My Speech at the People’s Climate March 2017

150,000 people. That’s the estimate for the People’s Climate March that took place on Saturday, April 29, 2017 in Washington, DC. Some estimates even put the number at 200,000!

(That’s 200,000 people, not 200,000 factorial, obviously. This is a bad math joke. You should probably just move along now.)

Before the march began, labor organizers held a rally in front of the Department of Labor, and I was asked to speak on behalf of the American Federation of Teachers and the Washington Teachers Union.

It was an honor to share a stage with such passionate and accomplished union activists. As a teacher, I felt a unique responsibility to draw attention to some of the (many) ways that public education is currently threatened, and how those threats on education are fundamentally threats on both democracy and the planet.

Here is a video of my speech and a written version of my remarks.

Thank you all so much! My name is Joe Herbert, and I am a high school math teacher right here in Washington, DC.

I’m here today because public education is the cornerstone of our democracy.

This current administration wants to destroy public education, just like it has no hesitation destroying the air we breathe and the water we drink.

We need to protect our kids and our planet. We need to prepare the children of today to be the engineers of tomorrow, the computer scientists of tomorrow – the artists, poets, clean energy experts, journalists, labor organizers, civil rights lawyers, and mathematicians of tomorrow.

For that, we need strong public schools. We need public schools that consider science, civics, and the arts to be just as important as math and English.

Today, there are students in American public schools who do not get any science or social studies education because they have extra math and reading.

Let me repeat: in 2017, there are students in American public schools not receiving an education – being deprived of an education – in science and social studies.

This is an abomination and a national embarrassment.

If this generation is to grow up to be an informed electorate, we need public schools that teach students to interpret evidence and think critically.

That’s what we’re here fighting for today.

But there is a sustained assault on public education, and it is very unfortunately bipartisan.

Right here in the Washington, DC, we have a Democratic mayor who has refused to give the Washington Teachers Union a fair contract.

So that as of today, our contract expired 1,672 days ago.

Many, many Democrats are cheerleaders for charter schools, which siphon money away from public schools and increase school segregation.

We need a renewed commitment to the principles of school integration set forth in the landmark Brown v. Board of Education case.

We need to fight back against these so-called “school choice” policies like charter schools and vouchers.

These policies increase racial segregation, harm public schools, and have a disproportionate impact on poor students, on black and brown students, on students with disabilities, and on students who are immigrants.

We must commit, as a society, to funding and supporting our public schools and our public school teachers, so that we protect the social good and pillar of democracy that public education represents.

And, we must guarantee that this education is afforded to every child, in every school, in every neighborhood in America.

This will ensure that the next generation has the critical thinking skills, the knowledge of history, and the understanding of science necessary to avoid repeating the mistakes that we see our politicians making today.

So let’s advocate for quality public schools at the national, state, and local level.

Let’s hold our elected officials from both parties accountable, and demand that they support public schools.

My students inspire me every single day, and it is because of them that I am optimistic about tomorrow.

Let’s leave them a healthy planet, and a brighter, safer, greener, and more just future.

Thank you! And thank you for standing up for our planet and our children!

Shuffle Tests: An Alternative Assessment to Combat Status (Part 2 of 2)

This is the second post in a two-part series about shuffle tests. In the first post, I discussed what a shuffle test is, as well as some of the successes and failures I experienced implementing a shuffle test in my classroom.

In this post, I will discuss some objections to the very notion of a shuffle test. I will also go into greater depth as to why this alternative assessment is a powerful tool to promote equity and combat status in the classroom.

Objections to the Concept of a Shuffle Test

In implementing a shuffle test in my classroom, I encountered two objections. One came from a student, one from a teacher.

A few of my higher-achieving students objected to the fact that their oral exam grade on a shuffle test would be dependent upon whether or not a classmate understands the material. My response to this was two-fold:

  1. Communicating clearly is an important part of doing mathematics, so part of what I’m assessing with this test is the ability of everyone in the group to explain math clearly to their classmates.
  2. My goal is for you to succeed, and the oral exam is not a “gotcha” test. If someone in a group isn’t able to explain a problem or answer my questions, I will simply say, without judgment, “it looks like this group isn’t ready yet,” and walk away to give you more time to discuss the problem. I’ll return again when you’re ready.

