What I Learned from my 3-Year Experiment with Project-Based Learning

I want to tell you the story of my journey in project-based learning and what it has taught me as a teacher an administrator.

In 2015, I took a job as director of E. O. Smith High School's Depot Campus. It's a program based on the Big Picture Learning model that focuses on learning through internships, advisories, and independent learning projects. I was excited. I had been experimenting with innovative models of education for years as a science teacher at the main campus. I'd created online and flipped classes, and had just finished what I considered my greatest achievement, a mastery-based, differentiated biology class. I saw the Depot program as an extension of what I'd been doing and an opportunity to continue to learn about and develop constructivist learning approaches like project-based learning. But as I walked past the several unfinished student murals on the walls of the little school, I should have known I was in for a challenge to my assumptions.

Year 1: Netflix and disillusionment

It is a pretty cool idea, when you think about it--kids engaged in things they're interested in. And if you take a look back at some of my blog posts from that first semester, you will pick up on the excitement I felt. But if you continue to follow my posts through the winter and into the spring, the story takes a twist. What I expected to see was students motivated by autonomy, mastery, and purpose and engaged in their independent work because they had the freedom to work on what interested them. But instead of working on their own projects during their daily "self-directed learning time," many watched YouTube or Netflix, played video games or just chatted. Only a couple of students ever completed the independent projects they had planned with their advisors that year.

I was disillusioned. What was going on?

Year 2: Hitting an invisible wall

So I gathered my team, and we all agreed that we needed a bit more structure. We reasoned that the students lacked basic organizational and other project-management skills, so we created "Project Block" to replace their "self-directed learning time." During this block, I would guide the students through a series of guided science projects, each one focusing on a different "project" skill, such as planning, troubleshooting, or research. Each project also allowed for student choice, and during the course of the year I tried a few different methods of tracking and monitoring student progress on their work. The goal of Project Block was that each student would successfully complete an independent project after the initial 8 canned projects.

While nearly all students successfully completed the first 8 structured projects, only a couple of students successfully completed the fully independent project. It was as if they hit an unseen wall when they were fully on their own.

Year 3: Back to the drawing board

Maybe what the students needed was guidance all the way through the project process. So instead of a series of canned projects, the following year I worked with my students in small groups to develop science project ideas and carry them out. One group built a maze and started running experiments to see if hamsters could learn the maze. The other decided to try aquaponics. They grew trout in a fish tank and circulated the water into a second tank containing lettuce plants. The fish water fertilizes the pants, and the plants clean the water.

Both were very cool projects. It was exciting at first, but as time went on, it became harder and harder to maintain the students' momentum for new learning. The hamster project team resisted my efforts to encourage them to do more research and improve their experiments. The aquaponics team never developed any experiments, and though they successfully grew both trout and lettuce, they spent most of the time during Project Block on their phones, YouTube, or chatting.

Year 4: A new hypothesis

During the summer after the third year, I read a new book by John Hattie and his colleagues called Visible Learning for Science. In that book, he addresses the nagging question of why inquiry-based learning, problem-based learning and other such approaches, which make so much intuitive sense, have such low effect sizes. In other words, why aren't they effective? His answer is that they are effective if they are used at the right time--after the students have learned the basics. They are useful methods for extending and deepening learning, but not for initial knowledge acquisition.

And so I had a new hypothesis. Students could not do independent projects, not because they lacked generic project skills, but because they lacked the fundamental content-specific knowledge and skills needed or the specific projects they were working on. So I began the fourth iteration of our project-based learning experiment. I decided to teach students the basics of electricity and electrical circuits using the CASTLE curriculum, and then have them each do an independent electricity project after they mastered the skills they needed.

Year 4, Take 2:

The CASTLE curriculum is really cool. It is a hands-on, self-guided curriculum in which students read instructions, set-up experiments, make and test predictions, and then read summaries of concepts. But about half of my students were not able to progress quickly enough. This could have to do with the reading level, but I had solved that by making a series of instructional videos, and students still struggled to progress through the curriculum on their own. The CASTLE curriculum, it turns out, is actually more of a constructivist approach, and I'm not sure why I thought it would solve my problem in the first place. I think I was driven primarily by the need for a canned curriculum that required little lesson prep.

