learning, and improved social skills across various racial and socioeconomic groups. These
findings align with the effect paradigm, which focuses on measuring the outcomes of

23
collaborative versus independent learning. This demonstrates that collaborative learning
strategies, when designed with CRP principles, can significantly enhance academic performance
and social skills across diverse student groups. Specifically, collaborative learning environments
can foster critical consciousness by encouraging students to engage in mutual support and active
participation, thus preparing them to navigate and challenge societal inequities.
The racial and socioeconomic diversity of these samples is crucial as it reflects broader
societal disparities in educational opportunities and outcomes. However, limited research exists
on the impact of peer collaboration on educationally disadvantaged minority students, especially
within urban settings (Samaha & De Lisi, 2000). This underscores the need for further
investigation into the effectiveness of collaborative learning strategies within diverse educational
contexts, particularly among marginalized student populations, to address the unique challenges
they may face.
Characteristics of Collaborative Learning Method
Combining insights from seminal studies on collaborative learning with the assistance of
technology support, an analysis of the characteristics influencing collaborative learning methods
reveals a multifaceted view of educational strategies. Johnson et al. (2010) explores how
collaborative learning through social annotation tools, specifically the HyLighter tool, impacts
reading comprehension and critical thinking in college English courses. Their study found that
students who engaged in collaborative annotation not only improved their understanding of the
text but also developed higher-order thinking skills due to the interactive discussions facilitated
by the annotations. This highlights the significance of interaction within a learner community,
demonstrating that the quality and nature of interactions can significantly influence learning
outcomes.

24
Further, comparative studies such as those by Wendt and Rockinson-Szapkiw (2014) and
Kwon and Cifuentes (2007, 2009) delved into how the mode of collaboration and the structure of
activities influence learning outcomes. Wendt and Rockinson-Szapkiw’s (2014) findings suggest
that face-to-face collaborative learning may be more effective in reducing science
misconceptions among middle school students than online collaborative methods. This suggests
that direct, in-person interactions may facilitate more immediate feedback and richer
communication, which are critical for conceptual understanding in science education. In a
different context, Kwon and Cifuentes (2007, 2009) compared individual versus collaborative
concept mapping, finding that while individual concept mapping might be more effective for
learning specific science concepts due to the focused and personalized nature of the task,
collaborative concept mapping leads to higher quality outputs and deeper conceptual
understanding. This is because group engagement allows students to benefit from diverse
perspectives and collective problem-solving, even though it does not specifically address the
comparison between online and face-to-face methods.
Borokhovski et al. (2015) underscore the effectiveness of structured interaction
treatments in collaborative learning, demonstrating that well-designed, technology-enhanced
collaboration can significantly outperform traditional instructional methods. Their meta-analysis
indicates that when collaborative learning activities are carefully designed to include structured
interactions, clear objectives, and supportive technology, they lead to substantial gains in student
achievement. For example, structured peer discussions and problem-solving tasks supported by
interactive software can help students articulate their reasoning, receive timely feedback, and
refine their understanding.

25
Chen et al. (2018) show that computer-supported collaborative learning (CSCL)
positively affects knowledge acquisition, skill development, and learner perceptions. Their
research highlights that CSCL environments, which integrate tools like shared digital workspaces
and communication platforms, facilitate effective collaboration by making it easier for students
to share resources, communicate asynchronously or synchronously, and collaboratively construct
knowledge. These environments support the social constructivist approach, where learning is
seen as a social process enhanced by interactive engagement within a community.
These studies collectively highlight the pivotal role of intentional design in collaborative
learning activities and the indispensable role of technology in enhancing these processes. They
align with social constructivist theories that advocate for knowledge construction through active
engagement and interaction within a learning community (Kreijns et al., 2003; Resta &
Laferriere, 2007; as cited in Chen et al., 2018). Technology integration supports collaboration
and amplifies the learning process, enabling dynamic, accessible, and engaging interactions
among learners (Schacter, 2000). Schacter (2000) specifically found that CSCL environments led
to improved student engagement and learning outcomes in diverse subjects, emphasizing the
importance of well-designed digital tools in facilitating effective collaborative learning.
The transformative potential of collaborative learning methodologies in education
highlights the critical interplay between structure, technology, and social interaction in
maximizing learning outcomes. Carefully considering collaborative learning characteristics—
such as the interaction mode, activity structure, and technology integration—is essential for
designing and executing effective educational strategies. These insights offer valuable guidance
for optimizing collaborative learning environments, enhancing educational benefits, fostering a
deeper understanding, critical thinking, and an engaging learning experience.

26
The Present Synthesis
Collaborative learning is recognized for fostering deeper understanding, enhancing
communication and social skills, increasing student engagement, promoting higher-order
thinking, and preparing students for real-world collaboration. Despite these recognized benefits,
previous research has predominantly focused on white students. This current synthesis explores
the connection between collaborative learning and academic achievement among culturally
diverse students, particularly Black and Latinx students, in the United States. Given the
persistent educational disparities faced by these groups, this synthesis aims to analyze the
effectiveness of collaborative learning strategies in promoting academic achievement among
diverse student populations, including at least 40% of Black and Latinx students.
Building on Hattie’s (2023) synthesis, which emphasizes the positive impact of
collaborative learning on student engagement and achievement, this study aims to expand these
findings to culturally diverse classrooms. For example, Caldwell (2006) and Johnson et al.
(2010) found that active and collaborative learning environments promote deeper understanding
and real-world application, significantly enhancing learning and performance outcomes. These
findings suggest the potential for such environments to benefit diverse student populations,
including Black and Latinx students, by providing engaging and culturally responsive
educational experiences. Similarly, Wendt and Rockinson-Szapkiw (2014) emphasize the
benefits of collaborative learning in enhancing student motivation, engagement, and knowledge
acquisition. This aligns with research indicating that collaborative learning can be particularly
effective for minority students by fostering a sense of belonging and validating their cultural
backgrounds, thereby supporting the hypothesis that collaborative learning improves academic
achievement among these student groups (Gay, 2000; Ladson-Billings, 1995).

27
The role of technology in collaborative learning has also been a subject of investigation.
Studies suggest that CSCL can enhance students’ cognitive structures and knowledge
organization. Kwon and Cifuentes (2009) and Schacter (2000) provide evidence that CSCL
environments can be as effective as individualized adult-student tutoring, highlighting the
potential of technology to support diverse student populations. However, technology integration
can sometimes hinder learning. For example, Wendt and Rockinson-Szapkiw (2014) discuss how
technology integration can be distracting if students are not adequately trained to use it
effectively or if technical issues arise. Additionally, Kirschner et al. (2018) found that poorly
designed digital tools can lead to cognitive overload, reducing the effectiveness of collaborative
learning. These examples underscore the need to examine the context and implementation of
technology to ensure it supports rather than hinders learning.
For the first research question (What is the effect of collaborative learning on academic
achievement among Black and Latinx students?), I hypothesize that collaborative learning will
significantly improve academic achievement among racially diverse student populations in the
United States. This hypothesis is grounded in the belief that collaborative learning strategies,
which promote engagement and learning through group interaction, are particularly beneficial for
Black and Latinx students. These strategies aim to improve academic achievements by
addressing systemic educational disparities and valuing the cultural experiences and perspectives
of diverse student populations, thus potentially leading to more equitable educational outcomes.
For the second research question (Does inclusion of technology explain variability in the
effect of collaborative learning on academic achievement among Black and Latinx students?), I
hypothesize that the impact of collaborative learning on student achievement will be significantly
larger when technology is integrated into the learning process. This hypothesis is based on the

28
belief that technology-enhanced collaborative learning modalities, such as CSCL environments,
can address systemic inequities and meet specific educational needs by providing engaging and
culturally responsive learning experiences. For example, tools like interactive software and
digital workspaces can facilitate better communication and resource sharing, which are
particularly beneficial in diverse classrooms.
For the third research question (To what extent does the effect of collaborative learning
on academic achievement vary depending on the percentage of the sample that is Black and/or
Latinx?), I hypothesize that the impact of collaborative learning on academic achievement will
vary according to the proportion of Black and/or Latinx students in the study population.
Drawing on the conditions paradigm, which posits that educational outcomes are influenced by
contextual factors, I predict that higher percentages of Black and/or Latinx students in the sample
will correlate with stronger positive effects of collaborative learning. This is due to the culturally
relevant and supportive environment that collaborative learning can foster, which is particularly
beneficial for these student groups.
These hypotheses aim to summarize how collaborative learning affects academic
achievement in culturally diverse United States classrooms, with a specific focus on Black and
Latinx students. By evaluating the impact of different collaborative learning modalities, the role
of technology, and the demographic composition of the student body, this study seeks to provide
insights into effective strategies for promoting equitable educational outcomes.
Methods
This meta-analysis, inspired by Hattie’s influential work (2023), explores the relationship
between collaborative learning and academic achievement among culturally diverse United
States students, focusing on African American and Latinx participants. Using a rigorous search

29
strategy and systematic screening, the meta-analysis aims to comprehensively synthesize
research literature, providing valuable insights into the effectiveness of collaborative learning for
diverse student academic achievement.
Literature Search
The articles incorporated into this synthesis were part of Hattie’s (2023) meta-analysis,
and the meta-analyses for all influences examined in that study are listed on the Visible Learning
website. This study focused on the influence of “collaborative learning” under the domain of
teaching strategies, finding a list of 7 relevant meta-analyses on the website. Out of these, five
were ultimately included in this synthesis. One meta-analysis by Williams (2009) was located
but did not provide the included studies and attempts to obtain the necessary information were
hindered by the unavailability of contact information. Despite diligent efforts, neither I nor my
committee could locate the author’s contact details. Consequently, we decided to leave this
information out of the dissertation. The second meta-analysis, by Jansen (2014), could not be
located even after attempting to email the authors. Additionally, it was discovered that the
author’s name was misspelled on Visible Learning website. The third meta-analysis, by
Borokhovski et al. (2015), was located, but only 25 studies were found. I emailed the author for
assistance with the remaining twenty, but there has been no response. Around two hundred fiftyfive studies were found in the fourth meta-analysis by Chen et al. (2018). However, I emailed the
author for assistance with the remaining hundred studies, and unfortunately, there has been no
response. The meta-analysis by Borokhovski et al. (2012), Tenenbaum et al. (2019), and Jeong et
al. (2019) were all located with the corresponding number of indicated studies.
The meta-analyses listed on the Visible Learning website can be located through the
University of Southern California (USC) Library Database. A list of the studies in all of the