This response satisfied students’ initial concerns, and I didn’t hear any other complaints.

The objection from the teacher standpoint is that a shuffle test is not as rigorous an assessment as a traditional test. Here are my thoughts on that:

  1. The oral exam component provides accountability to ensure that students aren’t just mindlessly copying from the “smart” kid in the group.
  2. This test serves a dual purpose: it is a test, but it is also a status intervention. Thus, you cannot look at its benefit to the classroom purely through the lens of its ability to measure student understanding. It is also a profound way to get ALL students collaborating on and persevering through challenging problems.

This is Deeper than Just an Alternative Assessment

It’s worth exploring a bit more that second point above: a shuffle test is fundamentally a status intervention.

If you’re unfamiliar with the notion of status, I highly recommend reading everything that Ilana (Lani) Horn has ever written. Lani is a researcher at Vanderbilt who studies equitable teaching. She identifies a key problem that hinders student learning: status, which she defines as, “the perception of students’ academic capability and social desirability.”

As Lani notes, “The word perception is key to this definition.” Students who perceive themselves as “good at math” don’t always want to listen to or value ideas from their peers whom they perceive as being “bad at math”. Similarly, students who have struggled for years in math often don’t feel like they have anything valuable to add to the conversation.

Lani has discussed many status interventions, such as establishing and maintaining norms, assigning competence, broadening students’ definition of “smartness” in math class, and visibly random groupings. I have also written on the importance and ease of using visibly random groups.

The status interventions listed above are primarily focused on the culture around everyday classroom tasks and activities. As wonderful as it is to promote positive interdependence among students during class with these status interventions, these interventions should not just be relegated to classroom activities, but should also include assessments. As Lani points out, “Assessment is one of the most powerful ways teachers communicate their values to students” (Strength in Numbers, p. 56).

Lani discusses shuffle quizzes as an example of what “Paul Black and Dylan Wiliam and their colleagues refer to … as assessment for learning“. That is, the primary purpose of the assessment is not to evaluate students, but rather, to promote their learning (Strength in Numbers, p. 56).

This was definitely the case in my classroom. This was the first question on my shuffle test:

Screen Shot 2017-03-18 at 1.51.30 PM

We had worked with composite figures before, but I had never given students a question in which “cutting out” part of a shape decreased its area but increased its perimeter.

Some groups were able to handle this entirely on their own, but I noticed others in which all of the students were leaving incorrect answers to the perimeter and moving on to the next question. There were two ways this was happening:

  1. They only added the straight parts of the shaded region’s perimeter.
  2. They treated perimeter like area: they found the total perimeter of the rectangle, then subtracted away 14 cm for the dotted lines.

For these groups, I intervened by asking the following two questions:

  1. What does perimeter mean?
  2. Trace for me with your pencil or finger the perimeter of the shaded region.

These two questions were enough to trigger a lightbulb moment in at least one student in each group. Additionally, I think that every student’s memory of that lightbulb moment will be much stronger than it would have been if I had “taught” them how to do this type of problem through direct instruction, since they struggled and thought about this funky perimeter for a significant and sustained period of time before finally seeing how the perimeter needs to be treated as fundamentally different than the area.

When Lani discusses assessments for learning and the research behind them by Black, Wiliam, et al., it is primarily in the context of formative assessments that will guide the teacher’s future instruction.

I used my shuffle test as an end-of-unit summative assessment, and I was very happy with its ability to serve that purpose. That being said, after seeing the incredibly rich mathematical thinking and discussion that my students engaged in during the shuffle test, I definitely want to try to incorporate this more frequently as a formative tool in my classroom.

Shuffle Tests: An Alternative Assessment to Combat Status (Part 1 of 2)

Screen Shot 2017-03-28 at 8.39.13 PMI have often found that many students underperform on traditional tests. They may know the material, they may work hard and be able to explain things in class, and they may even study, but when faced with a traditional, individual, silent testing environment, their stress and anxiety get the better of them and they freeze up.

