Powerful Learning with PDSA: Path for Improvement Part 8

It's time for PDSA (Plan-Do-Study-Act) Cycle 2 in John Dues' journey to reduce chronic absenteeism in his schools. His team is using PDSA to quickly test ideas and learn on a small scale. Find out what happened and how PDSA can be a powerful tool for learning.

TRANSCRIPT

0:00:02.2 Andrew Stotz: My name is Andrew Stotz, and I'll be your host as we dive deeper into the teachings of Dr. W. Edwards Deming. Today, I'm continuing my discussion with John Dues, who is part of a new generation of educators striving to apply Dr. Deming's principles to unleash student joy in learning. The topic for today is Powerful Learning with the PDSA Cycle, Part 2. John, take it away.

0:00:26.7 John Dues: It's good to be back, Andrew. Yeah, like you said, we, I think for the past three episodes or so, we've been working towards getting a better definition of our problem specific to this chronic absenteeism issue that we're working on this year. I don't know if you remember from last episode, but we have this team working and they've basically said we don't have enough information quite yet to write this precise problem statement. So we decided to gather information running the Plan-Do-Study-Act cycle. And last time we focused on the first cycle. This episode, we're gonna focus on a subsequent PDSA cycle, sort of along those same lines. For folks that are watching and perhaps just joining for the first time, I'll kind of share my screen and do a little bit of a review so that everybody can see or know what we're talking about, even if they're just listening for the first time. So we've talked about this improvement model. We're working through this four step improvement model. So set the direction or challenge is the first step. Grasp the current condition is the second step. Third step is establish your next target condition and then fourth, experiment to overcome obstacles.

0:01:44.3 John Dues: And we're doing all this with a team, people working in the system. People have the authority to work on the system and someone with the System of Profound Knowledge knowledge. right. And so, you know, we've talked about setting that challenge or direction. And as we're grasping the current condition, we've actually decided to skip to step four and experiment a little bit so we can get a deeper understanding of this problem that we've been working on. And you'll remember probably as well, did the screen change for you so you can see the chart now?

0:02:21.8 Andrew Stotz: Yeah.

0:02:22.9 John Dues: Yeah. Great. So I thought it'd be helpful to show this again too. So this is our process behavior chart of the chronic absenteeism rates dating back to the 2016/'17 school year. So we have eight years of data in regards to this problem. And you'll remember when we talked about set the direction or the challenge, we wanna basically cut this chronic absenteeism rate we're seeing coming out of the pandemic by a lot. So we're hovering around this 50% chronic absenteeism rate. We wanna cut it to 5%. So that means, you know, 50% or more of our kids, or right around 50% of our kids are missing 10% or more of the school year.

0:03:06.2 John Dues: And this is a trend that we're seeing all over the United States right now. And the other thing that we talked about is when we looked at this process behavior chart, that it's basically like there's a pre-pandemic system of chronic absenteeism, and then there's a post-pandemic system of chronic absenteeism. So, you know, before the pandemic, the rates were too high, but nowhere near to where they are now. So, you know, prior to the pandemic, we were sort of hovering around the, you know, 20 to 30% of kids chronically absent. And then, you know, coming out of the pandemic, it's been more like that, that 50% number that we've, that we've talked about.

0:03:49.4 Andrew Stotz: And so to reiterate for the listeners or the viewers, this is the chronic absentee rate at your school, as opposed to nationwide, which I remember last time you talked about, it's about 30% nationwide, and pre-pandemic, it was about 16%.

0:04:06.5 John Dues: Yeah. Right around there. So, yeah, so I'm talking about the four schools that make up our school system in Columbus, Ohio.

0:04:15.8 Andrew Stotz: Yep.

0:04:16.6 John Dues: And, you know, we have a pretty high percent of our kids are economically disadvantaged. And so the rates in schools that have that demographic tend to be more like ours, in that 40, 50% range. And then, but all schools coming out of the pandemic had much higher rates than what they had pre pandemic. No matter your affluence levels. It's just, just like a lot of things the schools with the most kids living in poverty get hit the hardest when you have these problems, basically. So, yeah, yeah. So what we were, we were studying this problem, and, you know, we have some idea of what's causing our challenges, but we've started running these PDSA cycles to dig into that a little bit more, and I'll, I'll, I'll stop sharing. So that's not distracting. And so we ran this first PDSA cycle we talked about last time, and now we're running, or we've just gotten finished running a second PDSA cycle. So for folks that are new to that, what that means is that we are basically running an experiment to test an idea, an idea about how to improve chronic absenteeism.

0:05:26.6 John Dues: And to do that we plan the intervention, then we do or run the experiment, we study it and then we act on that information. 'Cause that's where the PDSA comes from. So basically the objective specific to PDSA 2 is were or we designed a individualized intervention based on responses we get from interviews with kids using this five whys sort of empathy interview template. Right? And then after we do that, what's happening is that students are actually. So after the five whys is completed with the student, we move right into creating the plan of the PDSA still with that student. So they're part of the process. So that's also sort of a key, I think innovation of this particular round of PDSAs is the student is sitting there as we design the intervention. A student that has some issues with chronic absenteeism. And then basically in this particular plan, we decided we're gonna collect detailed attendance data for two weeks to evaluate the effectiveness of that.