30
meta-analyses was compiled in a Google Spreadsheet on a Google Drive folder for the
collaborative learning influence. Each study was first searched for using the USC Library
Database. If it was not found in the USC Library Database, the study was also searched for using
Google Scholar, ProQuest, ERIC, and PsycINFO. Once the article was located, it was
hyperlinked in the Google Sheets document along with recorded information regarding the
influence: the authors, the year of publication, and a record of whether or not a copy of the article
was added to the Google Drive folder.
However, due to reasons stated prior regarding locating and including meta-analyses,
only 520 studies from 791 reports, or 65.7% of the original studies in meta-analyses included in
Visible Learning, were identified and screened for inclusion. Among the 520 reports, 24 (4.6%)
were unavailable, 54 (10%) were duplicate reports and removed, leaving 442 reports to be
retrieved. The initial screening process was meticulously documented in a spreadsheet, capturing
essential details such as: the availability of the report; the source from which it was obtained;
whether the research was conducted in the United States; whether it provided information on the
racial/ethnic composition of the sample; and the proportions of participants from Asian, Black,
Latinx, White, and other ethnic groups within the study. Of the 442 reports that were retrieved,
eight reports met inclusion criteria and were retained for coding (Figure 1).
Inclusion Criteria
The current synthesis employed specific inclusion criteria to determine the eligibility of
studies for inclusion. These criteria encompassed various characteristics that were required for a
study to be included. Firstly, the study needed to have been included in a prior meta-analysis
from the Hattie overview and have a sample situated in the United States of America, with at
least 40% of the sample composed of Black or Latinx students. Additionally, to ensure the

31
validity and reliability of our meta-analysis, only studies that included a control condition were
considered. This criterion ensures a robust comparison between collaborative learning
interventions and traditional or other instructional methods, allowing for a more accurate
assessment of the effects of collaborative learning on academic achievement.
Achievement had to be measured at the student level, with reports providing sufficient
data to calculate an effect size. This criterion was crucial to ensure that the studies included in
the meta-analysis could contribute to a quantifiable synthesis of the effect of collaborative
learning on academic achievement. Information for inclusion was gathered from each study’s
abstract and/or methods section. Studies were excluded if they lacked data on the country or
race/ethnicity of participants. All relevant inclusion data were meticulously recorded in a Google
Sheets document. Initially, seven meta-analyses were identified, but two were excluded due to
unavailability, leaving five meta-analyses for review. From these, 791 records were screened,
271 unavailable records were omitted, resulting in 520 reports for retrieval and screening.
Among these, 54 reports were duplicates and 24 reports were unavailable, leaving 442 reports to
be assessed for eligibility.
Exclusions were made for reports not conducted in the United States (k = 288) and those
lacking racial/ethnicity data (k = 422). Some reports met both of these criteria. This resulted in
the exclusion of 442 reports. Ultimately, eight reports met all inclusion criteria. One study was
later excluded due to insufficient information for coding, leaving seven studies in the final
synthesis. This stringent inclusion process ensured that the meta-analysis remained aligned with
the research objectives and maintained high integrity. The process is visually summarized in
Figure 1, highlighting a systematic approach to examining the effects of collaborative learning
on academic achievement among Black and Latinx students.

32
Data Extraction
The coding procedure was undertaken by the primary researcher (myself), under the
supervision of the dissertation co-chairs, with the participation of other graduate students from
the cohort who received training for this specific meta-analysis. Weekly meetings were
conducted for coding training and discussions to ensure consistency and reliability in coding. To
ensure the validity and reliability of our meta-analysis, we considered experimental studies with
control conditions. During the data extraction process, we ensured that experimental studies
included a control condition to maintain the integrity of our comparative analysis. The training
commenced with thoroughly explaining the coding process, utilizing a preread article as a guide,
and providing step-by-step instructions on data input. Subsequently, each cohort member could
independently code articles to assess their comprehension of the process. Additionally, two new
articles were assigned for individual coding, and collaborative practice sessions were organized
among the cohort students to facilitate mutual support and enhance coding skills.
To foster a comprehensive understanding of the coding process, each cohort member had
the chance to share their coding approach and findings in weekly training sessions. Through
open discussions and collective decision-making, the cohort members would reach a consensus,
with the guidance and assistance of the supervising professors, whenever necessary. A dedicated
session was devoted to practicing the coding of effect sizes across multiple studies. The coding
practice sessions were ongoing to further refine and enhance the coding skills of all involved.
The coding guide consisted of various items related to the meta-analysis characteristics,
report characteristics, participant and sample characteristics, predictor influences, outcome
measures, research design, and the calculation of effect sizes. Meta-analysis characteristics that
were coded encompassed the meta-analysis name. Regarding report characteristics, coding

33
included publication type, data sources, the year of data collection, and whether the report
utilized an overlapping dataset. Setting characteristics that were coded involved the study’s
location by region and school level. Participant and sample characteristics encompassed
information on whether the sample was analyzed as a whole, as a subgroup, or both; percentages
pertaining to race and ethnic groups; grade level; gender; percentage of the sample identified as
low-income or economically disadvantaged; percentage of the sample in special education; and
the percentage of the population that were English language learners. Coding for technology use
included whether and how technology was used in the classroom, types of technology utilized
(e.g., computers, tablets, educational software), frequency of technology use, and the context in
which technology was implemented. Coding for influence and predictor measures included the
influence definition in the report, how the influence was measured, the reliability of the survey
instrument used, and how the researcher manipulated the influence. Outcome measures slated for
coding encompassed various types, including state standardized tests and GPA; descriptions of
outcomes; the domain of the outcome; the unit of analysis; the timing of the influence; and
whether the data collection was simultaneous or longitudinal. Research design coding included
the type of control condition, and how the control group was managed and compared to the
experimental group. The coding guide is available in the Appendix B.
By following this systematic coding procedure and employing multiple quality control
measures, including regular training sessions, collaborative discussions, and ongoing practice,
the data extraction process for this dissertation aimed to ensure the accuracy, reliability, and
comprehensiveness of the collected data. Following the extraction of data for studies that met the
inclusion criteria, an experienced graduate student collaborator reviewed the data for accuracy to
ensure the reliability of the coding and data extraction process. The agreement rate with the

34
validator was determined by dividing the number of data points that matched by the total number
of data points extracted. The agreement rate between the researcher and the validator was
approximately 99%, indicating a high level of consistency.
Computing Effect Sizes and Data Analysis
Effect sizes for intervention studies will be computed as standardized mean differences
(SMD) on achievement between collaborative learning groups and control groups. Calculations
will be based on means, standard deviations, and sample sizes for intervention and control
groups whenever possible. In cases where direct calculations are not feasible, effect sizes were
derived from standard error statistics. For studies involving multiple treatment conditions
compared to a single control condition, separate effect sizes were calculated for each intervention
condition. All intervention effect sizes were converted to bias-corrected Hedge’s g, a
standardized effect size correcting for slight positive bias in effects observed with small samples
(Hedges, 1981).
The meta-analysis of this intervention data was analyzed using the metafor and
clubSandwich R packages (Pustejovsky, 2019; Viechtbauer, 2010). Random-effect modeling was
employed throughout the analysis. To address the dependency between effect size estimates
within studies and account for potential model misspecification, a multi-level modeling approach
with a robust variance estimator (RVE; Pustejovsky & Tipton, 2020) was adopted. A randomeffects model was fitted to estimate the pooled effect size for the relationship between
collaborative learning and achievement. Heterogeneity among effect sizes was assessed using Q,
τ2, and I2 statistics. The weighted average effect was reported with 95% confidence intervals
(CI) (Borenstein et al., 2011).

35
Mixed-effects meta-regression models were employed to explore the sources of
heterogeneity in the effect size estimates. Separate models were used to examine the effects of
moderators, including the percentage of the sample that is Black, the percentage of the sample
that is Latinx, and the inclusion of technology. Additionally, the possibility of publication bias
and funnel plot asymmetry was examined through Egger’s regression test (Egger et al., 1997)
and by considering publication status as a moderator in meta-regression models.
Results
In total, seven studies meeting the inclusion criteria were incorporated into the final
sample. Of these, six were published in journals, while one remained as an unpublished
dissertation. These studies yielded a total of 10 samples and 34 effect sizes. Nine samples
originated from published sources, with the remaining sample from the unpublished dissertation.
The publication timeline of these studies spans from 2000 to 2014.
As for the context of the studies, three studies involved computer-based concept
mapping, two studies focused on face-to-face versus online collaboration, one study examined
peer collaboration on nonverbal intelligence tests, and one study involved communal problemsolving with either computer simulations or physical tools. In terms of cultural relevance, only
one study explicitly incorporated culturally relevant pedagogy, while the remaining six studies
did not explicitly mention culturally relevant aspects. Regarding technology use, three studies
used computer-based concept mapping tools, two studies utilized educational platforms like
Blackboard and Edmodo for web-assisted and online learning, one study involved computer
simulations for problem-solving tasks, and one study did not specify the use of any particular
technology. Detailed information regarding the authors, sample sizes, and effect sizes for these

36
studies can be found in Appendix A of the supplementary materials, alongside other pertinent
study characteristics.
Overall Average Effects/Correlations
The analysis revealed a statistically non-significant negative average effect size of
collaborative learning on achievement (g = –0.0267, see Table 1). This suggests that the
intervention did not significantly affect achievement outcomes among the populations studied.
However, the relatively limited number of studies and effect sizes analyzed emphasize the need
for careful interpretation and potential expansion in future research to confirm these findings.
Interestingly, the expected stronger relationship between collaborative learning and academic
achievement for more diverse samples, including Black and Latinx students, was not observed.
This is in contrast to the effect size of 0.45 reported in Hattie’s overview, which found a more
substantial positive impact of collaborative learning on academic achievement within
populations across diverse student populations.
Most studies demonstrated that collaborative learning had some positive effects on
learning outcomes, although the degree of effectiveness varied. For instance, Kwon and
Cifuentes (2007, 2009) and Serpell et al. (2006) found that collaborative learning could improve
comprehension and problem-solving skills compared to non-collaborative methods. Wendt and
Rockinson-Szapkiw (2014) showed that face-to-face collaboration was more effective than
online collaboration in reducing misconceptions and improving science literacy. These findings
suggest that while collaborative learning generally has beneficial effects, the mode of
collaboration and specific educational contexts are critical factors in determining its success.
The studies also revealed several challenges associated with collaborative learning. Both
Kwon and Cifuentes (2007, 2009) and Wendt and Rockinson-Szapkiw (2014) found that