As a result, I’ve been looking for alternative ways to have students demonstrate mastery, especially for my English Language Learner class, because they seem to particularly struggle with traditional tests.

This post is the first in a two-part series about an extremely exciting and positive experience I had while implementing one type of alternative assessment: the shuffle test. Not only is a shuffle test a great way for all students to access and demonstrate mastery of challenging content, it also has the added benefit of being a very effective status intervention.

In this post, I’ll discuss the mechanics of a shuffle test and talk about my experience implementing it. In the next post, I’ll go more into the pedagogy of why this is a powerful tool to promote equity in the classroom.

What is a Shuffle Test?

  • Students are randomly assigned to groups
  • The test itself consists of a small number of rich, challenging problems
  • Each student submits their own written solutions for an individual grade
  • Each group receives a collective group grade for an oral exam in which students explain the group’s solution to one of the questions

Some Important Points

Because the students are working in groups, the questions can and should be more difficult than those you would give to students on an individual test.

A key part of the shuffle test is the group grade for the oral exam. Traditionally, for the oral exam portion of a shuffle test you randomly choose one student to explain the group’s solution to one of the test’s questions. This means that all students need to ensure that all group members understand each problem on the test fully. This forces them to work together and support each other.

I actually did the oral exam portion of the shuffle test a little bit differently than the “one student at random answers for the group” model that I originally learned about. When I first learned about shuffle tests, I saw a video of an oral exam. In the video, after one student in the group was randomly selected to explain the group’s solution, one of the other group members pretty much checked out since they were now “off the hook”.

To mitigate this, I had students take turns explaining their group’s solution. I used equity cards to randomly select who would start explaining. When a student was selected, they were the only one in the group allowed to talk. After they had talked for a bit, I would use the cards to randomly select another group member to pick up the explanation where their colleague had left off. This forced all students to pay careful attention during the oral exam because they knew that they would be the one explaining at any minute.

My Experience: What Went Well

Really, the impetus to write this blog post was that I needed to share how shockingly sublime my classroom was during the shuffle test. I have developed a pretty decent culture of collaboration and group work in my classes, but it (like most things) is imperfect – a perpetual work in progress that requires constant vigilance and upkeep on my part.

The day of the shuffle test, however, was truly a dream come true: after I reviewed the procedure and expectations for the shuffle test, kids got into their groups and immediately started working together. They found the questions to be really challenging, but they just buckled down on working through them together rather than asking me anything. I think that by calling it a “test”, they more easily internalized the notion that they should rely on each other rather than me. 

I spent a significant amount of time casually milling about the room, subtly eavesdropping on their phenomenal discussions and making mental notes to myself, but really, my presence was not necessary for the excellent math thinking and learning that they were doing.

In one of my classes, I had been worried because a student who is a bit of a … *character* … had been absent for several preceding classes and had missed some of the key material on the test. I was worried that this student might act out as a defense mechanism (as was often the case), and that their group members would get frustrated and turn on them.

What actually happened was the exact opposite: this student was the most engaged I’ve seen all year, and worked really hard to pull their weight in the group. The other group members – including a super high-achieving, high-status student who always needs everyone to know that they’re right – responded by helping this student with the material they had missed, and welcoming them into the group.

My Experience: What Could Have Gone Better

The test that I gave was too long for one period. While some groups almost finished on day one, others were only about halfway done. I had suspected that I might need to budget two days for this, but there were some definite downsides to continuing into a second day.

The most fundamental downside was that students lost a sense of momentum and focus that they had had on day one. While the class still went well on the second day, with kids mostly on-task and working together on the math, that idyllic classroom utopia that I described above had returned a bit to its normal, human (and thus flawed) state.

Additionally, I had several kids who were out on a field trip, which messed up the groups. Even worse, my Geometry classes are at the end of the day, so as the field trip ended and students started to arrive back at school (at different times depending on the metro train their field trip group had caught), they trickled back into the class, forcing me to rearrange the groups several times.