In any case, it was failing for most of my students, so I decided to give them two options: Continue with the CASTLE curriculum, or switch to lessons delivered by me. I would put together mini-lessons and hands-on activities and projects on a variety of science topics selected by the students. All but one chose the lessons.

And thus the final nail was hammered into the coffin of the 3-year long experiment in project-based learning.

Some may say that it's just a problem with this particular population of students. Many of our students struggle with attention, hyperactivity, impulsivity, anxiety, depression, and other issues that affect their ability to stay on task, complete work, and in general, find success at school. But it turns out that this is exactly the kind of student for which the project-based approach is often offered as a solution. And what is the goal of the projects anyway, if not learning? How did the projects become and end in themselves? When did "independence" start to trump learning?

For me, the experience of the past three years has been a cautionary tale of confirmation bias. I started with the assumption of project-based learning and then went through several cycles of trying to justify it. Ultimately, like a creationist finally coming to terms with the evidence for evolution, I have to go where the evidence leads.

Year 4, Update:

It is now March of Year 4 of my experiment, and my new approach of explicit instruction followed by projects has devolved into pure guided project work. It proved too difficult to use an explicit instruction approach while having to create a curriculum from scratch and on-the-fly.

For several weeks now, we have been working on optimizing baking soda rockets, and the students love it. They look forward to Project Block, but I have to remind myself that having fun is not the same as learning. Video games are fun. We need both.

Looking ahead

At this point, there are two paths I am considering:

1. Spend the summer developing a science curriculum based on Hattie's layered approach: Explicit instruction first, then projects.

2. This may seem like it comes out of nowhere, but I am toying with the idea of trying Mike Schmoker's approach next year. In his book, FOCUS: Elevating the Essentials to Radically Improve Student Learning, he argues for a greater emphasis on literacy across the curriculum. I am considering a reading- and writing-based approach to science at the Depot. Not only would this be a way to help them acquire the foundation of science knowledge they need, it would focus our energies where they need to be focused. Many of our students are at-risk and behind in their basic skills, so building confidence in reading, writing and math could trump any benefits hands-on science could offer.

My current working hypothesis is that engagement is not about student interest, but about their sense of competence. And I think this holds true for reading, writing, and math as well as anything else. My real goal is to prepare these kids for college or career, and I want to do whatever will move them in that direction.

You might think that this entire four year sequence represents a retreat from independent and personalized learning, and if you thought that, you would be correct. But I believe we must go where the evidence leads, and we do what is best for the students, even if it means abandoning, or at least revising our beliefs about learning. As John Hattie says, we must not "hold any instructional strategy in higher esteem than students' learning."

It's good to be wrong

Depot students having a blast during fitness block.

"I feel like you treat us like delinquents. I feel like we're always being watched with suspicion."

We were sitting around the table, having a little conference. I had been sensing some negative vibes lately between the students and me, and so I called some of them together. The student who just spoke had been quiet, but he drummed up his courage and got right to the heart of the matter.

I was glad he did, because it's good to be wrong. That is, it's good to know you're wrong, because then you can change.

There had been several recent disciplinary incidents, and I had responded by taking away some privileges. On top of that, this year we've vastly increased accountability at the Depot. We added the CHIPS incentive system, ALEKS math block, competency ratings and progress tracking, and careful credit accounting. All this came together to generate lots of stress and negativity. The kids in the meeting had also complained about the Chips and how some students may not be able to go on the outing to Lake Compounce.

As I listened to what they had to say that day, I knew they were right. I knew the importance of focusing on support--that accountability is a poor driver of improvement, but I had lost focus. We had increased support along with the increased standards, but it wasn't enough, and according to our recent school climate survey, compared to last year at this time, more students felt like we don't care.

Time for a course correction. As Brian Tracy says, an airliner is almost never heading straight for it's destination. It's constantly correcting its course. It's constantly wrong, but the pilots know it, and they constantly adjust.

So I reinstated the privileges I had taken away. I reduced the Chips requirement for the latest reward outing, and finally, I made it easier to get math credit through the ALEKS system.