37
collaboration did not always yield superior results over individual efforts, especially in an online
context. Wendt and Rockinson-Szapkiw (2014) identified issues such as delayed feedback and
reduced non-verbal cues in online collaboration, which could hinder learning effectiveness.
These challenges underscore the need for effective implementation strategies to maximize the
benefits of collaborative learning. The meta-analysis revealed a substantial amount of
heterogeneity, as indicated by Cochran’s Q (Q = 133.00, p < 0.0001), tau-squared (τ² = 0.1733),
and I-squared (I² = 74.73%). The value of τ² suggests considerable variation among the true
effect sizes, while the high I² value indicates that 74.73% of the variation in effect sizes is due to
heterogeneity among the studies. This high level of heterogeneity implies that the effect of the
intervention on achievement varies significantly across different contexts and implementations.
Given the diverse settings and methodologies applied in the included studies, such variability is
expected. Therefore, while the overall effect is negligible, further analysis to explore potential
moderating factors is necessary to understand the sources of this heterogeneity.
Publication Bias
The analysis from Egger’s regression model indicated evidence of funnel plot (Figure 2)
asymmetry in the dataset (b = –0.0332, SE = 0.671, p < 0.05). The moderator analysis for
publication status suggested no significant difference in effect sizes between published and
unpublished studies. However, the limited data calls for careful interpretation and the need for
additional research.
Moderator Analyses
This study focuses on two critical moderators: the proportion of Black and/or Latinx
students and the incorporation of technology in classroom settings (Table 2). These moderators
were chosen due to their significant relevance to the research context. Analyzing the variations in

38
sample sizes of Black and Latinx students provides insights into the distinct experiences within
these demographic groups. Examining the use of technology in the classroom offers insights into
how digital tools and platforms can enhance or impede the effectiveness of collaborative learning
and its impact on academic achievement. By exploring how these moderators interact with the
primary variables of interest, the study aims to deepen our understanding and provide valuable
insights into the complex interplay of factors shaping educational outcomes. Due to the small
sample size, the statistical power of the moderator analyses is limited. Most subgroups included
more than three calculated effects. Specifically, we analyzed a wide range of subgroups across
various categories. In some cases, where there were two or fewer reports in each category, larger
categories were constructed to ensure a more robust analysis. For example, in the technology
incorporation analysis, all reports involving any form of technology were grouped together,
rather than separating web-assisted learning tasks from online learning assessments. However,
for the ethnicity analysis, Black and Latinx samples were analyzed separately. All studies
included a percentage of Black students, while only a portion of studies reported the percentage
of Latinx students. These adjustments were necessary to provide a more robust analysis and
ensure the findings were interpretable. This limitation should be considered when interpreting
the results.
Moderator: Percentage of Latinx students
The analysis revealed a slightly negative but statistically non-significant relationship
between the percentage of Latinx students and the impact of collaborative learning on academic
achievement scores (b = –0.0013, SE = 0.0027). This indicates that as the percentage of Latinx
students increases, the effect of collaborative learning on academic achievement becomes
slightly more negative, although this relationship is not statistically significant. The confidence

39
interval for this relationship ranges from –0.0067 to 0.0041, suggesting that the percentage of
Latinx students is not a significant moderator of the effectiveness of collaborative learning. The
non-significant results indicate that the demographic makeup, in terms of Latinx student
percentages, does not significantly alter the impact of collaborative learning on academic
achievement. This underscores the need to consider other contextual factors that may influence
the outcomes of collaborative learning interventions beyond demographic composition alone.
Moderator: Percentage of Black Students
The analysis of the Black moderator showed a regression coefficient of –0.0001 with a
standard error of 0.0069. This suggests a very slight, non-significant negative relationship
between the percentage of Black students and the effect of collaborative learning on academic
achievement. The 95% confidence interval for this relationship spans from –0.0141 to 0.0139,
indicating that the effect is statistically non-significant. This finding suggests that the percentage
of Black students does not significantly moderate the impact of collaborative learning on
academic achievement. While the studies did not explicitly focus on CRP as a primary variable,
Serpell et al. (2006), which exhibited positive effects, incorporated elements aligned with
culturally relevant educational practices, particularly through the communal learning context.
This underscores the general gap in explicitly integrating CRP in collaborative learning
interventions. Addressing this gap could provide valuable insights into how collaborative
learning strategies can be tailored to meet the diverse cultural needs of students. The lack of
significant impact implies that collaborative learning’s effectiveness, or lack thereof, is stable
across varying percentages of Black students. This stability highlights the need to consider other
contextual factors that might influence the outcomes of collaborative learning interventions, as
demographic variables alone do not appear to explain variations in effectiveness.

40
Moderator: Technology
The analysis for the technology moderator revealed the effects of collaborative learning
on achievement did not vary based on the presence of technology and were not statistically
significantly different from zero with or without technology. Without technology, the regression
coefficient indicated an effect size (g) of –0.0316 for the effect of collaborative learning on
achievement, with a 95% confidence interval ranging from –0.4034 to 0.3402. This suggests a
slight, non-significant negative effect of collaborative learning without technology on academic
achievement. In contrast, when technology was present, the effect of collaborative learning was
–0.0210, with a 95% confidence interval ranging from –0.4134 to 0.3714. This indicates a nonsignificant effect of collaborative learning with the presence of technology on academic
achievement. The difference between the groups was not significant, 0.0106 (SE = 0.0707).
Overall, these findings imply that the presence or absence of technology does not significantly
alter the impact of collaborative learning on academic achievement. These results suggest that
while technology can be a useful tool in education, its mere presence does not guarantee
enhanced outcomes from collaborative learning, indicating the need for thoughtful integration of
technology to maximize its benefits. Several studies involved the use of technology in their
interventions, highlighting its role in facilitating collaborative learning. Caldwell (2006) utilized
the Blackboard course delivery system, while Kwon and Cifuentes (2007, 2009) employed
computer-based concept mapping tools. These studies found that technology could be a useful
tool for enhancing collaborative learning. However, Wendt and Rockinson-Szapkiw (2014)
noted that face-to-face collaboration was more effective than online collaboration, suggesting
that technology might not always be the optimal medium for collaborative learning.

41
Discussion
This study discusses the findings of this meta-analysis, examining the relationship
between collaborative learning and student achievement in racially diverse samples in the United
States and the extent to which technology explains variability in this relationship among Black
and Latinx students.
Summary of Key Findings
The analysis revealed several key findings regarding the impact of collaborative learning
on the academic achievement of Black and Latinx students. Collaborative learning, in its current
implementation, does not significantly impact academic achievement for these student groups
possibly due to the lack of culturally responsive practices in many implementations.
Additionally, the inclusion of technology did not significantly moderate this relationship. The
percentage of Black and Latinx students in the sample also did not significantly alter the
effectiveness of collaborative learning.
Research Question 1 (What is the effect of collaborative learning on academic
achievement among Black and Latinx students?), key findings indicate that collaborative
learning has no detectable effect on the achievement of Black and Latinx students, with an
average effect size of g = −0.0267, 95% CI [–0.395, 0.341]. This effect size suggests that
collaborative learning, in its current implementation across the studied contexts, does not
significantly contribute to academic success for these student groups. The findings do not
support the hypothesis that collaborative learning enhances academic achievement among Black
and Latinx students. This may be due to the absence of culturally responsive practices in many
implementations, as emphasized by Serpell et al. (2006), who found that CRP-aligned communal
learning contexts are essential for improving academic outcomes for African American students.

42
The results highlight significant variability in the effectiveness of collaborative learning,
depending on specific conditions and contexts. This is reflected in the high level of heterogeneity
among the included studies, indicated by an I² value of 74.73% and a Q value of 133.00. Most
studies demonstrated that collaborative learning had some positive effects on learning outcomes,
although the degree of effectiveness varied. For instance, Kwon and Cifuentes (2007, 2009) and
Serpell et al. (2006) found that collaborative learning could improve comprehension and
problem-solving skills compared to non-collaborative methods. Furthermore, Wendt and
Rockinson-Szapkiw (2014) showed that face-to-face collaboration was more effective than
online collaboration in reducing misconceptions and improving science literacy.
Given the small and near-zero effect size, it is evident that other factors may influence the
effectiveness of collaborative learning, such as implementation quality, student engagement, and
curriculum integration. Therefore, future research should explore additional contextual factors,
such as socio-economic status, school resources, and teacher-student ratios, to better understand
their impact on collaborative learning’s effectiveness
Research Question 2 (Does inclusion of technology explain variability in the effect of
collaborative learning on academic achievement among Black and Latinx students?), the analysis
revealed that the presence of technology does not significantly moderate the relationship between
collaborative learning and academic achievement. Black and Latinx students who engaged in
collaborative learning using technological tools showed an average effect size of g = –0.0210,
95% CI [–0.4134, 0.3714], compared to those who did not use technology g = –0.0316, 95% CI
[–0.4034, 0.3402]. In both cases, the effect sizes are very close to zero and negative, indicating a
minimal and non-significant impact on academic achievement.