My goal for the next shuffle test is to have it be short enough that every group at least finishes their written solutions in one class period. While it would be ideal to do all of the oral exams in the period as well, I think that as long as students have answered every question and written up all of their solutions, it wouldn’t be too hard to have groups come after school or at lunch to complete their oral exam.

I would like to give a special shout-out and thank you to Bill Day and Julia Penn, who led the Math for America DC session that first introduced me to the concept of a shuffle test.

In my second post on shuffle tests, I discuss some objections to using shuffle tests (from both students and teachers), as well as more of the pedagogical theory behind why shuffle tests are a powerful way to promote equity in the classroom.

A Tool for Organized Planning

There are only two types of people who truly understand the volume of papers that teachers must manage: teachers and the family members of teachers. As the son of a teacher, I swore to myself that I would not have the piles upon piles of paper invade my home the way my dad’s papers used to invade ours growing up.

I wish I were writing a post with a silver bullet to keep papers and grading perfectly organized. I’m not. (But please do leave any suggestions in the comments!)

While I still struggle with paper organization, I have created a nice tool to keep my lesson planning organized: a custom-made googledoc spreadsheet that I use as a calendar.

This is my master planning tool. It’s color-coded by month, and I’ve pre-populated it with important holidays and dates in the DCPS calendar. It’s an easy-to-edit, big-picture view of my year that I can share with colleagues and co-teachers.

When I’m teaching a course for a second time, I often keep the old course map open to look at how I approached different topics. I also make notes to myself in the spaces to the right of the calendar with lessons learned along the way, such as the following note from the polynomials unit of my pre-calc class last year:

Screen Shot 2016-08-17 at 3.12.45 PM.png

Each year I start with a new blank slate and use the calendar to map out my units.

Please feel free to copy this calendar into a new googledoc to use for your own organizing.

What other things do you do to stay organized?

Update 8/21/16: Some people have been unable to copy the calendar from the view-only setting I created for the public link. Let me know if that’s the case, and I can give you temporary editing privileges.

Visibly Random Groupings: Why an Initially Terrifying Prospect Turns Out to be My Favorite Way to Promote Equity

Summer is often a time of reflection and change. We teachers make our New Year’s resolutions in August, not January. So as you think about the coming school year, and what small steps (or giant leaps) you’d like to take to do things differently, let me make an easy suggestion that changed my life this past school year: group students randomly.

Wait! Don’t go! Before you stop reading because, “that’s just crazy!” or “that’ll never work in my classroom” or “what about these ten problems with implementing random groupings that I immediately thought of?!”, allow me the time to argue that random groupings (1) are good for students, and (2) make your life easier. And yes, I promise to address your concerns about why this is the craziest thing you’ve ever heard of.

Good for Students: How Random Groupings Send Subtle Yet Important Messages that Promote Equity

In my opinion, the single most important reason to group students randomly is the fact that when we don’t, students will try to read into the groups we assign and the roles they should fill within those groups.

If we group homogeneously, they might think, “Yep, I’m in the dumb kids’ group. No surprises there. I’ll just pray I can make it through this class without looking like an idiot,” or alternatively, “I’m in the smart kids’ group! That means my ideas matter the most when we discuss things as a class.”

If we group heterogeneously, they might think, “I’m the smart one in this group; these other kids are not going to be able to keep up with my genius or help me solve this problem,” or alternatively, “I’m clearly the dumb one here. I’ll just copy down whatever the smart kid says.”

Even if students aren’t conscientiously narrating these thoughts to themselves, they are certainly internalizing the subtext of the group they have been placed in and their “role” with that group.

Furthermore, these unproductive thoughts are often filtered through the lens of race and gender and any other way that students can feel marginalized in a math class. Whether we like it or not, students carry the baggage of cultural norms and biases, social expectations, and status into the classroom.

Random groupings therefore send a powerful message: everyone is equal, everyone’s ideas are valid and deserve to be heard.

