Loading Add to favorites

by Chris Mattatall & Keith Power


snips 18

Students with learning disabilities benefit from high quality instruction, just as any other student would, provided that they: (1) are given ample opportunities to be engaged in the learning process, (2) have multiple opportunities to respond to instruction, (3) receive accurate and timely corrective feedback on their responses, and (4) are given opportunity to practice those corrected responses, even to the point of overlearning, if they are to master, retain, and apply their new knowledge (Greenwood et al., 2002; Harper & Maheady, 2007; Peladeau et al., 2003).  Furthermore, some researchers posit that students need to perform these new learning tasks with at least 80%-90% accuracy if they are to adequately benefit from the instruction (Brophy & Good, 1986).  Benefits of ensuring that students perform with high rates of accuracy include increased engagement and maintenance of attention (Reitz, 1994), as well as motivation that is developed through feelings of self-efficacy (Bandura, 2001; Covington, 2000; Ryan & Deci, 2000; Seifert, 2004).  In addition to these well-understood constructs, many cognitive and academic skills - as well as much of the material contained in grade-level curriculum - is cumulative and hierarchical by nature.  That is, higher order cognitive functions, and curriculum involving higher levels of skill, are dependent upon student’s prior knowledge or mastery of basic skills (Harper & Maheady, 2007; Kame’enui & Simmons, 1990).  In other words, in order for a student to benefit from the application of solving a math problem, they must first be able to read the problem and know how to apply the necessary computation in order to solve it.  Establishing the necessary foundational requisite skills requires multiple applications of instruction, correct strategy use and practice, and corrective feedback with subsequent corrective practice.

Unfortunately, in relation to students with learning disabilities, research has shown that these students often receive less, not more, opportunities than their peers within general education settings, to receive the instruction, corrective feedback, and practice required to learn and master new material (Greenwood, Maheady, & Carta, 2002; McLeskey & Waldron, 2007).  There are many reasons cited in research that attempt to explain this phenomenon.  Including students with varying abilities and needs in the general education classroom and ensuring that each student within the class is meeting their learning potential has never been an easy task.  Teachers tend to move along through the curriculum even when students have not yet mastered foundational skills or material because of the perceived or stated requirement to cover the entire grade-level curriculum in one school year (Mattatall, 2011).  Teachers also find it difficult to differentiate instruction effectively across varying levels of need and to provide ample time for additional practice and corrective feedback (Mathes & Babyak, 2001; Timperley & Parr, 2007).  Increasing the range of academic and social behaviours that can be reasonably accommodated in the general education setting will require new ways of thinking, planning, designing access to curriculum, and teaching (McLeskey & Waldron, 2007).   However, it is unlikely that any new approach will be sustained unless it is easy to use, easily managed by students and teachers, and fits within the daily routine of the classroom setting (Gersten & Bregelman, 1996; Gersten et al., 2000). One such approach that has been shown to fit naturally into the ebb and flow of the general education classroom, and for which there is substantial empirical evidence of effectiveness for students with learning disabilities, is that of peer-mediated, or peer-assisted learning methods.

In the following video, Susan Craig from the Thunder Bay Catholic District School Board discusses the PALS program and its implementation in her school board:

Click here to view the transcript of this video.

Peer-Mediated Learning Approaches

Description of the Approach

snips 12

Peer-mediated learning approaches are known by several names: peer assisted learning, peer tutoring, and class-wide peer tutoring, to name the most common.  For this review, the term peer-mediated learning will be used.  Peer-mediated learning is a classroom-based practice where students work in pairs to complete activities.   One student (tutee) provides overt responses while the other student (tutor) provides immediate corrective feedback, clarification of concepts, or further instruction.  In some arrangements roles are reciprocal and students switch roles while completing activities together.  Activities that guide the peer interactions are often highly structured, and can even be somewhat scripted in some cases, especially for sessions involving young students.  In most cases, peer-mediated learning sessions are short in duration (e.g. 10 to 30 minutes) and involve focused work tasks around a teacher-determined activity, goal or skill. During peer-mediated learning sessions classroom teachers move throughout the classroom and are able to observe dyads at work, and provide frequent corrective feedback and instruction as needed. Students are placed in pairs based on the teacher’s judgement and assessment of their current abilities, with the requirement that one of the students in the dyad has a higher skill level than the other.  The idea here is that the student who currently possesses a higher level of skill will be able to model appropriate task completion, as well as provide adequate corrective feedback on the skill or activity in progress.   However, this does not mean that students with the higher skill level will not benefit from the approach.  Depending on how the classroom teacher arranges the assigned work and objective of the activity, both students in the dyad can benefit from the paired arrangement.