43
These findings do not support the hypothesis that technology-enhanced collaborative
learning significantly improves academic achievement among Black and Latinx students. The
integration of technology aligns with the principles of CSCL, which emphasizes the use of
digital tools to enhance collaborative learning. Studies by Kwon and Cifuentes (2007, 2009) and
Schacter (2000) indicate that technology can provide scaffolding and additional resources,
potentially improving learning outcomes. However, this meta-analysis did not find a significant
positive effect when technology was used.
The effectiveness of technology in collaborative learning can be heavily dependent on
how well it is implemented. Poor implementation or lack of proper training for teachers and
students might negate the potential benefits of technology. For instance, Wendt and RockinsonSzapkiw (2014) identified issues such as delayed feedback and reduced non-verbal cues in online
collaboration, which could hinder learning effectiveness. Additionally, the types of technology
used, differences in educational settings, student populations, curriculum design, and variability
in measuring academic achievement can impact the effectiveness of technology in collaborative
learning.
Research Question 3 (To what extent does the effect of collaborative learning on
academic achievement vary depending on the percentage of the sample that is Black and/or
Latinx?) the analysis revealed that the percentage of Black students does not significantly
moderate the relationship between collaborative learning and academic achievement. The
regression coefficient for Black students was b = –0.0001 (SE = 0.0069), with a 95% confidence
interval ranging from –0.0141 to 0.0139.
Similarly, the percentage of Latinx students also does not significantly moderate the
relationship between collaborative learning and academic achievement. The regression

44
coefficient for Latinx students was b = –0.0013 (SE = 0.0027), with a 95% confidence interval
ranging from –0.0067 to 0.0041. These findings suggest that the percentage of Black and Latinx
students in the sample does not significantly alter the effectiveness of collaborative learning. The
near-zero and non-significant effect sizes imply that collaborative learning’s impact on academic
achievement remains stable regardless of the proportion of Black and Latinx students.
The importance of culturally responsive pedagogy, which emphasizes incorporating
students’ cultural backgrounds into the learning process, is significant. Ladson-Billings (1995,
1998) and Gay (2002) argue that such practices help close the opportunity gap by valuing
students’ cultural experiences. For instance, in Serpell et al. (2006), elements aligned with
culturally relevant educational practices, particularly through the communal learning context,
were incorporated. This study highlights that if learning activities do not align with the cultural
experiences of Black and Latinx students, their impact may be diminished. Given this, the nonsignificant regression analyses indicate a need for further research into optimizing culturally
responsive teaching practices to enhance educational outcomes. The high heterogeneity among
studies suggests that future research should focus on identifying specific conditions and contexts
under which collaborative learning can be more effective for diverse student populations.
Alignment of Key Findings With Theory and Prior Research
The findings of this study offer important insights into the potential and limitations of
collaborative learning and the moderating role of technology for Black and Latinx students.
Constructivist theories, particularly those proposed by Bruffee (1993), suggest that collaborative
learning enhances learning through social interaction and the co-construction of knowledge.
However, this study’s findings indicate that factors such as implementation quality and
alignment with students’ cultural experiences are crucial for realizing these benefits. Inadequate

45
training for teachers and students or insufficient alignment of learning activities with the cultural
backgrounds of Black and Latinx students might have hindered its effectiveness.
Carlone and Johnson (2012) highlight the importance of understanding cultural
production within educational settings. Their work suggests that integrating students’ cultural
backgrounds into the learning environment can significantly influence educational outcomes.
Prior research on technology-enhanced collaborative learning, such as the studies by Kwon and
Cifuentes (2007, 2009) and Schacter (2000), suggests that technology can enhance collaborative
learning by providing scaffolding and additional resources. However, as noted in the summary,
this study found no significant moderating effect of technology on the relationship between
collaborative learning and academic achievement. This implies that the potential benefits of
technology were not realized, possibly due to poor implementation or lack of proper training.
Studies included in the meta-analysis, such as those by Johnson et al. (2010) and Wendt
and Rockinson-Szapkiw (2014), highlight the importance of well-designed collaborative tasks
and the role of technology in facilitating effective group interactions. The non-significant
findings point to the variability in how collaborative learning and technology were implemented
across different educational settings.
These insights highlight the need for future research to explore the specific conditions
and contexts under which collaborative learning and technology integration can be more
effective for Black and Latinx students. Addressing these factors is crucial for reducing the
opportunity gap and promoting equitable educational outcomes. Future studies should focus on
identifying best practices for implementing collaborative learning in culturally responsive ways
and leveraging technology to support diverse learners effectively.

46
Implications for Theory
The findings of this study have several implications for existing theories on collaborative
learning and educational technology, particularly in the context of racially diverse student
populations. The lack of significant impact of collaborative learning on the academic
achievement of Black and Latinx students suggests a need to revisit constructivist theories, such
as those proposed by Bruffee (1993). Constructivist theories emphasize the benefits of social
interaction and the co-construction of knowledge, but our findings indicate that these benefits are
not universally realized. Specifically, the negligible effect size (g = −0.0267) and high
heterogeneity (I² = 74.73%) suggest that implementation quality and cultural alignment are
critical factors. Without adequate training for teachers and alignment of learning activities with
students’ cultural backgrounds, the theoretical benefits of collaborative learning may not
materialize. As Bruffee (1999) noted, collaborative learning requires conditions where
knowledge can be socially constructed within a community, but this may not happen effectively
without cultural responsiveness.
The non-significant moderating effect of technology (effect size with technology: g =
–0.0210, without technology: g = –0.0316) challenges the optimistic claims made by proponents
of CSCL. While prior research, such as Kwon and Cifuentes (2007, 2009) and Schacter (2000),
suggests that technology can enhance collaborative learning by providing scaffolding and
additional resources, our findings indicate that technology integration alone does not
significantly impact academic outcomes. This suggests that CSCL theories need to consider the
complexities of technology implementation, including factors like the adequacy of teacher
training and the cultural relevance of technological tools.

47
Our findings regarding the percentage of Black and Latinx students (regression
coefficients: Black students b = –0.0001, Latinx students b = –0.0013) indicate that having a
higher percentage of these students in collaborative learning environments does not significantly
alter academic outcomes. This challenges assumptions in theories of culturally responsive
pedagogy (Ladson-Billings, 1995, 1998; Gay, 2002) that simply increasing the representation of
minority students will enhance learning outcomes. Our study suggests that culturally responsive
practices need to be more specifically integrated into collaborative learning activities to be
effective. As described by Carlone and Johnson (2012), incorporating students’ cultural
backgrounds into the learning process can make collaborative learning more relevant and
engaging, but this requires deliberate effort and design.
Overall, this study underscores the importance of context-specific factors in educational
theories related to collaborative learning and technology. Future theoretical work should focus
on integrating these contextual considerations, particularly the role of cultural responsiveness
and the nuances of technology implementation, to develop more comprehensive and applicable
frameworks for understanding and improving educational outcomes for diverse student
populations.
Implications for Practice
The findings of this study have significant implications for educators, policymakers, and
other professionals involved in the design and implementation of educational programs,
particularly for racially diverse student populations. One way to address this is by providing
comprehensive training for teachers, equipping them with effective collaborative learning
strategies and cultural responsiveness. Professional development programs should focus on
collaborative learning techniques (Johnson & Johnson, 2010), understanding and incorporating

48
students’ cultural backgrounds (Gay, 2002; Ladson-Billings, 1995), and fostering inclusive
classroom environments that promote equitable participation and engagement (Brown &
Crippen, 2016).
The study underscores the importance of cultural alignment in collaborative learning.
Designing activities that incorporate students’ cultural experiences can make learning more
relevant and engaging, potentially improving outcomes for Black and Latinx students. Educators
should consider integrating community-based projects, culturally relevant case studies, and
allowing students to have input in the selection of topics and activities (Carlone & Johnson,
2012). This approach addresses the finding that the theoretical benefits of collaborative learning
may not materialize without cultural responsiveness (Ladson-Billings, 2020).
While the immediate impact on academic achievement might not always be evident, the
long-term benefits of collaborative learning are substantial. Collaborative learning environments
can significantly enhance social skills, including communication, teamwork, and conflict
resolution (Samaha & De Lisi, 2000). These environments also increase student engagement and
motivation, making learning more enjoyable and stimulating (Kwon & Cifuentes, 2007, 2009).
Additionally, collaborative learning encourages critical thinking and creative problem-solving.
Serpell et al. (2006) demonstrated that communal problem-solving tasks help students develop
these essential skills. The practice of explaining concepts to peers and receiving feedback fosters
deeper understanding and better retention of knowledge (Kwon & Cifuentes, 2009). Peer support
and scaffolding within collaborative groups provide significant academic and emotional benefits,
creating a supportive learning environment for all students. Over time, these environments
contribute to better academic outcomes as the enhanced social skills, increased engagement, and
deeper understanding fostered by collaborative learning manifest in improved performance.

49
The non-significant moderating effect of technology on the relationship between
collaborative learning and academic achievement indicates the need for better integration of
technology in educational practices. Effective integration requires ensuring that both teachers
and students are adequately trained to use technological tools purposefully to facilitate
meaningful collaboration (Kwon & Cifuentes, 2007; Schacter, 2000). This involves providing
access to necessary technological resources and using digital platforms for group projects and
discussions (Wendt & Rockinson-Szapkiw, 2014). By focusing on these key areas—teacher
training, culturally responsive activities, and effective technology integration—educators and
policymakers can enhance the effectiveness of collaborative learning and support the academic
achievement of Black and Latinx students.
Limitations and Recommendations for Future Research
This meta-analysis has several limitations. The studies included varied in their definitions
and implementations of collaborative learning and technology use, which may have introduced
heterogeneity in the results. Additionally, the analysis did not account for other potential
moderating variables, such as socioeconomic status or prior academic achievement. The sample
sizes of the included studies were relatively small, which may limit the generalizability of the
findings and reduce the statistical power of the analysis. The slightly negative overall effect size
and non-significant moderating effect of technology highlight the complexity and variability in
how collaborative learning and technology are implemented and studied across different
contexts.
Collaborative learning benefits may take more time to manifest in academic achievement,
suggesting the need for longitudinal studies to capture these delayed effects. For instance, Wendt
and Rockinson-Szapkiw (2014) indicated that the cumulative effects of collaboration could

50
become more evident over extended periods. Standardized achievement tests may not fully
capture the multifaceted benefits of collaboration, such as enhanced social skills, critical
thinking, and student engagement. Future research should employ a combination of quantitative
and qualitative measures to provide a more comprehensive assessment of collaborative
learning’s impact.
The studies reviewed also highlighted the need for a supportive learning environment that
fosters effective collaboration. Factors such as group composition, task types, and the role of
technology significantly influence the outcomes of collaborative learning. Culturally relevant
pedagogical approaches can enhance the effectiveness of collaborative learning, particularly in
diverse classrooms (Serpell et al., 2006). Future research should focus on longitudinal studies to
understand the long-term effects of technology-enhanced collaborative learning. Investigating
the types of technology that are most effective in different collaborative learning contexts would
also provide valuable insights for educators, allowing for more tailored and effective
implementation strategies.
Conclusions
In conclusion, this meta-analysis reveals that collaborative learning has no detectable
overall effect on student achievement in racially diverse samples, and the role of technology as a
moderator is non-significant. These findings indicate that while collaborative learning and
technology have been considered beneficial, their effectiveness can vary significantly depending
on implementation and context. The slight negative effect observed is so small that it is not
statistically significant. Future research should continue to explore the multifaceted impacts of
collaborative learning and technology, with a focus on identifying specific conditions and
practices that enhance their effectiveness. This ongoing exploration is crucial for informing

51
educational strategies aimed at supporting the academic success of minority students and
promoting equitable educational outcomes.

