Schoolhoming: Cohorts

(I refer to due diligence and research and interviews throughout this; myself and some like-minded friends are working to collect actual data to design a better educational system. The best way to do that is by conducting interviews with real people. Interviews are anonymized and the data is synthesized to reveal root causes and suggest best approaches.)

The first thing to consider when designing something is who you are designing for. As my friends and I are thinking about how to do education better, we are thinking about:

  • all students
  • all teachers
  • parents who can’t afford private tutors or schools and who have to work
  • parents, students, and teachers who are concerned about contracting or spreading Covid-19

We are not solving for school staff or administrators, as we believe the parameters of those jobs will change based on the designs that work best for the students, the teachers, and the parents.

The next thing to think about are “extreme users.” Just as an airline pilot and a first-time flyer don’t experience the security lines at an airport the same way, there are types of students, teachers, and parents who will be the outliers as far as needs go. If your design satisfies oppositional extreme users, it probably works great for mainstream users.

In this case, the student population is where we really want to look for extremes. Here are some of the categories we’ve come up with:

  • students with ability mismatches (you probably call this “disability;” one or more elements of the student’s mental, physical, or emotional toolkit is mismatched with their environment. This might be that English is not their primary language or it might be that they are deaf. It might be that they are autistic or that tthey are in a wheelchair)
  • students with at least one parent who is available and an educator (these students are likely to have better outcomes)
  • students with siblings who are also school age
  • students in abusive home environments
  • students who rely on public school for food
  • autodidacts

We are working to develop a set of actionable user personas for student populations, but this is good enough for now as we are in our due diligence phase.

What is a cohort?

Forget everything you know about how schoolchildren have to be arranged into classes based on age and neighborhood (here’s a good video on the industrial school process). Schoolhoming borrows the concept from homeschooling that kids in the same family stay together, with the small class size allowing personal attention. We’ll leverage data science and algorithms to create cohorts that keep families together and friends together…basically, if kids are going to hang out together anyway, they are in the same cohort. This reduces the number of branching possibilities in a contact tracing map…which is another way to say that it reduces the chances of getting or transmitting Covid-19.

a sketch of a possible cohort composition

I’ve sketched one possible cohort of 12 students plus a teacher. (In reality I expect cohorts to go up to 25 students but I’m designing for best until I have to compromise). Each color represents a family; there are 5 only children (or without school age siblings), two sets of two siblings, and one set of three siblings. One of the only children is the teacher’s kid. (This is a type of torture for both teacher and child, but it reduces contagion risk a lot.) One of the sibling pairs was put in this cohort because they are friends with one of the Purple Family (the three sibs). One of the singletons is in this cohort because they are friends with another one. (“Friends” can also mean cousins, or they attend the same church, or are in the same scouting troop…the point is they don’t live under the same roof but are likely to associate outside of school). The other students are in this cohort due to other criteria; perhaps they live near where the cohort assembles for school, or their parent works nearby making commute easier, or they are a good match for special needs (IB, AP, ESL, etc) and that teacher.

Already living together is weighted heavily in the algorithms that assemble the cohorts. Special needs for mental and emotional mismatches are also weighted heavily. Location proximity is not weighted as highly, but location accessibility is weighted heavily for students and teachers with mobility issues. Teacher expertise with other physical disabilities is a factor. Student languages are a factor.

This unit stays together from kindergarten through 12th grade. When this gets started, there may be cohorts that include kindergarteners and seniors, but after one or two cycles cohorts are likely to be K-8th and 6th-12th.

Cohorts and Coronavirus

In the diagram I’ve included the next tier of contact tracing concerns for contagion. There are 8 nodes with this cohort, as compared to 30 nodes in a standard public school class (30 families). This reduces contagion risk, and if the adults follow good safety protocols it’s pretty solid. But, assuming the worst happens and the virus appears in a cohort, just that cohort can be quarantined rather than shutting down the education for everyone.

Additionally, safety protocols such as checking temperature, providing masks and ensuring good mask discipline, and effectively disinfecting class spaces are easier with smaller groups that are geographically dispersed. Teachers or older kids should use forehead scanning thermometers before admitting students to the classroom, and may want to check again at lunch and at the end of each day. Social distancing is easier as well, but the physical spaces for classes are going to be covered in another post.