Peer-mediated learning approaches have been implemented at all grade levels, and across many subject areas (e.g. reading, math, science, social studies).   They have also been implemented within and across grade levels (e.g. same-age peers, older to younger students), within various settings (e.g., classrooms, alternate arrangements, group homes), and in various formats (e.g., triads, small group, class-wide dyads).  For this review, only research that examines the effectiveness of peer-mediated learning approaches for same-age/same-grade students in paired arrangements will be considered.  Same-age pairings are by far the most frequently used format; perhaps because it is much easier for individual classroom teachers to implement this arrangement in their own classrooms.

Objective and Benefits of the Approach

Peer-mediated learning approaches are based upon the notion that students need frequent opportunities to respond (demonstrate their understanding or skill) to instruction and to receive immediate corrective feedback, guidance and praise, when needed.   Peer-mediated learning has been shown to: (1) increase the number of opportunities for students to respond to instruction, (2) increase the amount of time on task, (3) increase the amount of feedback a student receives on their responses, and (4) increase the immediacy of the feedback (Bowman-Perrott et al., 2013).  Each of these components have been empirically linked with increased academic achievement for students (Bowman-Perrott et al., 2013; Greenwood et al., 1992; Maheady et al., 1988).

Parameters for the Review of the Research

To identify relevant studies published in scholarly peer-reviewed journals, a university-based library search engine, which included multiple online databases, as well as Google Scholar, were searched for abstracts including the years 2000 to 2014.  A variety of parameters were used to find as many studies as possible on peer-mediated learning approaches in K-12 schools.  This search yielded a large number of studies (over 1000).  Further parameters were placed on the search in order to further refine the obtained results.  Studies within the studies contained in this review were carried out at all grade levels (K-12), across multiple subject areas, and involving multiple moderators (gender, socio economic status, disability status, etc.).

Specific studies not included in this review which also involved substantial empirical data on the effectiveness of peer-mediated approaches for students with learning disabilities, include those studies that employ a developed program such as Peer Assisted Learning Strategies (Fuchs, Fuchs, Mathes, & Simmons, 1997) or Corrective Reading Program (Englemann, Hanner, & Johnson, 1989).  It may be worth noting that the PALS program (Fuchs et al., 1997) out of Peabody College of Vanderbilt University is modelled after and upon one of the earliest models of peer-mediated learning approaches (Classwide Peer Tutoring: CWPT; Delquadri, Greenwood, Whorton, Carta, & Hall, 1986) developed at the Juniper Gardens Children’s Project of the University of Kansas.  These two models, as well as one notable twelve-year investigation on the effects of peer-mediated learning approaches (Greenwood et al., 1989), contain some of the most extensive data on the effectiveness of peer-mediated approaches (Maheady, Mallette, & Harper, 2006).


The meta-analysis by Bowman-Perrott et al. (2013) examined effects of peer-mediated learning across 23 single-case research experiments for 511 students with learning disabilities in Grades 1-12. The overall effect size for students with learning disabilities was 0.75, indicating that moderate to large academic benefits can be attributed to peer mediation.  Students without learning disabilities had a lower effect size of 0.65.  The overall effect size of peer-mediated learning is highest for vocabulary learning (ES = 0.92), followed by mathematics (ES = 0.86), reading (ES = 0.77), spelling (ES = 0.74) and social studies (ES = 0.57).  Effect sizes were slightly higher for middle and secondary students (ES=0.74) than they were for elementary students (ES = 0.69).