52
References
Borenstein, M., Hedges, L. V., Higgins, J. P., & Rothstein, H. R. (2011). Introduction to MetaAnalysis. John Wiley & Sons. https://doi.org/10.1002/9780470743386
Borokhovski, Bernard, R. M., Tamim, R. M., Schmid, R. F., & Sokolovskaya, A. (2015).
Technology-supported student interaction in post-secondary education: A meta-analysis
of designed versus contextual treatments. Computers and Education, 96, 15–28.
https://doi.org/10.1016/j.compedu.2015.11.004
Borokhovski, Tamim, R., Bernard, R. M., Abrami, P. C., & Sokolovskaya, A. (2012). Are
contextual and designed student-student interaction treatments equally effective in
distance education? Distance Education, 33(3), 311–329.
https://doi.org/10.1080/01587919.2012.723162
Brown, J. C., & Crippen, K. J. (2017). The Knowledge and Practices of High School Science
Teachers in Pursuit of Cultural Responsiveness. Science Education (Salem, Mass.),
101(1), 99–133. https://doi.org/10.1002/sce.21250
Bruffee, K. A. (1999). Collaborative Learning: Higher Education, Interdependence, and the
Authority of Knowledge. Second Edition. Johns Hopkins University Press.
Burr, V., Dick, P. (2017). Social Constructionism. In: Gough, B. (eds) The Palgrave Handbook
of Critical Social Psychology. Palgrave Macmillan, London. https://doi.org/10.1057/978-
1-137-51018-1_4
Caldwell, E. R. (2006). A comparative study of three instructional modalities in a computer
programming course: Traditional instruction, Web-based instruction, and online
instruction (Order No. 3227694). Available from ProQuest Dissertations & Theses
Global; ProQuest One Academic. (305285277).

53
http://libproxy.usc.edu/login?url=https://www.proquest.com/dissertationstheses/comparative-study-three-instructional-modalities/docview/305285277/se-2
Carlone, H., & Johnson, A. (2012). Unpacking “culture” in cultural studies of science education:
cultural difference versus cultural production. Ethnography and Education, 7(2), 151–
173. https://doi.org/10.1080/17457823.2012.693691
Capar, G., & Tarim, K. (2015). Efficacy of the cooperative learning method on mathematics
achievement and attitude: A meta-analysis research. Educational Sciences-Theory &
Practice, 15(2), 553–559. https://doi.org/10.12738/estp.2015.2.2098
Chen, J., Wang, M., Kirschner, P. A., & Tsai, C.-C. (2018). The Role of Collaboration,
Computer Use, Learning Environments, and Supporting Strategies in CSCL: A MetaAnalysis. Review of Educational Research, 88(6), 799–843.
https://doi.org/10.3102/0034654318791584
Dillenbourg, P., Järvelä, S., & Fischer, F. (2009). The evolution of research on computersupported collaborative learning. In N. Balacheff, S. Ludvigsen, T. d. Jong, A. Lazonder,
& S. Barnes (Eds.), Technology-enhanced learning (pp. 3–19). Springer. 10.1007/978-1-
4020-9827-7_1
Dillenbourg, P. (1999). What do you mean by collaborative learning?. In P. Dillenbourg (Ed.),
Collaborative learning: cognitive and computational approaches (pp. 1–19). Elsevier.
Dillenbourg, P., Baker, M., Blaye, A., & O’Malley, C. E. (1996). The evolution of research on
collaborative learning. In E. Spada, & P. Reiman (Eds.), Learning in humans and
machine: towards an interdisciplinary learning science (pp. 189–211). Elsevier.
Gay, G. (2002). Preparing for Culturally Responsive Teaching. Journal of Teacher Education,
53(2), 106–116. https://doi.org/10.1177/0022487102053002003

54
Hattie, J. (2023). Visible Learning: The Sequel : A Synthesis of Over 2,100 Meta-Analyses
Relating to Achievement (New edition). Routledge.
https://doi.org/10.4324/9781003380542
Hattie, J. (2009). Visible learning: A synthesis of over 800 meta-analyses relating to
achievement (1st ed.). Routledge. https://doi.org/10.4324/9780203887332
Jeong, H., Hmelo-Silver, C. E., & Jo, K. (2019). Ten years of Computer-Supported
Collaborative Learning: A meta-analysis of CSCL in STEM education during 2005–
2014. Educational Research Review, 28, 100284-.
https://doi.org/10.1016/j.edurev.2019.100284
Johnson, D. W., Johnson, R. T., & Stanne, M. B. (2000). Cooperative learning methods: a
meta-analysis. Minneapolis University of Minnesota.
Johnson, T. E., Archibald, T. N., & Tenenbaum, G. (2010). Individual and team annotation
effects on students’ reading comprehension, critical thinking, and meta-cognitive skills.
Computers in Human Behavior, 26(6), 1496–1507.
https://doi.org/10.1016/j.chb.2010.05.014
Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why Minimal Guidance During Instruction
Does Not Work: An Analysis of the Failure of Constructivist, Discovery, Problem-Based,
Experiential, and Inquiry-Based Teaching. Educational Psychologist, 41(2), 75–86.
https://doi.org/10.1207/s15326985ep4102_1
Koschmann, T. (1996). Paradigm shifts and instructional technology: an introduction. In T.
Koschmann (Ed.), CSCL: theory and practice of an emerging paradigm (pp. 1–23).
Lawrence Erlbaum Associates, Inc.

55
Kwon, S. Y., & Cifuentes, L. (2007). Using Computers to Individually-generate vs.
Collaboratively-generate Concept Maps. Educational Technology & Society, 10(4), 269–
280.
Kwon, S. Y., & Cifuentes, L. (2009). The comparative effect of individually-constructed vs.
collaboratively-constructed computer-based concept maps. Computers and Education,
52(2), 365–375. https://doi.org/10.1016/j.compedu.2008.09.012
Kyndt, E., Raes, E., Lismont, B., Timmers, F., Cascallar, E., & Dochy, F. (2013). A metaanalysis of the effects of face-to-face cooperative learning. Do recent studies falsify or
verify earlier findings? Educational Research Review, 10, 133-149.
https://doi.org/10.1016/j.edurev.2013.02.002
Ladson‐Billings, G. (1995). But that’s just good teaching! The case for culturally relevant
pedagogy. Theory into Practice, 34(3), 159–165.
https://doi.org/10.1080/00405849509543675
Ladson-Billings, G. (1998). Just what is critical race theory and what’s it doing in a nice field
like education? International Journal of Qualitative Studies in Education, 11(1), 7–24.
https://doi.org/10.1080/095183998236863
Loes, C. N. (2022). The Effect of Collaborative Learning on Academic Motivation. Teaching
and Learning Inquiry, 10, 1–17. https://doi.org/10.20343/teachlearninqu.10.4
Mayer, R. E. (2012). Information processing. In K. R. Harris, S. Graham, & T. Urdan (eds.), Apa
educational psychology handbook.Vol. 1: Theories, constructs, and critical issues (pp85–
99). American Psycological Association.
Piaget, J. (1962). Play, dreams and imitation. Norton.

56
Radkowitsch, A., Vogel, F., & Fischer, F. (2020). Good for learning, bad for motivation? A
meta-analysis on the effects of computer-supported collaboration scripts. International
Journal of Computer-Supported Collaborative Learning, 15(1), 5–47.
https://doi.org/10.1007/s11412-020-09316-4
Rienties, B., Alcott, P., & Jindal-Snape, D. (2014). To let students self-select or not: That is the
question for teachers of culturally diverse groups. Journal of Studies in International
Education, 18(1), 64-83. doi:10.1177/1028315313513035
Samaha, N. V., & De Lisi, R. (2000). Peer Collaboration on a Nonverbal Reasoning Task by
Urban, Minority Students. The Journal of Experimental Education, 69(1), 5–21.
https://doi.org/10.1080/00220970009600646
Saqr, M., Elmoazen, R., Tedre, M., López-Pernas, S., & Hirsto, L. (2022). How well centrality
measures capture student achievement in computer-supported collaborative learning? – A
systematic review and meta-analysis. Educational Research Review, 35, 100437-.
https://doi.org/10.1016/j.edurev.2022.100437
Scager, K., Boonstra, J., Peeters, T., Vulperhorst, J., & Wiegant, F. (2016). Collaborative
learning in higher education: Evoking positive interdependence. CBE Life Sciences
Education, 15(4), ar69-. https://doi.org/10.1187/cbe.16-07-0219
Schunk, D. H. (2020). Learning Theories, an Educational Perspective (8th ed.). Boston, MA:
Pearson Education Inc.
Serpell, Z. N., Boykin, A. W., Madhere, S., & Nasim, A. (2006). The Significance of Contextual
Factors in African American Students’ Transfer of Learning. Journal of Black
Psychology, 32(4), 418–441. https://doi.org/10.1177/0095798406292466
Slavin, R. E. (2019). Educational Psychology: Theory and Practice. London: Pearson.