Only about 2,500, or 1.7% of nearly 150,000 cases of COVID-19 in the U.S. between Feb. 12 and April 2 were in children, according to the U.S. Centers for Disease Control and Prevention (CDC), and the R0 is 2.2 That means that one case of Covid-19 creates 2.2 more cases, so if one member of a cohort gets the virus, most of the rest of the cohort will be fine. 1 out of every 10,000 kids develops serious symptoms. The school district my son is in has 187,830 students, so we expect 18 or 19 of them to get a serious case (in fact, two teenage kids of a work colleague just got out of a two-month hospital stay) if we don’t create super-spreader situations such as going back to school in the same manner we did school back in January.

The goal of the cohort approach is to reduce the R0 for student populations, so that a case of Covid-19 just doesn’t have access to enough potential hosts to keep up a 2.2 new case rate. Slowing that transmission should help reduce the probability of a child getting a severe case. However, education doesn’t exist in a vacuum; families need to observe good safety protocols at home and out and about, and kids attending school should limit their in-person interactions outside the cohort. Communities need to increase health care capacity, such as training up more health care providers, creating more hospital bed space, providing testing and PPE. The full phrase is “lower the curve and raise the line” and I don’t know of any community that has been doing the second part in the US.

My son’s high school is 2600 students. Cramming that many people into one building is a super-spreader. Fully online education doesn’t address psychological health and leaves out many special needs groups, even when it’s done well. The only two real paths I see are cohorts or seminar style education, and seminars don’t work for younger kids.

Cohorts and Education

The great thing about the small group size is that everyone can work at their own speed. As one of the “smart kids” I spent my school years being bored out of my mind. In our interviews with students and parents, almost no one views the experience of school fondly, and the education is either too challenging, too confusing, not challenging enough, or meaningless. Everyone is motivated by different things, and intakes information differently. A teacher can adapt to that with a small class.

Which is incredibly different and hard on the teacher. In order to assess success, there need to be standards, which means that teachers will need to meet those standards for each student while also customizing the educational experience. This is one of the reasons we are hoping to do effective research and create good user personas for students; rather than 12 different classes in one cohort there are probably only 3 or 4 max. Homeschool parents say that the curriculum takes a ridiculous amount of time to assemble, and it’s hard to give the right amount of attention to each kid. Younger kids need more hand-holding, while older kids are moving through topics really fast.

However, the older kids can become a force multiplier for the teacher, helping teach the younger ones as a way to reinforce their own grasp of the subject as well as demonstrate mastery. Younger students learn academics and social skills from the older ones. This should be scaled up, with one cohort teaching another cohort on a topic via online technology. Online cohort interaction should be a regular and frequent part of this program, to build relationships, diversity, and inclusion.

Another good thing about cohort teaching is that you don’t have to be on an industrial bell schedule. Each segment can be shorter, in line with actual human cognition and attention. For those of us who are self-motivated, being able to go as fast as we want helps our retention…but we are in the minority. Most students are more like my son. He used to crastinate but now he’s gone pro.

More advanced kids will need specialist teachers. Rather than introducing new organisms into the classroom, expert teachers will telepresence in (whether than means on screen or inhabiting a robot depends on needs). The cohort teacher can participate and upskill, or can use this time to focus on the younger kids. Specialized education (think high school science, AP, IB, fine arts) will probably pull students from multiple cohorts into an online environment with a specialist teacher.

For advanced high school students, cohorts might include seminars or be replaced by seminars. Seminar-style education is sometimes called Oxford style. Student and professor meet one-on-one once a week for a long, in-depth discussion. The professor assigns the student work for the following week, including research and writing. Since the professor’s knowledge of the subject is broad and deep, the direction can follow lines that the student is most interested in, or needs the most.

Cohorts and Teachers

A cohort approach requires a LOT more teachers than currently exist in America, and requires that the teachers be able to span K-12. Obviously this means we need to remunerate teachers at a competitive rate. As a long-time counter-terrorism professional, I personally think federal CT funding should be pulled from law enforcement and given to school districts, where it will do more good…but overall, teacher pay and benefits is a huge social issue that America must address.