Similar findings are reported in Scruggs, Mastropieri & Marshak (2012) experimental study.  Students (n=157) in ten middle school social studies classes were assigned at random to a traditional instruction condition, or an experimental condition involving classwide peer mediation involving 18 weeks of instruction that covered seven units of History.  All students were assigned the same pretest-posttest measures.   The experimental condition was more effective than the traditional condition in facilitating learning gains over the 18 week period.  Once again, students with learning disabilities reported higher effect sizes (ES =0.74), than those experienced by students without learning disbilities (ES = 0.57).

Kunsch, Jitenra & Sood (2007) report on 17 experimental studies in K-12 classrooms using random assignment of participants to treatment and control conditions.  This meta-analysis examines the effects of peer-mediated instruction in mathematics for students with learning disabilities and those at-risk for mathematics disabilities.  A total of 460 students with learning disabilities in mathematics and 643 students at-risk of math disabilities, along with 627 students without disabilities, were participants.  The overall mean effect size across all studies for students with learning disabilities and those at-risk of learning disabilities in mathematics was moderate (ES = 0.47).  However, the resulting effect size is small (ES = 0.21) for those with identified LD in math, and small to moderate (ES = 0.66) for students at-risk for learning disabilities.

The Okilwa & Shelby (2010) review of 12 studies reported on the effects of peer-mediated learning approaches for middle and high school participants (Grades 6-12) with learning disabilities in two classroom-type settings; special education classrooms (7 studies) as well as general education classrooms (5 studies).  Subject areas include english, social studies, history, math, and science.   Noteable are the results reported from the special education classroom setting.  Students with learning disabilities were paired together to complete activities in a peer tutoring format, with one student having a higher skill level than the other.  Three of the studies reported moderate effect sizes (ES = 0.40, 0.63, and 0.68) in Grade 9-12 mathematics, Grade 10 history, and Grade 7/8 social studies respectively.  Students with learning disabilities in general education classrooms were paired with students without learning disabilities in 4 of the 5 studies.  Fuchs, Fuchs, & Kazdan report a small to moderate effect size of 0.34 on assessment of reading comprehension after 16 weeks of dyad reading assignments (5 times every two weeks).

In the other study, which involved a randomized control trial, Mastropieri et al. (2006) randomly assigned 13 Grade 8 science classes to use classwide peer-mediated approaches to teach science curriculum.  Their study reported that students with LDs obtained significantly higher pre to posttest scores on science unit tests as the result of the peer-mediated arrangement compared to students in the control condition.  As well, student with LDs in peer-mediated condition also outperformed students with LDs in the control condition on the state-wide high stakes test at the end of the school year, though the results were not statistically significant.

Implementation & Recommended Practice


snips 5

The studies used in this review reported small to moderate mean effect sizes across all of their studies (ES = 0.21 to 0.75) for students with learning disabilities who were involved in peer-mediated learning approaches.  In most cases effect sizes were in the moderate range.  The lowest effect size (ES = 0.21: Kunsch, Jitendra, & Sood, 2007) were obtained in research studies involving “complex mathematics content” (p.8).

Regardless of grade level, students in peer-mediated conditions either made significant personal gains in achievement or outperformed peers in control conditions.  Gains were shown in all subject areas that used peer-mediated approaches.   Likewise, greater academic gains were achieved by students without learning disabilities who were engaged in peer-mediated approaches compared to their peers in classrooms not using this approach.  The meta-analysis by Bowman-Perrott et al. (2013) showed that no difference in student outcomes were evident between studies with a large dosage of peer-mediated work (over 1480 minutes: ES = 0.75) and low dosage of work (as few as 280 minutes: ES = 0.75) suggesting that the nature of the work (e.g., increased opportunities to respond, error correction, and feedback) was as important as the number of opportunities for students to work in pairs.  However, as Kunsch et al. (2007)  show, the more complex the material used with students in peer-mediated formats the more time is required to see positive effects in student achievement.