57
Slavin, R. E. (1980). Cooperative Learning. Review of Educational Research, 50(2), 315–342.
https://doi.org/10.3102/00346543050002315
Slavin, R. E. (1983). When does cooperative learning increase student achievement?
Psychological Bulletin, 94(3), 429–445. https://doi.org/10.1037/0033-2909.94.3.429
Stahl, G., & Hakkarainen, K. (2020). Theories of CSCL. International handbook of computersupported collaborative learning. Springer.
Tenenbaum, H. R., Winstone, N. E., Leman, P. J., & Avery, R. E. (2020). How Effective Is Peer
Interaction in Facilitating Learning? A Meta-Analysis. Journal of Educational
Psychology, 112(7), 1303–1319. https://doi.org/10.1037/edu0000436
Visschers-Pleijers, A. J. S. F., Dolmans, D. H. J. M., de Leng, B. A., Wolfhagen, I. H. A. P., &
van der Vleuten, C. P. M. (2006). Analysis of verbal interactions in tutorial groups: a
process study. Medical Education, 40(2), 129–137. https://doi.org/10.1111/j.1365-
2929.2005.02368.x
Wendt, J. L., & Rockinson-Szapkiw, A. (2014). The effect of online collaboration on middle
school student science misconceptions as an aspect of science literacy. Journal of
Research in Science Teaching, 51(9), 1103–1118. https://doi.org/10.1002/tea.21169
Whipple, W. R. (1987). Collaborative learning: recognizing it when we see it. AAHE bulletin, 4,
3–5.
Yosso, T. J. (2005). Whose culture has capital? A critical race theory discussion of community
cultural wealth. Race, Ethnicity and Education, 8(1), 69–91.
https://doi.org/10.1080/1361332052000341006

58
Tables
Table 1
Overall Average Effect of Collaborative Learning
Outcome NStud NS NES g 95% CI
low/high
τ2 I
2 Q
Achievement 7 10 34 –0.0267 –0.395/0.341 0.1733 74.73 133.00**
Note. Nstud = number of studies. Nsamples = number of samples. NES = number of effects. g =
Hedges’ g (average pooled effect). CI = confidence interval (low estimate / high estimate). ** p
< 0.0001

59
Table 2
Results of Moderator Analyses
Moderator NStud NS NES b (SE) g
95% CI
low/high
Publication status
Published 6 9 22 — –0.0277 –0.4472/ 0.3918
Unpublished 1 1 12 –0.0332 (0.671) –0.0332 –8.5636/ 8.4971
Ethnicity
% Black 7 10 34 –0.0001 (0.0069) — –0.0141/ 0.0139
% Latinx 5 8 18 –0.0013 (0.0027) — –0.0067/ 0.0041
Technology
No tech 4 7 20 — –0.0316 –0.4034/0.3402
Yes tech 6 7 14 0.0106 (0.0707) –0.0210 –0.4134/0.3714
Note. NStud = number of studies. NS = number of samples. NES = number of effects. b =
unstandardized regression slope coefficient (moderator effect). SE = standard error. g = Hedges’
g (average pooled effect). CI = confidence interval (low estimate / high estimate).

60
Figures
Figure 1
PRISMA Chart

61
Figure 2
Publication Bias Funnel Plot

62
Appendix A: List of Studies Meeting Inclusion Criteria, Alphabetically by Author
Author Year % Black % Latinx Technology N g v
Caldwell 2006 86.7 — No 40 –0.1037 0.1001
Caldwell 2006 86.7 — No 40 –0.2104 0.1006
Caldwell 2006 86.7 — No 40 0.3752 0.1018
Caldwell 2006 86.7 — No 40 0.2467 0.1008
Caldwell 2006 86.7 — No 40 –0.6461 0.1052
Caldwell 2006 86.7 — No 40 0.2556 0.1008
Caldwell 2006 86.7 — Yes 40 –0.3905 0.1019
Caldwell 2006 86.7 — Yes 40 –0.2434 0.1007
Caldwell 2006 86.7 — Yes 40 0.1253 0.1002
Caldwell 2006 86.7 — Yes 40 0.0695 0.1001
Caldwell 2006 86.7 — Yes 40 –0.2159 0.1006
Caldwell 2006 86.7 — Yes 40 0.135 0.1002
Kwon 2009 30.7 26.3 Yes 121 0.2647 0.0334
Kwon 2007 62 34 Yes 62 –0.5727 0.0672
Samaha 2000 10 78 No 40 0.0271 0.1003
Samaha 2000 10 78 No 50 –0.2162 0.0854
Samaha 2000 10 78 No 34 0.3793 0.1214
Samaha 2000 10 78 No 40 0.4668 0.103
Samaha 2000 10 78 No 50 0.3511 0.0861
Samaha 2000 10 78 No 34 0.6127 0.1248
Schacter 2000 18.3 15.5 Yes 24 –0.8082 0.0901
Schacter 2000 18.3 15.5 No 24 –0.9513 0.0928
Schacter 2000 18.3 15.5 No 24 –0.857 0.091
Schacter 2000 18.3 15.5 Yes 26 –0.237 0.0807

63
Author Year % Black % Latinx Technology N g v
Schacter 2000 18.3 15.5 No 26 –0.1906 0.0805
Schacter 2000 18.3 15.5 No 26 –0.4009 0.0817
Schacter 2000 18.3 15.5 Yes 26 0.1113 0.0803
Schacter 2000 18.3 15.5 No 26 –0.1818 0.0805
Schacter 2000 18.3 15.5 No 26 –0.1441 0.0803
Serpell 2006 55.6 — Yes 102 1.0907 0.0463
Serpell 2006 55.6 — Yes 102 0.5377 0.0419
Serpell 2006 55.6 — No 102 0.5784 0.0421
Serpell 2006 55.6 — No 102 0.7946 0.0436
Wendt 2014 35.3 2.4 Yes 90 –0.5569 0.0509

64
Appendix B: Coding Guide Prior to Coding
Coding column Coding title Coding options
Coder info
C-1 Date coded [text entry]
C-2 Coder [text entry]
Meta-analysis characteristics
M-1 Meta-analysis’ first author’s last name [text entry]
M-2 Meta-analysis Google Drive link [text entry]
Report characteristics
R-1 Report ID (reference #s) [text entry]
R-2 Article Google Drive link [text entry]
R-3 First author’s last name [text entry]
R-4 Year [text entry]
R-5 Title [text entry]
R-6 APA reference [text entry]
R-7 Publication type 1. Journal article
2. Book or book chapter
3. Dissertation
4. Master’s thesis
5. Policy report
6. Government report
7. Conference paper
8. Other
–99. Can’t tell
R-8 Data sources 1. Independent study
2. Regional/national data set
3. Other
–99. Can’t tell
R-9 Dataset name [text entry]; –99 missing/can’t
tell/not applicable
R-10 Data collection year indicated 0. No; 1. Yes
R-11 Year(s) data collected [text entry]; –99 missing/can’t
tell/not applicable

65
Coding column Coding title Coding options
R-12 On what page(s) did you find the data
source?
[text entry]
R-13 Overlapping datasets [text entry]; –99 No
Setting characteristics
S-1 Study Number 0. Single study
1. Study 1
2. Study 2
3. Study 3
S-2 Location [text entry]; –99 Missing/can’t
tell/not applicable
S-3 Region 1. Northeast
2. South
3. Midwest
4. West
5. National
–99. Can’t tell
S-4 On what page(s) did you find the
location?
[text entry]; –99 Missing/can’t
tell/not applicable
S-5 School level 1. Preschool
2. Elementary School: K–5
3. Middle School: 6–8
4. High School: 9–12
5. Undergraduate
6. Graduate School
7. Other (Specify)
–99. Can’t tell
S-6 Other school level (specify) [text entry]; –99 Missing/can’t
tell/not applicable
Tech Technology 0. No tech
1.Yes tech
Participant and sample characteristics
P-1 Sample 0. Overall sample; 1. Subgroup
P-2 Subgroup specification [text entry]; –99 Missing/can’t
tell/not applicable
P-3 Subgroup overlap 0. No; 1. Yes; –99. N/A
P-4 Subgroup overlap explanation [text entry]; –99 Missing/can’t
tell/not applicable

66
Coding column Coding title Coding options
P-5 Sample size (at start) [text entry]; –99 Missing/can’t
tell/not applicable
P-6 On what page(s) did you find the sample
size?
[text entry]; –99 Missing/can’t
tell/not applicable
P-7 Sample characteristics 1. Sample at Start
2. Analysis Sample
3. Both, but they are the same
4. Both, and they are not the same
5. Neither
–99. Can’t Tell/Not Applicable
P-8 Sample characteristics specification [text entry]; –99 Missing/can’t
tell/not applicable
P-9 %White [text entry]; –99 Missing/can’t
tell/not applicable
P-10 %Black [text entry]; –99 Missing/can’t
tell/not applicable
P-11 %Hispanic [text entry]; –99 Missing/can’t
tell/not applicable
P-12 %Asian or Pacific Islander [text entry]; –99 Missing/can’t
tell/not applicable
P-13 %Native American or American Indian [text entry]; –99 Missing/can’t
tell/not applicable
P-14 %Other [text entry]; –99 Missing/can’t
tell/not applicable
P-15 On what page(s) did you find the
racial/ethnic distribution?
[text entry]; –99 Missing/can’t
tell/not applicable
P-16 Grade level -1. Preschool 8. Grade 8
0. Kindergarten 9. Grade 9
1. Grade 1 10. Grade 10
2. Grade 2 11. Grade 11
3. Grade 3 12. Grade 12
4. Grade 4 13. Undergraduate
5. Grade 5 14. Graduate
6. Grade 6 15. Other (Specify)
7. Grade 7 –99. Can’t tell
P-17 On what page(s) did you find the grade
level?
[text entry]; –99 Missing/can’t
tell/not applicable
P-18 % Female [text entry]; –99 Missing/can’t
tell/not applicable