On the other hand, a cohort approach has been much more appealing to the teachers we’ve interviewed. The smaller class sizes is a big plus, as is the ability to stay with kids through their school careers (*some* kids…there are always some kids you want to see the north end of heading south). All teachers said they would be much less concerned about their physical health with this plan, vice reopening schools. No teacher reported enjoying teaching online, although a few said it was fine, given the alternatives. A couple teachers mentioned that they appreciated not having to worry about school shootings. All teachers wondered how administration and equitable grading would be handled.

We’re still working with teachers to figure those things out; we assume that cohorts will be grouped somehow, and those teachers will have regular online conferences and shared online space where they can standardize lesson plans, etc. We’re looking atLaloux’s Reinventing Organizations for models on evolving the educational system into an ecosystem, or a “shared power, decentralized network,” which will be crucial for coordinating a large number of distributed small cohorts.

Illustration of the Laloux Culture Model from shift314

If you are a teacher and you have thoughts on this, and would be willing to be interviewed, please let me know!

Cohorts and Parents

Cohorts will also allow us to provide a safe place for kids to be for the whole day. In the last month I’ve interviewed 86 parents who are stressed out of their minds because they have to be at work for 8 hours but their kids will be at home, and are too young to be left alone. Even if the parent is working from home. Trust me, you can’t be both mentally at work and be a child’s caregiver.

In truth, school and work have never matched up. The school schedule is so much shorter than the work day (plus commute) that parents have been having to lie, cheat, and steal to balance both. This is a chance to make things even better than they were.

However, we can’t ask kids to be in hardcore learning mode for 10 hours a day, nor can we ask teachers to be “on” for that long, and then turn around and do all the behind-the-scenes work. I do believe that cohorts can assemble early and end late, but much of that time should be downtime for both the teacher and the students. Some parents will be able to come in during the day to be the adult (they are already part of the super-organism that either has covid or doesn’t, but they should be temperature checked and wear a mask and all that). Some parents will pick kids up for other activities. Parents whose employers have recognized that employees are humans and may have kids, and thus have flexible work schedules, will be able to coordinate with the teacher.

The idea of adding a second adult to the cohort, sort of a teacher’s assistant who focuses less on educating and more on childcare, has been suggested a couple times. This adds expense but doesn’t really change the contagion math, so I would love to see businesses contribute funding for this so that their employees can be at work, and so that the business doesn’t have to stand up its own childcare facility with the huge expense and covid risk that entails.

If you are in HR or a similar field, or an executive, please consider this…if you have thoughts and would like to be interviewed for our research, please let me know. If you are an employee but you like this idea and would like to bring it to your employer’s attention, stay tuned – as we refine the design and stand up some test cohorts, we will announce it.

Cohorts and You

Does this make sense? Obviously we aren’t going to shift the entire US to a new paradigm by September, but we’ll address how to actually start changing stuff later. Right now we want to hear from you: if you are a student, would you prefer this style of education over 100% online, 100% back in school as usual, or whatever your school is planning for this year? If you homeschool, would you consider shifting to be part of a cohort? If you were a student, in homeschool or public or private, how does this compare?

If you are a teacher, does this sound better or worse than what you’re planning for this fall? What are your biggest concerns?

If you are a parent, would you be comfortable sending your kids to a schoolhome? What are your biggest concerns about education right now?


  1. I’m curious if this cohort, schoolhoming idea could work in tangent with public education’s efforts to provide virtual learning. For example, I have a son going into 5th grade. If we carefully created a cohort, keeping in mind the variables you mentioned and contact tracing, could you create a “mini class” inside that teacher’s virtual class? Typically my son’s class has about 25 students. We get 4 classmates to make a cohort of 5. They could rotate houses – daily, weekly, every other day, whatever – and parents can take turns basically being the virtual teacher’s assistant. All kids on their laptops, participating. Could have time set aside for specials like “PE” (e.g. bike ride, walk), music, art, and actual hands on learning. Thoughts? What are the major pitfalls to an arrangement like this?

    1. Turns out lots of people are thinking about this! Here’s a group on FB to find and join and coordinate:

      Given that kids under 12 are very unlikely to get covid-19 or be contagious, I expected all school districts to provide school for that age group…but since they are not, it’s at least safe for groups of families to do it at home.

      I’ve started reaching out to my son’s friends to offer our house for this, but I continue to worry about the low-income families that don’t have this option.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.