The research for this review demonstrated that when classroom teachers employed peer-mediated approaches to review, practice, clarify concepts, and provide opportunities for students to receive immediate corrective feedback and subsequent practice on those corrections, that positive effects on academic achievement occurred for students with learning disabilities even after short periods of time (i.e.; 8 total hours of peer learning sessions).  The success of this approach is likely due to the availability for students with learning disabilities to have multiple opportunities during classtime to respond to instruction, and receive immediate and individualized corrective and affirmative feedback on their understanding or demonstration of the material.  Providing students with these types of arrangements has been shown in previous research to be empirically linked to academic achievement for all students. Results from the review of the literature in this paper on peer-mediated approaches for students with learning disabilities appear to confirm this.

Implementing peer-mediated learning approaches in K-12 classrooms involves the following recommendations.

            1.  Effective sessions are well planned and highly structured, especially for young children.  Peer-mediated sessions are generally short in length (15 to 30 minutes) and involve distinct objectives and learning goals for each session.  Students in effective peer-mediated learning sessions are well aware of the learning objectives of each session.  It is recommended that classroom teachers spend some time training students first how to work in pairs, coach or provide effective corrective feedback to one another in their peer-mediated sessions.  

            2.  Students are paired appropriately.  In most cases student dyads involve students of different levels of skill; one higher than the other.  This allows for the more skilled student to provide appropriate modelling, corrective feedback, and necessary additional coaching or instruction.   However, too much of a gap between student skill is not recommended. Teacher discernment is needed to match students who will work well together. 

            3.  Effective peer-mediated learning sessions work best when teachers ensure that students have appropriate prior knowledge to complete the tasks required of them in paired lessons.  Peer-mediated approaches are best employed for the purpose of providing additional practice, review, engagement, corrective feedback, and opportunities to practice corrected work.

            4.  Peer-mediated approaches are to be considered supplementary in nature.  Classroom teachers provide initial and foundational explicit instruction, and peer-mediated approaches can be used to supplement and support the initial instruction of teachers, not replace it.


As discussed in this summary, peer-mediated learning approaches is an alternate term for peer assisted learning. The Peer Assisted Learning Strategies (PALS) program is a tier 1 intervention for reading difficulties (as identified in the Ministry of Educations’ document, learning for all) and the following list of resources includes videos, resource documents, and articles on PALS:

Click here to visit the LD@school website and access a video entitled, "The Tiered Approach".

Click here to visit the Ontario Ministry of Education's website and watch a video on PALS

Click here to access an overview (in PDF) on PALS, developed by OISE/UT

Click here to visit the Reading Rockets website and access on article entitled, "Best Practice for ELLs: Peer-Assisted Learning"

Click here to visit the Ontario Education Research Exchange website and read a program brief on PALS

Click here to view a PowerPoint presentation on PALS


Bandura, A. (2001). Social cognitive theory: An agentic perspective. Annual Review of Psychology, 52, 1-26.  doi: 10.1146/annurev.psych.52.1.1

Bowman-Perrott, L., Davis, H., Vannest, K., Williams, L., Greenwood, C., & Parker, R. (2013). Academic Benefits of Peer Tutoring: A Meta-Analytic Review of Single-Case Research. School Psychology Review, 42(1), 39-55.

Brophy, J., & Good, T. (1986). Teacher behavior and student achievement. In M. C. Wittrock (Ed.), Handbook of research on teaching (3rd ed., pp. 328–375). New York: Macmillan.

Chard, D. J.,  Vaughn, S., & Tyler, B-J. (2002). A Synthesis of research on effective interventions for building reading fluency with elementary students with learning disabilities. Journal of Learning Disabilities, 35(5), 386-406. doi:10.1177/00222194020350050101

Covington, M. (2000). Goal theory, motivation, and school achievement: An integrative review. Annual Review of Psychology, 51, 171-200.  doi: 10.1146/annurev.psych.51.1.171

Gersten, R. & Brengelman, S. U. (1996). The quest to translate research into classroom practice: The emerging knowledge base. Remedial and Special Education,17, 67–74. doi:10.1177/074193259601700202

Gersten, R., Chard, D., & Baker, S. (2000). Factors enhancing sustaineduse of research-based instructional practices. Journal of Learning Disabilities, 33, 445–457. doi:10.1177/002221940003300505

Greenwood, C. R., Carta, J. J., Hart, B., Kamps, D., Terry, B., Arreaga-Mayer, C., et al. (1992). Out of the laboratory and into the community: 26 years of applied behavior analysis at the Juniper Gardens Children’s Project. The American Psychologist, 47 (11), 1464–1474.  doi:10.1037/0003-066X.47.11.1464

Greenwood, C., Maheady, L., & Carta, J. (2002). Classwide peer tutoring programs. In M. Shinn, H. Walker, & G. Stoner (Eds.), Interventions for academic and behavior problems II: Preventive and remedial approaches (pp. 611–649). Washington, DC: National Association of School Psychologists.