67
Coding column Coding title Coding options
P-19 On what page(s) did you find the %
female statistic?
[text entry]; –99 Missing/can’t
tell/not applicable
P-20 % Low income/economically
disadvantaged
[text entry]; –99 Missing/can’t
tell/not applicable
P-21 On what page(s) did you find the % lowincome statistic?
[text entry]; –99 Missing/can’t
tell/not applicable
P-22 % Special education [text entry]; –99 Missing/can’t
tell/not applicable
P-23 On what page(s) did you find the %
special education statistic?
[text entry]; –99 Missing/can’t
tell/not applicable
P-24 % English learners [text entry]; –99 Missing/can’t
tell/not applicable
P-25 On what page(s) did you find the %
English learner statistic?
[text entry]; –99 Missing/can’t
tell/not applicable
Influence/predictor measures
I-1 Report’s name for influence [text entry]
I-2 Influence definition [text entry]; –99 Missing/can’t
tell/not applicable
I-3 On what page(s) did you find the
influence definition?
[text entry]; –99 Missing/can’t
tell/not applicable
I-4 How is the influence measured? [text entry]; –99 Missing/can’t
tell/not applicable
I-5 On what page(s) did you find the
description of how the influence was
measured?
[text entry]; –99 Missing/can’t
tell/not applicable
I-6 Reliability 0. No; 1. Yes; –99. Unsure, N/A
I-7 Alpha coefficient (reliability) [text entry]; –99 Missing/can’t
tell/not applicable
I-8 Alpha coefficient from what source? 1. Data from this coded study
2. Data from the study for which
the survey was derived
–99. Unsure, N/A
I-9 On what page did you find the alpha
coefficient?
[text entry]; –99 Missing/can’t
tell/not applicable
I-10 How was the influence manipulated by
the researcher?
[text entry]; –99 Missing/can’t
tell/not applicable

68
Coding column Coding title Coding options
I-11 On what page(s) did you find the
description of how the researcher
manipulated the influence?
[text entry]; –99 Missing/can’t
tell/not applicable
Outcome measures
O-1 Outcome type 1. State standardized tests (statewide testing)
2. National standardized tests
(SAT/ACT/NAEP/PISA/TIMS
S)
3. GPA
4. Knowledge diagnostic test (e.g.,
researcher/instructor developed
test)
5. Other achievement
O-2 Outcome name [text entry]; –99 Missing/can’t
tell/not applicable
O-3 Outcome description [text entry]; –99 Missing/can’t
tell/not applicable
O-4 On what page(s) did you find the
description of the outcome?
[text entry]; –99 Missing/can’t
tell/not applicable
O-5 Domain of outcome 1. Mathematics
2. English Language Arts
3. Science
4. Social science
5. General academics
6. Other (specify)
O-6 Domain of outcome (specified) [text entry];
–99 Missing/can’t tell/not
applicable
O-7 What is the unit of analysis? 1. Student
2. Teacher
3. Classroom
4. School
5. Other (specify)
–99. Unsure/not applicable
O-8 Other unit of analysis [text entry]; –99 Missing/can’t
tell/not applicable

69
Coding column Coding title Coding options
O-9 Timing of influence and outcome
measure collection
1. Simultaneously
2. Longitudinally
–99. Unsure
O-10 Specify timing [text entry]; –99 Missing/can’t
tell/not applicable
O-11 On what page(s) did you find the timing
of data collection described?
[text entry]; –99 Missing/can’t
tell/not applicable
Effect sizes and research design
E-1 Sample size (for relationship/effect) [text entry]; –99 Missing/can’t
tell/not applicable
E-2 On what page(s) did you find the sample
size?
[text entry]; –99 Missing/can’t
tell/not applicable
E-3 Direction of relationship between
influence and outcome
0. Null/No Relationship
1. Positive
2. Negative
3. Mixed
–99. Unclear
Evidence of direction 1. Sign of correlation coefficient
2. Comparing means
3. Indication in text
–99. Can’t tell/unclear
On what page(s) did you find the
direction of the relationship?
[text entry]; –99 Missing/can’t
tell/not applicable
In what table did you find the direction of
the relationship?
[text entry]; –99 Missing/can’t
tell/not applicable
E-4 Type of Research Design 1. Descriptive study
2. Correlational study
3. One-group/single-group preexperimental design
4. Quasi-experiment
5. RCT/true experiment (2+
groups)
–99. Can’t tell
E-5 Is there a treatment group and a control
group?
0. No; 1. Yes; –99. Unclear
E-6 Is there random assignment to treatment
and control groups?
0. No; 1. Yes; –99. N/A
E-7 On what page did the researchers specify
random assignment?
[text entry]; –99 Missing/can’t
tell/not applicable

70
Coding column Coding title Coding options
E-8 Level of assignment 1. Student
2. Teacher
3. Classroom
4. School
5. Other (specify)
–99. Unsure/not applicable
E-9 Other level of assignment [text entry]; –99 Missing/can’t
tell/not applicable
E-10 Is there matching of treatment units to
comparison units?
0. No; 1. Yes; –99. N/A
E-11 Matching characteristics [text entry]; –99 Missing/can’t
tell/not applicable
E-12 On what page(s) did the researchers
indicate matching and matching
characteristics?
[text entry]; –99 Missing/can’t
tell/not applicable
E-13 Did the researchers report prior-influence
or pre-test statistics?
0. No; 1. Yes; –99. Can’t tell
E-14 On what page(s) did the researchers
report pre-test statistics?
[text entry]; –99 Missing/can’t
tell/not applicable
E-15 In what table did the researchers report
pre-test statistics?
[text entry]; –99 Missing/can’t
tell/not applicable
E-16 Regression 0. No; 1. Yes; –99. Can’t tell
E-17 On what page did the researchers specify
using regression?
[text entry]; –99 Missing/can’t
tell/not applicable
E-18 Multi-level/hierarchical modeling 0. No; 1. Yes; –99. Can’t tell
E-19 On what page did the researchers specify
multi-level modeling?
[text entry]; –99 Missing/can’t
tell/not applicable
Correlational studies
EC-1 What is the correlation coefficient? [text entry]; –99 Missing/can’t
tell/not applicable
EC-2 On what page did you find the correlation
coefficient?
[text entry];–99 Missing/can’t
tell/not applicable
EC-3 In what table did you find the correlation
coefficient?
[text entry]; –99 Missing/can’t
tell/not applicable
EC-4 What is Fisher’s Z? [text entry]; –99 Missing/can’t
tell/not applicable
EC-5 What is the variance (v)? [text entry]; –99 Missing/can’t
tell/not applicable
EC-6 Screenshot of effect size calculation [Image]; –99 Not Applicable

71
Coding column Coding title Coding options
EC-7 What is the t-statistic? [text entry]; –99 Missing/can’t
tell/not applicable
EC-8 On what page did you find the t-statistic? [text entry]; –99 Missing/can’t
tell/not applicable
EC-9 In what table did you find the t-statistic? [text entry];–99 Missing/can’t
tell/not applicable
EC-10 What is the effect size (r)? [text entry]; –99 Missing/can’t
tell/not applicable
EC-11 What is Fisher’s Z? [text entry]; –99 Missing/can’t
tell/not applicable
EC-12 What is the variance (v)? [text entry]; –99 Missing/can’t
tell/not applicable
EC-13 Screenshot of effect size calculation [Image]; –99 Not applicable
EC-14 What is the p-value of the t-test? [text entry]; –99 Missing/can’t
tell/not applicable
EC-15 On what page did you find the p-value? [text entry]; –99 Missing/can’t
tell/not applicable
EC-16 In what table did you find the p-value? [text entry]; –99 Missing/can’t
tell/not applicable
EC-17 What is the effect size (r)? [text entry]; –99 Missing/can’t
tell/not applicable
EC-18 What is Fisher’s Z? [text entry]; –99 Missing/can’t
tell/not applicable
EC-19 What is the variance (v)? [text entry]; –99 Missing/can’t
tell/not applicable
EC-20 Screenshot of effect size calculation [Image]; –99 Not applicable
How many groups are compared in the Ftest?
[text entry]; –99 Missing/can’t
tell/not applicable
What is the F-statistic of the F-test? [text entry]; –99 Missing/can’t
tell/not applicable
On what page did you find the F-statistic? [text entry]; –99 Missing/can’t
tell/not applicable
In what table did you find the F-statistic? [text entry]; –99 Missing/can’t
tell/not applicable
What is the respective t-statistic? [text entry]; –99 Missing/can’t
tell/not applicable
What is the effect size (r)? [text entry]; –99 Missing/can’t
tell/not applicable

72
Coding column Coding title Coding options
What is Fisher’s Z? [text entry]; –99 Missing/can’t
tell/not applicable
What is variance (v)? [text entry]; –99 Missing/can’t
tell/not applicable
Screenshot of effect size calculation [Image]; –99 Not applicable
EC-21 What is Mₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EC-22 On what page did you find Mₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EC-23 In what table did you find Mₜ? [text entry];–99 Missing/can’t
tell/not applicable
EC-24 What is SDₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EC-25 On what page did you find SDₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EC-26 In what table did you find SDₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EC-27 What is Nₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EC-28 On what page did you find Nₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EC-29 In what table did you find Nₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EC-30 What is M꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EC-31 On what page did you find M꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EC-32 In what table did you find M꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EC-33 What is SD꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EC-34 On what page did you find SD꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EC-35 In what table did you find SD꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EC-36 What is N꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EC-37 On what page did you find N꜀? [text entry]; –99 Missing/can’t
tell/not applicable

73
Coding column Coding title Coding options
EC-38 In what table did you find N꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EC-39 What is the effect size (r)? [text entry]; –99 Missing/can’t
tell/not applicable
EC-40 What is Fisher’s Z? [text entry]; –99 Missing/can’t
tell/not applicable
EC-41 What is the variance (v)? [text entry]; –99 Missing/can’t
tell/not applicable
EC-42 Screenshot of effect size calculation [Image]; –99 Not applicable
EC-43 Frequency of yes/favorable outcome for
“treatment” group
[text entry]; –99 Missing/can’t
tell/not applicable
EC-44 Frequency of no/unfavorable outcome for
“treatment” group
[text entry]; –99 Missing/can’t
tell/not applicable
EC-45 Frequency of yes/favorable outcome for
“control” group
[text entry]; –99 Missing/can’t
tell/not applicable
EC-46 Frequency of no/unfavorable outcome for
“control” group
[text entry]; –99 Missing/can’t
tell/not applicable
EC-47 On what page did you find the
contingency table/data for the
Contingency table?
[text entry]; –99 Missing/can’t
tell/not applicable
EC-48 In what table did you find the
contingency table/data for the
contingency table?
[text entry]; –99 Missing/can’t
tell/not applicable
EC-49 What is the effect size (r)? [text entry]; –99 Missing/can’t
tell/not applicable
EC-50 What is Fisher’s Z? [text entry]; –99 Missing/can’t
tell/not applicable
EC-51 What is the variance (v)? [text entry]; –99 Missing/can’t
tell/not applicable
EC-52 Screenshot of effect size calculation [Image]; –99 Not applicable
EC-53 r-index calculated? 0. No 1. Yes –99. N/A
EC-54 Effect size from original meta-analysis [text entry]; –99 Missing/can’t
tell/not applicable
EC-55 On what page did you find the effect size
from the original meta-analysis?
[text entry]; –99 Missing/can’t
tell/not applicable
EC-56 In what table did you find the effect size
from the original meta-analysis?
[text entry]; –99 Missing/can’t
tell/not applicable
Experiments (including quasi-experiments and randomized control trials)