Before learning about random groupings, I used to make groups by trying to “spread out” the strong and weak students, i.e. I’d make sure that my strong students weren’t too clustered, and that the weaker students always had someone who could support them. While I did this with the best of intentions, this approach is problematic:

  • I was assigning “strong” and “weak” classifications to students based on prior achievement. Despite the cumulative nature of much of mathematics, a quality math task will invite innovative and creative ideas that can come from anyone, even a student who struggles with some basic skills.
  • There are many ways to be “smart” in math class. Most people tend to assume that being “good at math” means being able to do basic arithmetic computations quickly and accurately. While that’s a great skill to have, most mathematicians (myself included!) will readily admit to struggling with that particular skill. When we assign groups based on prior achievement, we’re only validating a narrow definition of what constitutes being “good at math”.

By grouping students randomly, we send the message that there are multiple ways to be smart at math (seeing patterns and connections, drawing a graph or diagram to visualize a problem, communicating ideas clearly, etc.).

Obviously your students won’t fully internalize the idea that there are lots of ways to be smart at math simply from random groupings alone; this message needs to be explicitly reinforced many times over. But random groupings are a powerful way to put your money where your mouth is. This can start to chip away at the status barriers preventing full participation and engagement.

If we truly believe that all students can be successful in mathematics (and if you don’t, why on earth are you teaching?), then visibly random groupings are the only way to organize the classroom.

For more about why random groupings are essential, listen to this tumblr post from Ilana Horn, check out her blog Teaching Math Culture, or read her book, Strength in Numbers: Collaborative Learning in Secondary Mathematics.

But What About…? Calming Your Fears

I will be honest: even after reading enough to be convinced that visibly random groupings are a good idea in theory, I was terrified of actually implementing them. What if the random groups create nightmarish social dynamics? What if all the weak kids get put into one group and completely flounder?

To that second concern, I’ll note that if a bunch of “low” students end up in one group together, oftentimes they will realize that there are no “strong” students to hide behind, and they will rise to the challenge. And if they don’t, then at least they’ll be concentrated and you can spend extra time with their group!

But back to the broader concern that randomization can create bad social dynamics. This is obviously true, and you don’t have to live with those bad dynamics. Supporting randomized groupings doesn’t mean you have to blindly accept a group that you know won’t work together well. The key is that you change groups based on social, not academic considerations. For example, you could say, “I know you two always get chatty together, so I’m going to switch you to another group.” Or, as was the case in one of my classes last year, if two students have associations with rival gangs, they should definitely not be seated together!

As long as the groups are visibly randomized from an academic standpoint, with any conscientious tweaks occurring for purely social reasons, your classroom will still get the positive benefits outlined above.


So how can you actually implement this in your classroom? Ilana Horn shows an example of a chart you could hang in your room with different group roles. Simply shuffle cards with students’ names on them and place them randomly into the slots. Screen Shot 2016-07-06 at 11.00.15 PM

That’s way more sophisticated than what I did! At the beginning of this past school year, I was so nervous about random groupings that I decided to start out simple: I only did random groupings for a few specific, group-work tasks. I would visibly shuffle cards with students’ names, then put them down four at a time on desks to create the groups, saying the students’ names as I went.

Later in the year, I realized that random groupings are actually the greatest thing ever and I started using randomization to create my general seating chart. I would shuffle the cards, then have a student read them to me in order as I wrote students’ names on a blank seating chart. I would then allow students to silently raise their hands to request to move closer to the front if they had bad eyesight and had been placed too far back.

This whole process took about five minutes, and created a visibly randomized seating chart with student input. I found it was generally best to do this at the end of class and then have the new seating chart posted as students came in the next day.

Good for Teachers: How Random Groupings are Actually the Greatest Thing Ever from a Selfish Point of View

I hope that I’ve convinced you that random groupings are both good for your students and logistically feasible. But now for the selfish part: they will make your life as a teacher significantly easier.

Creating seating charts randomly has freed up soooooo much time! I cannot even begin to describe how much time I wasted two years ago trying in vain to balance the delicate Rubik’s Cube of student personalities and abilities, only to have each new seating chart seem even worse than the previous one. By making my seating charts randomly, I essentially gave myself hours of extra time to devote to lesson planning and grading.