Harper, G. F.  & Maheady, L. (2007). Peer-mediatated teaching and students with learning disabilities. Intervention in School and Clinic, 43(2), 101-107.  doi:10.1177/10534512070430020101

Kame’enui, E. J., & Simmons, D. C. (1990). Designing instructional strategies: The prevention of academic learning problems. Upper Saddle River, NJ: Merrill/Prentice Hall.

Kunsch, C. A., Jitendra, A. K., & Sood, S. (2007). The Effects of Peer-Mediated Instruction in Mathematics for Students   with Learning Problems: A Research Synthesis. Learning Disabilities Research & Practice, 22(1), 1-12. doi:10.1111/ j.1540-5826.2007.00226.x

Mathes, P. G., & Babyak, A. E. (2001). Peer-Assisted Learning Strategies for first-grade readers with and without additional small-group mini-skills lessons. Learning Disabilities Research and Practice, 16, 28–44. doi: 10.1111/0938-8982.00004

Mattatall, C. A. (2011). A study of how one ontario school board used peer assisted learning strategies and data-informed decision-making to address reading failure at Grade one  (Unpublished doctoral dissertation). Queen’s University, Kingston, Ontario, Canada.

McLeskey, J. & Waldron, N.L. (2007). Making differences ordinary in inclusive classrooms. Intervention in School & Clinic, 42(3), 162-168. doi: 10.1177/10534512070420030501

Okilwa , N.S. & Shelby, L. (2010) The Effects of Peer Tutoring on Academic Performance of Students With Disabilities in Grades 6 Through 12: A Synthesis of the Literature. Remedial and Special Education,  31, 450-463, doi:10.1177/0741 932509355991

Péladeau, N., Forget, J., & Gagné, F. (2003). Effect of paced and unpacedpractice on skill application and retention: How much isenough? American Educational Research Journal, 40(3), 769–801.   doi:10.3102/00028312040003769

Reitz, A. (1994). Implementing comprehensive classroom-based programs for students with emotional and behavioral problems. Education & Treatment of Children, 17(3), 312–331. Retrieved from http://www.educationandtreatmentofchildren.net

Ryan, R. & Deci, E. (2000). Intrinsic and extrinsic motivation: Classic definitions and new directions. Contemporary Educational Psychology, 25, 54-67.  doi: 10.1006/ceps.1999.1020

Scruggs, T. E., Mastropieri, M. A., & Marshak, L. (2012). Peer-Mediated Instruction in Inclusive Secondary Social Studies Learning: Direct and Indirect Learning Effects. Learning Disabilities Research & Practice, 27(1), 12-20. doi:10.1111/j.1540-5826.2011.00346.x

Seifert, T. (2004). Understanding student motivation. Educational Research, 46, 137-149.  doi: 10.1080/0013188042000222421

Timperley, H. S. & Parr, J. M. (2007). Closing the achievement gap through evidence-based inquiry at multiple levels of the education system. Journal of Advanced Academics, 19, 90-115. doi: 10.4219/jaa-2007-706

A special thanks to the Thunder Bay Catholic District School Board for allowing us to observe and take pictures of a PALS class in action!

horizontal line teal

Dr. Mattatall is an assistant professor of special education in the Faculty of Education at Memorial University of Newfoundland.  His research interests include peer-mediated learning approaches to address reading and mathematics disabilities in primary and elementary school children, as well as the effects of data-informed collaboration to help educators meet the needs of all students within schools.

Mr. Power is a doctoral student in the Faculty of Education at Memorial University of Newfoundland.  His research interests include childhood development, teacher education and development, and comprehensive school health.