This last adjustment was the result of me looking more closely, after that little conference, at the sources of stress on the kids. I suspected that a few of the behavioral incidents and the slow accumulation of reward Chips were actually due to kids struggling with their math. They were nearing the end of their courses, and only the most difficult topics were left.

It may be coincidence, but since I adjusted the ALEKS requirements, there has been a notable change in behavior and attitude among a few of the students I had been concerned about, and spirits, in general, have been higher since the changes.

Lesson learned.

Being wrong is good stuff.

Change is tough: The slogging phase

"I'm done with my chores now. It's time. I say my prayer and head out on the hunt. The sun isn't up yet; it's cold; the fields are sopping. Brambles scratch my ankles, branches snap back in my face. The hill is a sonofabitch but what can you do? Set one foot in front of another and keep climbing.

An hour passes. I'm warmer now, the pace has got my blood going. The years have taught me one skill: how to be miserable. I know how to shut up and keep humping. This is a great asset because it's human, the proper role for a mortal. It does not offend the gods, but elicits their intercession. 

My bitching self is receding now. The instincts are taking over. Another hour passes. I turn the corner of a thicket and there he is: the nice fat hare I knew would show up if I just kept plugging."

-Steven Pressfield, The War of Art

Not sure exactly where I am in the hype cycle right now, but it's somewhere after the "Peak of Inflated Expectations." I'm in that phase when blogging about all the cool things that are happening, like our awesome Gallery Night and progress on our progress tracking system would seem empty. It would violate Ryan Holiday's rule of not talking about projects until they are almost entirely done. He says it's too easy to trade doing for talking about doing--to blog, post, tweet and brag about all the cool things we are up to and in the end produce nothing. Talk, Holiday says, is the enemy of work, and I'm in the work phase.

It's that phase when the brambles grab at your legs, your boots are wet and your legs are jelly. It's that phase when you wonder if it's all for nothing, and are tempted to let everything just slide back to where it was.

But it's also that phase when you can't stop--when you know you have to keep going, because you know in your heart and mind you are on the right track.

Last year was my first year in administration--my first year at the Depot, and I started with Barry Jentz and Jerome Murphy's Entry Plan approach. I rode over the top of the hype cycle right about then, disillusioned with the autonomy-based approach to education, and landed in the "Trough of Disillusionment." We did a program evaluation and immediately started up the slope with a set of changes we thought were needed.

It's tough not knowing where you are on that slope. I hope we're headed into the "Plateau of Productivity." There's a bunch of stuff in place, and data is coming in, but it's too early to tell for sure. So right now, we keep plugging, waiting for that rabbit to show up.

And I think it's right around the corner.

A rubric for the new year

At school, we set clear goals with students, establish clear criteria for success, and track progress with rubrics. Why don't I use this approach at home?

It's kind of funny to think about. Imagine pulling out a rubric during dinner time to score your younger children on their table manners, or pulling out a "relationship rubric" to talk to your teens about their relationships with their significant others (or to your wife about your own relationship).

But this is exactly the sort of thing we're trying to do at The Depot--using rubrics and setting goals for improving our students' noncognitive skills. We're trying to be intentional and objective about helping our students grow in important ways. Why should home be any different?


Some people really don't like rubrics at all. Maybe they feel like they are artificial. Maybe they feel like that's not how humans operate.

And maybe I don't talk about specific goals and criteria for success with my own kids when I discuss life skills with them, but maybe I should. Having clear expectations, rules, criteria for success, and feedback is essential to improvement, whether you're improving your basketball skills, your fitness, or your relationship skills.

Of course we don't want to make this all about scores or constant judgement, but using data doesn't have to be the way. It can be a powerful tool for avoiding natural cognitive biases that cloud our view of how we're actually doing. It's about being willing to look at things as they really are, and improving.


To various degrees I try to track my own stuff, like cholesterol, fitness, meditation, and reading. I recently discovered I had high cholesterol, so I started a series of experiments to see what might bring it down. I started with a baseline data point (here's our current paleo-primal winter meal plan), then I cut back on a bit saturated fats for a month. This meant skipping my coconut milk breakfast smoothies and cooking in vegetable oil instead of bacon grease. Then I got another blood test. Then I started up my old supplement regime again for a month, and retested. The results are in the figure on the right.