Additionally, students used to feel personally insulted if I made a new seating chart that separated them from their friends, and they would lobby incessantly for specific seating arrangements. When students get a new seat that they don’t like but understand that it arose from a visibly randomized process, they complain less and for a shorter amount of time.

I also think that using randomization has made me a better teacher. It’s a great reminder to view all students as having the potential to make meaningful contributions, and it helps me avoid deficit thinking. When I began using random groupings this past school year, I found myself expecting more from the students who struggled the most, and as a result, they rose to higher challenges. (Mostly. Ish. Depending on the kid. This is education after all; silver bullets don’t exist!)

On the opposite end of the spectrum, random groupings are a good reminder that we can’t always tell who is going to struggle with a given task and who is going to need support. I had a student in my Geometry class this past year who demonstrated very early on that she was incredibly strong with the procedural skill of solving algebraic equations and was also quite adept at correctly setting up equations from geometric situations. I therefore thought of this student as one of my “stronger” students. When we started working on proofs, a lot of kids needed support and feedback, and I found myself being spread quite thin around the room. I ended up not checking in on this student (or the group she was working with) because I thought that she was one of my highest flyers and would be fine without me. It wasn’t until I read her proofs that I saw she was grasping at straws trying to form a coherent and logical argument.

I had let my preconceived notions of this student’s intelligence get in the way of recognizing that she needed more guidance and support to complete the task at hand. Random groupings can be a helpful reminder that we can’t necessarily judge future success from prior achievement in a subject as complicated, disparate, and interconnected as mathematics.

Concluding Thoughts

If you’re not already using random groupings, I hope you’ll consider this for next school year. It has the advantage of being good for students’ conceptions of themselves and others in math class, and also making your life easier. Furthermore, this is a fairly easy and straightforward new strategy to implement.

If you already use random groups, what have your experiences been? Do you have other ways that you put students into groups randomly? Please share!

Will random groupings magically solve all status issues in your classroom? Of course not. But it is one powerful piece of the puzzle that can help all students recognize that they have valuable ideas to contribute to the class.

Thoughts? Questions? Concerns? Please let me know!


A New Beginning or More of the Same?

There is a blind faith in data and numbers that pervades much of our society, particularly education reform circles. People who worship at the altar of this faith look at a school with low test scores and assume that the school (and by extension the teachers and principal) are “bad”. They often fail to ask some basic questions, such as:

  1. What is the quality of the test that the students took? Does it even measure what it purports to measure?
  2. Is the exam developmentally appropriate?
  3. How many of these students are emergent bilinguals? Were there any language accommodations given other than extended time and a dictionary?
  4. How many of these students experience violence or hunger or fear or insecurity on a daily basis? How does this affect their ability to concentrate?
  5. Did the students even try? Were they invested in the exam? Does it really represent the students’ best effort and understanding?

I could go on and on and on. This is just a small sampling of the many factors that make student standardized test data significantly less than reliable. Education reformers will tell you that there are big, complicated algorithms to account for things like student background and prior knowledge. As a mathematician, I am not intimidated into complacence by talk of a “complicated algorithm”. Rather, I note that the American Statistical Association has discredited the current obsession with test scores.

This brings me to the big news in DCPS this week: Kaya Henderson has announced that she will be leaving as chancellor. Will Mayor Bowser appoint someone who understands the vast limitations of standardized tests? Will she appoint someone who views PARCC results with serious skepticism? Someone who genuinely recognizes and trusts the professionalism of the district’s teachers (beyond the mere lip service that Kaya Henderson is so fond of)?

I’m not holding my breath for the kind of chancellor this district needs, but maybe I’ll be pleasantly surprised. New beginnings are new beginnings, and maybe DC will finally decide to listen to solid research instead of blindly chasing meaningless data points.

If you’re interested in more information about how and why standardized testing is ruining education, check out the article I wrote for the Washington Post last year. I like my original title better than the one they used: The Real Education Crisis.