74
Coding column Coding title Coding options
EE-1 What is Nₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EE-2 On what page did you find Nₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EE-3 In what table did you find Nₜ? [text entry]; –99 Missing/cannot
tell/not applicable
EE-4 What is N꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EE-5 On what page did you find N꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EE-6 In what table did you find N꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EE-7 What is Mₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EE-8 On what page did you find Mₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EE-9 In what table did you find Mₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EE-10 What is SDₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EE-11 On what page did you find SDₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EE-12 In what table did you find SDₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EE-13 What is M꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EE-14 On what page did you find M꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EE-15 In what table did you find M꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EE-16 What is SD꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EE-17 On what page did you find SD꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EE-18 In what table did you find SD꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EE-19 What is the effect size (d)? [text entry]; –99 Missing/can’t
tell/not applicable

75
Coding column Coding title Coding options
EE-20 What is the variance (v)? [text entry]; –99 Missing/can’t
tell/not applicable
EE-21 Screenshot of effect size calculation [Image, –99 N/A]
EE-22 What is SEₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EE-23 On what page did you find SEₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EE-24 In what table did you find SEₜ? [text entry]; –99 Missing/can’t
tell/not applicable
EE-25 What is SE꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EE-26 On what page did you find SE꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EE-27 In what table did you find SE꜀? [text entry]; –99 Missing/can’t
tell/not applicable
EE-28 What is the effect size (d)? [text entry]; –99 Missing/can’t
tell/not applicable
EE-29 What is the variance (v)? [text entry]; –99 Missing/can’t
tell/not applicable
EE-30 Screenshot of effect size calculation [Image]; –99 Not applicable
EE-31 What is the t-statistic? [text entry]; –99 Missing/can’t
tell/not applicable
EE-32 On what page did you find the t-statistic? [text entry]; –99 Missing/can’t
tell/not applicable
EE-33 In what table did you find the t-statistic? [text entry]; –99 Missing/can’t
tell/not applicable
EE-34 What is the effect size (d)? [text entry]; –99 Missing/can’t
tell/not applicable
EE-35 What is variance (v)? [text entry]; –99 Missing/can’t
tell/not applicable
EE-36 Screenshot of effect size calculation [Image]; –99 Not Applicable
EE-37 What is the p-value of the t-test? [text entry]; –99 Missing/can’t
tell/not applicable
EE-38 On what page did you find the p-value? [text entry]; –99 Missing/can’t
tell/not applicable
EE-39 In what table did you find the p-value? [text entry]; –99 Missing/can’t
tell/not applicable

76
Coding column Coding title Coding options
EE-40 What is the effect size (d)? [text entry]; –99 Missing/can’t
tell/not applicable
EE-41 What is variance (v)? [text entry]; –99 Missing/can’t
tell/not applicable
EE-42 Screenshot of effect size calculation [Image]; –99 Not applicable
EE-43 How many groups are compared in the Ftest?
[text entry]; –99 Missing/can’t
tell/not applicable
EE-44 What is the F-statistic of the F-test? [text entry]; –99 Missing/can’t
tell/not applicable
EE-45 On what page did you find the F-statistic? [text entry]; –99 Missing/can’t
tell/not applicable
EE-46 In what table did you find the F-statistic? [text entry]; –99 Missing/can’t
tell/not applicable
EE-47 What is the effect size (d)? [text entry]; –99 Missing/can’t
tell/not applicable
EE-48 What is variance (v)? [text entry]; –99 Missing/can’t
tell/not applicable
EE-49 Screenshot of effect size calculation [Image]; –99 Not applicable
EE-50 Frequency of yes/favorable outcome for
treatment group
[text entry]; –99 Missing/can’t
tell/not applicable
EE-51 Frequency of no/unfavorable outcome for
treatment group
[text entry]; –99 Missing/can’t
tell/not applicable
EE-52 Frequency of yes/favorable outcome for
control group
[text entry]; –99 Missing/can’t
tell/not applicable
EE-53 Frequency of no/unfavorable outcome for
control group
[text entry]; –99 Missing/can’t
tell/not applicable
EE-54 On what page did you find the
contingency table/data for the
contingency table?
[text entry]; –99 Missing/can’t
tell/not applicable
EE-55 In what table did you find the
contingency table/data for the
contingency table?
[text entry]; –99 Missing/can’t
tell/not applicable
EE-56 What is the effect size (d)? [text entry]; –99 Missing/can’t
tell/not applicable
EE-57 What is the variance (v)? [text entry]; –99 Missing/can’t
tell/not applicable
EE-58 Screenshot of effect size calculation [Image]; –99 Not applicable
EE-59 d-index calculated? 0. No; 1. Yes; –99. N/A

77
Coding column Coding title Coding options
EE-60 Effect size from original meta-analysis [text entry]; –99 Missing/can’t
tell/not applicable
EE-61 On what page did you find the effect size
from the original meta-analysis?
[text entry]; –99 Missing/can’t
tell/not applicable
EE-62 In what table did you find the effect size
from the original meta-analysis?
[text entry]; –99 Missing/can’t
tell/not applicable

Abstract (if available)

Abstract This research synthesis examines the impact of collaborative learning on academic achievement among Black and Latinx students in the United States, addressing a notable gap in existing literature. Grounded in John Hattie’s (2023) meta-analysis, which shows a significant positive effect of collaborative learning on student achievement globally, this study focuses on whether these benefits extend to historically marginalized groups in the United States. By limiting the analysis to racially diverse samples, the research aims to provide nuanced insights into the effectiveness of collaborative learning for Black and Latinx students, with additional exploration of technology’s role as a moderating factor. The literature review explores collaborative learning’s theoretical foundations, defining characteristics, and outcomes, emphasizing constructivism and culturally responsive pedagogy. It highlights the complexity of defining collaborative learning and underscores the importance of understanding its impact on Black and Latinx students’ academic achievement. Building on Hattie’s work, this meta-analysis uses a rigorous search strategy and systematic screening to synthesize research literature comprehensively. The study employs specific inclusion criteria and a meticulous data extraction process, coding various characteristics to ensure the meta-analysis’s validity and reliability. Random-effect modeling and meta-regression are used to explore effect sizes and heterogeneity. Seven studies meeting the inclusion criteria were incorporated into the final sample, yielding 10 samples and 34 effect sizes. The analysis revealed a statistically non-significant negative effect size for achievement (g = –0.0267), with substantial heterogeneity among studies (I² = 74.73%), suggesting significant variation in effect sizes across different contexts. Moderator analysis indicated no significant relationship between the percentage of Black or Latinx students and the effectiveness of collaborative learning, nor did the presence of technology significantly alter the impact of collaborative learning on academic achievement. These findings highlight the variability in the effectiveness of collaborative learning and suggest the need for further research to explore the impact of culturally responsive teaching practices and the nuanced role of technology in enhancing the effect of collaborative learning on educational outcomes for diverse student populations.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Learning together: a meta-analysis of the effect of cooperative learning on achievement among Black and Latinx students

PDF

The effects of technology on the mathematical achievement of Black and Latinx students

PDF

A meta-analysis of formative assessment on academic achievement among Black and Hispanic students

PDF

Unveiling the visible impact: a meta-analysis on inquiry-based teaching and the effects on Black and Latinx student achievement

PDF

The relationship between working memory and achievement among Black and Latinx students

PDF

The effects of inclusive education on academic achievement for special education and general education students of color: a meta-analysis

PDF

Collaborative instructional practice for student achievement: an evaluation study

PDF

The intersections of culturally responsive pedagogy and authentic teacher care in creating meaningful academic learning opportunities for students of color

PDF

Examining the impact of LETRS professional learning on student literacy outcomes: a quantitative analysis

PDF

A meta-analysis exploring the relationships between racial identity, ethnic identity, and Black students' positive self-perceptions in school

PDF

Reducing statistics anxiety among learners in online graduate research methods courses

PDF

Pooling historical information while addressing uncertainty and bias for power analysis: a Bayesian approach for designing single-level and multilevel studies

PDF

Examining perspectives of academic autonomy in community college students: a quantitative study

PDF

An examination of tri-level collaboration around student achievement using the gap analysis approach: central office leadership factors

PDF

Effects of teacher autonomy support with structure on marginalized urban student math achievement

PDF

An examination of tri-level collaboration around student achievement using the gap analysis approach: School site leadership factors

PDF

Sustainable intervention for learning gaps in middle school mathematics: a gap analysis

PDF

The racialized experiences of Black students in online master’s degree programs at a Predominately White Institution

PDF

Problems and solutions for school counselors supporting Black and Latinx students in the 21st century

PDF

Considerations for personalized professional learning at International Academy of South East Asia: a gap analysis

Asset Metadata

Creator Mooney, Sheree Cheng (author)

Core Title Bridging gaps, building futures: a meta-analysis of collaborative learning and achievement for Black and Latinx students

School Rossier School of Education

Degree Doctor of Education

Degree Program Educational Leadership

Degree Conferral Date 2024-08

Publication Date 07/12/2024

Defense Date 06/27/2024

Publisher Los Angeles, California (original), University of Southern California (original), University of Southern California. Libraries (digital)

Tag collaborative learning,computer-supported collaborative learning,constructivist learning,meta-analysis,OAI-PMH Harvest

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Kho, Adam (committee chair), Patall, Erika A. (committee chair), Lyons-Moore, Akilah (committee member)

Creator Email shereecmooney@gmail.com,smooney@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113997NAS

Unique identifier UC113997NAS

Identifier etd-MooneySher-13226.pdf (filename)

Legacy Identifier etd-MooneySher-13226

Document Type Dissertation

Format theses (aat)

Rights Mooney, Sheree Cheng

Internet Media Type application/pdf

Type texts

Source 20240712-usctheses-batch-1181 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu