UPDATE: Student Grouping: What is effective?

This is an updated version of an earlier post, based on new research I included. The decision-tree has been updated! You can find a Google Doc version of this here.

Student Grouping: What is Effective?

How do we leverage student grouping to best promote achievement?

This is a question teachers and administrators ask themselves almost daily. Unfortunately, there are few clear or easy answers. But we can draw out a few general principles from recent research and other sources of knowledge that may help to inform our instructional practice.

It’s important to acknowledge there’s often a steady pressure on teachers to utilize group work. And for some teachers, grouping students by ability can make serving a wide disparity of different levels of students more manageable.

But there’s an often unstated assumption: group work is inherently superior to whole class or independent learning. But is group work always better than other modes of learning?

Tom Bennett, a British behavioral specialist, argues in an article in American Educator, “Group Work for the Good,” that there is little research to suggest group work is better for academic learning. Bennett cautions teachers to only “use group work when you feel it is appropriate to the task you want your students to achieve, and at no other time.

OK, but what are the times when group work is appropriate?

When I first investigated this, a particular passage from Douglas Fisher and Nancy Frey, in their book, Guided Instruction: How to Develop Confident and Successful Learners, struck me:

“. . . an understanding of memory systems has profound implications for instruction, which include creating systematic and intentional scaffolds of students’ understanding rather than leaving them alone to discover information independently. That’s not to say that students should not work together in collaborative learning; they should. We have argued for productive group work in which students interact with one another and generate ideas to produce individual works (Frey et al., 2009). But this work must center on the consolidation and application of content that students already know. It’s neither the time nor the place to introduce new information. Doing so would overload the working memory system and fail to ensure learning(Bold added).

In other words, Fisher and Frey suggest that new concepts and information should not be introduced during group work. Group work should instead be used for reinforcing, consolidating, and applying information students have already been exposed to.

But I later came across another study by Paul Kirschner, John Sweller, Femke Kirschner, and Jimmy Zambrano, “From Cognitive Load Theory to Collaborative Cognitive Load Theory,” that contradicts this. Instead, Kirschner et al. suggest that working as a group creates a collective working memory, and that therefore group tasks should be more complex.

They state, “… learning in a team is more effective than individual learning if the complexity of the to-be-learned material is so high that it exceeds the limits of each individual learner’s working memory.” They furthermore suggest that greater complexity will make the task more engaging for the group: “Collaboration will occur when the task is complex enough to justify the extra time and effort involved in collaborating with others.”

Therefore when assigning group work, ensure that the task is complex enough to warrant collaborative effort.

So we’re engaging groups of kids in complex tasks. Now how do we ensure students are productive during the times when they do work as a group? Here it can be instructive to look at some of the analysis coming out of the business sector. Fostering productive teams, after all, is critical to the success or failure of many modern businesses.

One finding from the business realm that will make immediate sense to educators is that creating a context that fosters shared identity promotes productivity. You can read more about this research in “Spaces the Signal Identity Improve Workplace Productivity,” in the Journal of Personnel Psychology. We know that giving our students a sense of belonging and recognizing who they are and what they bring is critical to fostering a positive school community. But it’s good to know that it also can improve group performance.

Another finding is that how a team communicates is what determines its effectiveness. As presented in an article, “The New Science of Building Great Teams” in Harvard Business Review, effective teams communicate more equitably and with higher engagement. And even more critically for consideration in a school context, socialization outside of formal meeting time has a huge influence on team effectiveness. What this means for educators is that fostering effective group work requires time and training. Furthermore, as described in a passage “Group Dynamics for Teams” by Daniel Levi, this training requires norming, socialization, and building cooperative skills. Educators know that many of our students struggle with social skills and working productively together. These skills must be taught and developed.

Similarly, moving into research from higher education, in “What makes a ‘good group’? Exploring the characteristics and performance of undergraduate student groups” the authors argue that if “team working skills” are “important as a learning outcome, they must be assessed directly alongside the task output.” In other words, if a teacher is going to utilize group work for a task, they must establish explicit learning objectives for the skills practiced in the group work itself, not only for the content of the task. This again reinforces the idea that when we do use group work, we must do so strategically.

This builds off of Robert Slavin’s review of educational research, as outlined in an ACSD article, “Cooperative Learning and Student Achievement,” which suggests that not only must effective group work have group goals and rewards, but also must hold each individual accountable for their contribution. Group work which incorporates only one aspect of those two critical components (group goals and individual accountability) demonstrates little benefit to learning, whereas group learning which incorporates both is far more effective.

Even adult research teams require training and practice to develop intrapersonal awareness, foster shared norms, and to understand that conflict is normal, as suggested by a paper “Creating and maintaining high-performing collaborative research teams: the importance of diversity and interpersonal skills.” The authors further suggest that fostering diverse teams is essential to productivity.

This latter insight, that diverse teams are more productive, may be one of the most useful within a classroom context. Various studies, as presented in an article on Harvard Business Review, “Why Diverse Teams Are Smarter” by David Rock and Heidi Grant, suggest that ethnic and racial diversity makes for more effective, deliberative, and innovative teams. This is an important consideration for teachers when forming groups.

A recent study by P. Karen Murphy et al., “Exploring the influence of homogeneous versus heterogeneous grouping on students’ text-based discussions and comprehension” provides useful guidance for teachers in deciding between homogenous vs. heterogenous grouping:

“. . . teachers’ goals and expectations for small-group discussions should guide their decision to compose the groups homogeneously or heterogeneously. For example, if teachers desire to focus on enhancing students’ basic comprehension or if they desire to support students’ engagement in the discussion, they may find that grouping the students homogeneously is more advantageous for low-ability students. Alternatively, teachers should employ heterogeneous ability grouping if their focus is on building students’ high-level comprehension of the text.”

In other words, group homogeneously to engage low-skilled students; group heterogeneously to deepen comprehension.

However, it’s important to note that research on homogeneous vs. heterogeneous grouping is mostly unclear. Ultimately, how a teacher chooses to group students must be strategic and based on the task and learning outcomes. But overall findings seem to suggest that our default should be mixing students of different backgrounds and ability.

A synthesis of findings on effective group work

Ok, so we’ve reviewed a fair amount of information on grouping. Let’s summarize what we have so far:


  • Use group work only when it is necessary to achieve the task you are planning
  • When assigning group work, ensure that the task is complex enough to warrant collaborative effort.
  • Create a classroom and school environment that fosters a shared identity
  • Provide norming, time for socialization, and training in the cooperative skills students will require to work productively as a team
  • Set explicit learning targets for group work skills when engaging in a group task, while holding each individual student accountable for their work within the group
  • Group students heterogeneously to promote greater critical thinking and creativity



Bennett, T. (2015). Group Work for the Good. American Educator. Retrieved from http://www.aft.org/ae/spring2015/bennett

Channon, S.B. (2017). What makes a ‘good group’? Exploring the characteristics and performance of undergraduate student groups. SpringerLink. Retrieved from https://link.springer.com/article/10.1007/s10459-016-9680-y

Cheruvelil, K. S., Soranno, P. A., Weathers, K. C., Hanson, P. C., Goring, S. J., Filstrup, C. T. and Read, E. K. (2014), Creating and maintaining high-performing collaborative research teams: the importance of diversity and interpersonal skills. Frontiers in Ecology and the Environment. 12: 31–38. doi:10.1890/130001. Retrieved from http://onlinelibrary.wiley.com/doi/10.1890/130001/abstract

Greenaway, K.H., Hannibal A. Thai, S. Haslam, A., and Murphy, S.C. (2016). Spaces That Signal Identity Improve Workplace Productivity. Journal of Personnel Psychology. 15(1), 35–43. Retrieved from https://www.researchgate.net/publication/301277968_Spaces_That_Signal_Identity_Improve_Workplace_Productivity

Fisher, D. & Frey, N. (2010). Scaffolds for Learning: The Key to Guided Instruction. ASCD. Retrieved from http://www.ascd.org/publications/books/111017/chapters/Scaffolds-for-Learning@-The-Key-to-Guided-Instruction.aspx

Kirschner, P.A., Sweller, J., Kirschner, F., and Zambrano, J. (2018). From Cognitive Load Theory to Collaborative Cognitive Load Theory. International Journal of Computer-Supported Collaborative Learning. Retrieved from https://link.springer.com/article/10.1007%2Fs11412-018-9277-y

Levi, D. (2001). Group Dynamics for Teams. Sage Publications, 322 pp. Retrieved from http://citeseerx.ist.psu.edu/viewdoc/download?doi=

