Hole in my Classroom

A while back I joined an #ecosys conversation about Dr. Sugata Mitra’s Hole in the Wall theory of learning. What Dr. Mitra did was put ATM style kiosks throughout India with internet access. In doing this he found that groups of kids were coming to these free computers and learning. Simplistically stated, Dr. Mitra believed that kids learned best in a small group of four around this one computer. I wanted to test this theory of group learning in my classroom. Bear with me while I explain the method to my madness…or just scroll to the results at the bottom.


Students will learn best if working in collaborative groups of four.


I took four learning standards in my Social Science curriculum that coincidently had to do with Ancient India. I created a four step model of “attacking” each standard.

  1. Identify the standard
  2. Create research questions based on the standard.
  3. Choose a method of gathering information to answer those questions. (notes, graphic organizer, concept map, etc.)
  4. Choose and create learning evidence for your standard. (essay, presentation, slide show, podcast, digital story, etc.)
For the first standard we went through the four steps as a whole class. I modeled each step and we discussed various examples for steps three and four. Once we modeled and went through the four step process the real work began. For the remaining three standards students worked independently with minimal teacher guidance. They were placed in one of three grouping scenarios. They were either working by themselves, with a partner, or in a group of four. Individuals working by themselves were responsible for doing all the steps independently with no help or discussion from peers. The partners could work together on all of the steps and even turn in one final piece of learning evidence for step four. The groups of four followed suit and did all their work collaboratively including turning in one piece of learning evidence for the whole group.

In order to validate the results, each student experienced each of the grouping scenarios. For example, if a student worked by themselves for the first standard they would be in a partnership for the second and a group of four for the third. This actually proved to be quite a logistical challenge to ensure all students experienced each grouping scenario. By the time we finished the three remaining standards all the kids rotated through the three group types.

The groupings were determined by myself in order to ability group the students. I wanted to make sure the results were not due to one student carrying to load but rather a true collaborative effort. While in some cases a strong student is a good idea, for this I wanted to keep everyone on a level playing field.

While students were working on their various activities they were given rough estimates of due dates but nothing firm to avoid anxiety. Once all the students had completed their work on the standards we took an assessment. The assessment was based directly on the three standards.

After the students took the assessments I graded them and matched the results of individual questions to determine which grouping scenario applies to which kid on which question. This was an exhausting process…


The results however were encouraging. The assessment itself consisted of short answer questions that directly aligned to the standards that the students were working on. In nearly every question on the assessment, the students that worked in groups of four performed higher than the partners and the individuals. I am not going to share the actual raw date but let's me honest, no one really looks at that stuff. However, the numbers did indicate higher scores for those students working in the groups of four which would align with Dr. Mitra's theory of the power of collaborative learning. Only one question did not have those same results but were within a tenth of a point average.

So based on this data kids performed best when working in groups of four. In addition, I surveyed the students to have them reflect on this process. Here are some of their thoughts:

“I think I learned the most in a group of 4 because the group each had different background knowledge.”

“I learned most in a group of four because there was ideas thrown everywhere. We debated and always chose [the] best ideas that I would never have thought of on my own.”
Final Thoughts:

First, students do learn more and better in a collaborative group. In my observations during the work, students in the groups of four were highly engaged in conversations about the work they were doing. Second, I took a very “back seat” approach to the learning going on. I was hands off and did very little direct instruction but rather fielded questions and gave direction when needed. When compared to previous assessments when I had a more direct instructional approach, students actually did better on this assessment. So what does this all mean? The less I taught, the more the students learned. When I gave them the power to choose how to learn, how to show their learning evidence, and let them “do” rather than “listen” they learned more.

As a result of this little “project” I will be providing more opportunities for students to work in collaborative groups as well as let go of the reins in the classroom a bit more.


sugata mitra said...

Great job! Thanks for trying. My only suggestions: Do let them make their own groups. They can change groups at any time. Anybody can look at what any other group is doing. Appoint a 'policeman' to look after law and order. Sit back and be invisible. All help to be provided by the policeman. You don't exist.....

Josh said...

Would this be the same as your "Granny Cloud" theory I have read about? That was another theory I was looking at working with. I like the idea of these small groups having an overseer keeping an eye and providing encouragement and redirection if needed. It is funny that the less I was teaching, the more the kids were actually learning. Very powerful!

Michael Josefowicz said...

Josh, I was so happy to find your post in today's twitter stream. It is so inspiring to see a classroom practitioner take an idea that was floated in an #ecosys convo and do further research in their class.

You have my gratitude for taking the time and having the focus to test an idea with a rigorous approach.

With all the talk I've heard about "action research" your simple, but thoughtful, research is a model of what can be done.

Mrs. Tenkely said...

Thank you for letting us take a peek inside your classroom. It is always nice to look at what strategies work and the approach that other educators take to ge there. Sounds like a success to me!

Anonymous said...

I just started following your blog on my google reader. what a practical and useful research. i've been looking in to collaborative learning as well. our school is moving towards e-learning where the learning is very much lead my the students. thanks so much for this post. i will definitely be implementing the 'group of four' method in my classroom.

Janet Abercrombie said...

I've been wondering about pairs vs. groups of four. I often have pairs "turn and talk" to reflect on a small part of a lesson or brainstorm ideas.

The REAL power comes when the pairs turn and talk to another pair. It's difficult to get them to stop talking (on-task talking, of course).

I need to try fours with other lessons. Thanks for the reminder.

Janet | expateducator.com

Dave Meister said...


Did you find that you interacted more with the groups of four because they were able to formulate more and better questions? It would have been interesting to do a teacher-presence-interaction map and chart to evaluate how the different groupings affected your interactions. Did you keep any data on your actions and activities during the experiment?

Interesting, deserves some follow up!