This month I'm drinking red wine every day. Since I saw a drop with the saturated fat reduction (the middle point on the graph), I'll kick it up a notch next month and try the Mediterranean diet (here's the tentative meal plan).


What if we had similar measurements for relationship variables and subjective well-being (actually, there are a bunch of ways we could measure these) and then came up with specific strategies to improve them?

Maybe it would look something like this: Hey, honey, how about we take this relationship test and then try something new, like going out every week, and take the test again later to see if things improve?

Or how about this: By the way, son, I've noticed you've been working towards some serious fitness and professional goals. Have you thought about setting any personal or relationship goals? Are you interested in some ideas about how you could track them?



I know there must be a balance between contentment and striving. There's a time to be content with who you are and where you are, and just be. But as long as we're talking about improvement, I know I need to be more intentional about it, whether at work or home. If there's something wrong with this approach at home, then why am I doing it at school, and if it works at school, then why am I not I doing it at home?

I don't plan on developing any relationship rubrics for myself or my kids right away, but I do plan on working towards more intentionality in helping everyone around me, including myself, improve. And maybe a little rubric for that wouldn't hurt.

Instead of a resolution, here's a rubric for the new year:

Just getting started	Developing	Meets expectations	Exceeds expectations
Bill offers to help his kids and wife establish new personal goals and tracking systems. He begins to work on a new tracking system for one of his own personal goals.	Bill offers to help his kids and wife establish new personal goals and tracking systems, and follows up with them during the year. He also establishes a new tracking system for one of his own personal goals.	Bill helps one of his kids and/or his wife establish new personal goals and tracking system, and follows through with them all year. He also establishes a new tracking system for one of his own personal goals.	Bill helps his kids and wife establish new personal goals and tracking systems, and follows through with them all year. He also establishes two new tracking systems for his own personal and relationship goals.

Continuously improving without constantly criticizing

Last week, two of our teachers were in my office talking about a new activity they were wrapping up in their advisories.  It was designed to check the students' progress half-way to their first Exhibition of the year, and it seemed to be working. Students were definitely more aware of where they stood, and how much they still needed to do.

I had liked the idea the week before, but I had become concerned about how long it was taking, so I spoke up: "My only question is, are we spending an inordinate amount of time on this?"

As if I really needed to nit-pik at this point.

Unfortunately, it was only the most recent example of me crossing over the line that separates actually supporting improvement and just being overly critical. It happens when I forget that continuous improvement doesn't come through knee-jerk criticism of everything I don't like, it comes through careful investigation and collaboratively taking one or two high-leverage actions at a time.

Interviewing a student

I'm grateful to my staff for being patient with me so far as I poke, prod and subject them to my "EntryPlan:" 39 one-on-one interviews with students, staff, and parents, three student surveys, three focus group questions, attendance reviews, requests for data, student work reviews and a month of engagement measurements.

It's all been a necessary part of my effort to get a good handle on where we are as a school so that we can be systematic, scientific, and intentional about improvement. But I know it's been a bit taxing, so the last thing I need to do is add unnecessary criticism on top of all that.

I can genuinely say my staff has done a phenomenal job over the past 7 years. The Depot is a place where students get more than a second chance: They find a home, a family, a school that works for them, and a place to pursue their interests, find success, build self-esteem and master the key skills they'll need for the rest of their lives.

But as in any school, there is lots of room for improvement. No matter how good we already are, as long as there are students here who haven't yet maximized their potential, we have work to do. I truly believe every single individual human has tremendous and unique potential, just waiting to explode, and I want to help make it happen for EVERY student (and staff).

This is a BIG task, and making it happen is going to take a deliberate, scientific approach, a lot of data collection, a lot of work, and a lot of courage. It's going to take us ruthlessly scrutinizing ourselves, not to judge or criticize, but to look at the data together and come up with one or two powerful ways we can improve, implement changes, and then repeat the process.

Meanwhile, remind me when I forget the difference between continuous improvement and constant criticism, and hey, let's enjoy the process together. Because improving yourself is actually lots of fun--it's challenging and fulfilling and can work right into an even bigger goal, like improving the world.

Let's be the change...