Murphy, P.K., Greene, J.A., Firetto, C.M., Li, M., Lobczowski, N.G., Duke, R.F., Wei, L., Croninger, R.M.V. (2017). Exploring the influence of homogeneous versus heterogeneous grouping on students’ text-based discussions and comprehension. Contemporary Educational Psychology. 51, 336-355 Retrieved from: http://www.sciencedirect.com/science/article/pii/S0361476X17302540

Pentland, A.S. (2012). The New Science of Building Great Teams. Harvard Business Review. Retrieved from https://hbr.org/2012/04/the-new-science-of-building-great-teams

Rock, D. and Grant, H. (2016.). Why Diverse Teams Are Smarter. Harvard Business Review. Retrieved from https://hbr.org/2016/11/why-diverse-teams-are-smarter

Slavin, R. (1988). Cooperative Learning and Student Achievement. ASCD. Retrieved from http://www.ascd.org/ASCD/pdf/journals/ed_lead/el_198810_slavin.pdf

Wang, Z. (2013). Effects of heterogenous and homogenous grouping on student learning. Chapel Hill. Retrieved from https://cdr.lib.unc.edu/indexablecontent/uuid:ac391807-1cca-447e-801d-d65183945ad0

Yee, V. (2013.). Grouping Students by Ability Regains Favor With Educators. NY Times. Retrieved from https://www.nytimes.com/2013/06/10/education/grouping-students-by-ability-regains-favor-with-educators.html

Group Work Decision Tree

Copy of Group Work Decision Tree - Page 1

Student Grouping: What is Effective? by Mark Anderson is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

A Revision on Group Work: Assign complex—rather than simple—tasks


Paul Kirschner, John Sweller, Femke Kirschner, and Jimmy Zambrano published a recent paper in the International Journal of Computer-Supported Collaborative Learning titled, “From Cognitive Load Theory to Collaborative Cognitive Load Theory,” in which they make the case that cognitive load theory needs to be reconsidered from a different frame for group tasks.

I found it interesting because I had created a synthesis on group learning a while back (read all about it here), and much of Kirschner et al.’s synthesis aligns with much of what I found, such as that group work requires the development of collaborative skills, the importance of clear, accountable roles and expectations for group tasks, or that heterogenous grouping should be our default when creating groups.

However, there was one key piece that seemed to contradict a finding I had drawn from Doug Fisher and Nancy Frey’s book, “Guided Instruction,” which is that new concepts and information should not be introduced during group work, as this would overload students’ working memory.

Kirschner et al. instead suggest that working as a group creates a collective working memory, and that therefore group tasks should be more complex. They state, “… learning in a team is more effective than individual learning if the complexity of the to-be-learned material is so high that it exceeds the limits of each individual learner’s working memory.” They furthermore suggest that greater complexity will make the task more engaging for the group: “Collaboration will occur when the task is complex enough to justify the extra time and effort involved in collaborating with others.”

However, they also caution that “In terms of cognitive load, if learners have not acquired [task-specific collaboration] skills prior to beginning on the collaborative task, the load induced here could be so high as to hinder collaborative learning.” This agrees with what I also found in my synthesis, which is that fostering effective group work requires time and training, with explicit learning objectives for the skills practiced in group specific tasks. As Kirschner et al. puts it, “We may need to be taught how to communicate and coordinate carrying out complex tasks in order to optimize transactive activities and construct better knowledge and skill schemas (Zambrano et al. 2018).”

Kirschner et al. also highlight the different considerations we need to make for students based on what they already know. In fact, they seem to suggest that students with low levels of domain-specific knowledge can benefit the most from engaging in group tasks, whereas students with higher level of expertise may have their learning hindered by the additional elements introduced by collaboration. Here’s two relevant quotes on this:

“When teams are composed of learners with a low level of domain-specific knowledge, these novices need to be involved in cognitively demanding search-based problem solving, whereas when they are knowledgeable, this is not the case as the learners can probably deal with the problems using their available knowledge base. Also, when teams are composed of learners with a low level of domain-specific knowledge, there is a greater potential for a larger increase in collective WM than when individuals have high levels of domain-specific knowledge required by the task.”

“With respect to cognitive load, if learners have relevant knowledge to carry out a task, communication and coordination activities may be unnecessary or even detrimental to learning. When there is little domain-specific knowledge, the cognitive load incurred by transactions could positively impact learning but where there is a great degree of expertise, and thus where transactions are either unnecessary for or detrimental to learning (Zambrano et al. 2017b), the cognitive load incurred could negatively impact learning.”

I’m glad I reviewed this article, as it has helped me to clarify my thinking about assigning group work.

To summarize the key point I’m revising my thinking about group work on:

Fromnew concepts and information should not be introduced during group work. Group work should be used for reinforcing, consolidating, and applying information students have already been exposed to


When assigning group work, ensure that the task is complex enough to warrant collaborative effort.

I’ll provide an updated decision-tree and guiding document to reflect this learning soon.

Shifting from Single to Twofold Vision


Seeing a school as an ecosystem is akin to shifting from a “single vision” to a “twofold vision,” as William Blake outlines it and Philip Pullman explains it:

. . . when it comes to vision, we need to be able to see contrary things and believe them both true: “Without Contraries is no progression” (The Marriage of Heaven and Hell), despite the scorn of rationalists whose single vision rejects anything that is not logically coherent. Blake was hard on single vision:

Now I a fourfold vision see
And a fourfold vision is given to me;
Tis fourfold in my supreme delight
And threefold in soft Beulahs night
And twofold Always. May God uskeep
From Single vision and Newtonssleep!
(“Letter to Thomas Butts”)

Fourfold vision is a state of ecstatic or mystical bliss. Threefold vision arises naturally from Beulah, which, in Blake’s mythology, is the place of poetic inspiration and dreams, “where Contrarieties are equally True” (Blake, Milton). Twofold vision is seeing not only with the eye, but through it, seeing contexts, associations, emotional meanings, connections. Single vision is the literal, rational, dissociated, uninflected view of the world characteristic, apparently, of the left hemisphere of the brain when the contextualising, empathetic, imaginative, emotionally involved right brain is disengaged or ignored.

William Blake and Me, Philip Pullman / The Guardian

We ain’t shooting for ecstasy nor poetic inspiration here, necessarily—not when it comes to school improvement. But if we can begin to see the wider complexity of connections and contexts beyond the linear machinations of everyday humdrum mundanity, then we might be getting somewhere.

Smorgasbord: Complexity, Reading, Morals, and Algorithms

Free vector graphic: Panda, Golf, Animal, Bear, Sports - Free ...

Stop wasting your time on item-analysis of standards and skills on state ELA tests, people

Tim Shanahan has some advice and candor that many principals and district leaders sorely need to hear.

“What makes the difference in reading performance isn’t practice answering certain question types, but practice in interpreting texts that are challenging–that pose barriers to meaning.

. . . The point isn’t that the standards should be ignored, but that teachers have to understand that reading comprehension tests do not/cannot measure single, separable, independent skills. These instruments provide nothing more than an overall indicator of general reading comprehension performance.”

This is the annual rigmarole that schools waste their ELA teachers’ time with at the beginning of each school year.

Stop it, folks. Just stop it. You’re not going to glean new insight about how to effectively teach literacy to your kids by doing intensive item analysis of the standards and questions on the ELA state test.

Instead, read real literature and engage your kids in learning about their world. Then you might actually have an impact.

A Spirited Reaction to One District’s Approach to Standards-Based Reading Instruction, Shanahan on Literacy

Teach morals by human example, not using cute animals

“Books that children can easily relate to increase their ability to apply the story’s lesson to their daily lives.”

But the study also notes that “The more a child attributed human characteristics to the anthropomorphic animals, the more they shared after reading the animal book.”

So as always, it’s about how the adults reading the books with children help them pay attention to and understand what’s most important.

Human Characters, Not Animals, Teach Children Best Moral Lessons, Neuroscience News

If laptops are detrimental to learning in college classrooms, then . . .

“We find that allowing any computer usage in the classroom—even with strict limitations—reduces students’ average final-exam performance by roughly one-fifth of a standard deviation.”

Should professors ban laptops?, Education Next

Bellwether reviewed NY’s ESSA plans and provides a useful critique.

NY has a strong foundation, but it’s accountability measures may be too complex for parents and the public to make sense of, as well as too vague.

An Independent Review of New York’s Draft ESSA Plan, Bellwether Education

A Friendly Reminder: Schools are Complex

“At least on paper, it is difficult to tell what separates the schools at the bottom of the list from those at the top, which cuts to the core of what makes school turnaround so difficult: nobody knows precisely what works.

‘The problem is that there is no silver bullet to turnaround interventions,’ said Priscilla Wohlstetter, a distinguished research professor at Columbia University’s Teacher College. ‘It’s a really tough thing to figure out what makes the difference in schools.’”

For $582 Million Spent on Troubled Schools, Some Gains, More Disappointments, NY Times

And school closures? Also complex

Make sure to read behind the headlines on the new CREDO study. There’s a lot of unknowns and nuance to their findings.

Matt Barnum does a nice job of drawing those out in this Chalkbeat piece.

“…the study can’t explain why closures happen more often in certain communities. For instance, if low-achieving schools with many white students are especially likely to be located in rural areas where there are fewer alternative schools, that may help explain the results.

Another explanation could be that the expansion of charter schools in high-minority areas puts additional fiscal and enrollment pressure on districts and charters — as charters expand, other schools may close as their enrollment declines.

What is clear, though, is that black and low-income students and communities are especially likely to have a school closed.”

Schools with more students of color are more likely to be shut down — and three other things to know about a big new study, Chalkbeat

A NY City Council bill could make public the algorithms that affect the public

An innovative–and some would say, long overdue, bill has been introduced in the City Council by Bronx Councilmember Vacca.

This would make the algorithm that the city uses to sort students for high school would be made transparent.

Showing the Algorithms Behind New York City Services, NY Times

Smorgasbord: NY State Test Results, Incoherency, and Teacher Shortages

NY State test results have been released: trends are positive

This year’s tests can actually be compared directly to last year’s, so inferences are slightly more valid. Statewide, ELA proficiency went up 1.9 points and math 1.1.

It will be interesting to see what narratives spring out of this. Even more interesting will be how anti-charter constituents spin the positive results from charters.

Look for all sides spinning these results in the way that suits them best.

State Education Department Releases Spring 2017 Grades 3-8 ELA and Math Assessment Results, NYSED

Speaking of measurement: How can we measure SEL?

Some interesting suggestions here from a recent design challenge:

  1. How quickly kids answer questions on an on-line test (too quickly means less self-control/engagement)
  2. Asking kids questions about a video to assess their perspective-taking abilities

Building a Modern Marshmallow Test: New Ways to Measure Social-Emotional Learning, EdWeek

It should go without saying that laptops alone do not a quality education make

You know, like, how are you actually using the laptops?

Do Laptops Help Learning? A Look At The Only Statewide School Laptop Program, NPR Ed

How we teach history depends on where we teach it

I’ve argued before that one of the biggest problems with what we teach students across our nation is that it’s completely incoherent, and we do little to nurture a collective sense of values, knowledge, and civic engagement.

Here’s that problem in action:

Virginia’s standards of learning for U.S. history to 1865 include “describing the cultural, economic and constitutional issues that divided the nation” and “explaining how the issues of states’ rights and slavery increased sectional tensions.” Alabama fifth-graders “identify causes of the Civil War from the Northern and Southern viewpoints.”

Contrast that with Delaware, where school districts set their own curriculum but a syllabus for the eighth grade suggesting what might be covered during instruction says that abolition meant that the American people could for the first time “seriously claim to be living up to their commitment to the principle of liberty rooted in the American state papers.”

In Michigan, curriculum also is decided locally, though the state’s social studies standards for the Civil War and Reconstruction in eighth grade include the instructions: “Explain the reasons (political, economic, and social) why Southern states seceded and explain the differences in the timing of secession in the Upper and Lower South.”

Civil War lessons often depend on where the classroom is, Associated Press

Teacher shortages in high needs areas, such as SPED and math, with no end in sight

One of the suggestions here for addressing this makes a lot of sense to me:

“Make teacher certification national instead of state by state. Prospective teachers must pass an exam specific to the state they want to work in. But if a teacher wants to move from, say, Pennsylvania to California, they can’t immediately apply for jobs there. By having a national certification exam, teachers would have more mobility to go where they’re needed.”

Schools throughout the country are grappling with teacher shortage, data show, CNN

One way of addressing teacher shortages in SPED: draw from the paraprofessionals

They’re already in the field. Make it easier for them to transition into teaching.

Makes sense to me. But one thing to be aware of: paras have great experience in managing behaviors and working with kids, but may not have a strong background on content.

Which is why having a strong curriculum and departmental teams that can support adaptation and implementation of that curriculum are so critical.

With principals in ‘crisis mode,’ new Washington state law taps into thousands of potential teacher recruits, Seattle Times