0:06:39.7 Andrew Stotz: When you said this one, are you talking about the PDSA one or two?

0:06:43.5 John Dues: Two. The one. The one you just got done running. The one we're talking about. So the PDSA 2 ran for two weeks. So when I say experiment, I'm not talking about, you know, like a randomized controlled trial that can last a year or two years or five years before you get the results. I'm talking about something you can do in a day, a week, two weeks. My general rule is not to go over a month with these PDSA cycles. It starts to feel like it's too long. I wanna get data back quicker than that on an intervention. And so that's what we did with this PDSA cycle 2. And it was really, the plan was built around this key question. The key question was, will involving students in the design of an individualized intervention to address their chronic absenteeism lead to an increase in their average daily attendance rate during that period of intervention. So we're not taking that for granted just because we're sitting with the kids creating a plan with them. We don't know, we don't know what's gonna happen exactly. And basically step one of that plan was this five wise interviews that I talked about.

0:07:50.2 John Dues: So basically we had four staff members. So each one was assigned a student at their campus that they chose to work with on this initial intervention. And they took a piece of sticky paper and up top they basically wrote, here's our problem, the student's name. So let's say James is not coming to school consistently. And when students miss a lot of school, they're at risk of falling behind academically. And right below that problem statement, then they wrote, why are you not coming to school consistently? 'Cause that's the first why question. So that's sort of the first part of this five whys interview. So it's very simple. You need chart, paper and marker in about 20 minutes to do this. Step two is, then they used the information that they gathered from that five whys interview to design the intervention with the student. And basically what they did was they designed the intervention around the root cause that they got to at the bottom of that five whys sequence. So basically, you know, when they said that, when they asked that first question, you know, why are you not coming to school consistently? The student is then going to say something, right? I miss the bus almost every day.

0:09:10.7 John Dues: And so the next question, the next why question is built on the previous answer from the student. So why do you miss the bus every day? And you kind of keep going. And it doesn't always happen perfectly. Sometimes it takes three questions, sometimes it takes a little more than five. But generally speaking, once you drill down with those five whys, you'll get to sort of a root cause from the interviewee, right? And so then they're basically saying like, you know, based on that root cause we identified, what do you think we can do to improve your daily attendance? And then now they're sort of transitioning from the five whys into the planning of the intervention. And sort of that was step two of the plan. And step three is then actually starting to track the student's daily attendance as they do whatever that plan is across the 10 school days that are in that particular cycle. So that's the plan phase. You know, we had a key question that we designed around, and then the team also makes predictions about what they think is gonna happen during that cycle. That's the plan.

0:10:23.5 John Dues: And then, so then they move into, once the plan's in place, you run the experiment. That's the do, right? And so in this particular do, PDSA cycle two here, that team collected both the quantitative data, so that's the five whys interview, and then the quantitative data, that's the daily attendance data. And so, you know, I mentioned that they had chosen four students to do this work with. And so what you're doing in the do is saying, did you do the plan? Basically. And that shouldn't be assumed because things may happen that interrupt the plan or derail the plan or make you change the plan. The guy that's designing these PDSAs is pretty good project manager. He knows improvement science, he knows the Deming stuff pretty well. So largely this experiment ran as planned. It's also for a pretty short time frame, so there's less time for it to go sort of off the rails and go wrong. But one of the student's attendance, it was so poor during this two week period that they never actually held the empathy interview. So you're noting stuff like that. You plan to interview four kids, you actually only interviewed three kids.

0:11:34.9 John Dues: And here's the reason why, we couldn't even get to the part of interviewing the student, this particular student, because he was not there during that two week period when they were gonna do the interviews. And so then the data comes in, right? The data comes in and now we're ready to study what actually happened during that two week period. So with the first student, interestingly, during the course of the PDSA, that daily attendance rate did go up. Right? And he was actually, he did miss a couple days, so it wasn't perfect, but he brought a note and there was actually a medical diagnosis. There was like a excuse reason for those particular days. Right. The second student was that student I was just referring to. So it was, his attendance was so poor that weren't even able to do the interview. That happens that you know, in real life. Third student's, this was a real success story. She was a part of her, the plan that she designed. She was meeting with one of the staff members at the school she attends that she chose one-on-one at the start of the day and she had significantly increased attendance during this two week period.

0:13:04.0 John Dues: And then the last student had this initial bump early in the PDSA and then had four straight sort of missed days at the end of that two week period also due to an illness, due to the flu. So you know, four kids, one uh two kids, pretty good success. Two kids still had some absences throughout the period. And then one kid sort of really didn't break the cycle. Nothing changed. And in fact the interview didn't even occur. Which means there was also no intervention that occurred with that particular student. Even so, with the student where we couldn't do the interview during that period, part of the plan was to get him a more reliable bus stop and we actually were able to change his bus stop, communicate that to the family and then that still didn't lead to increased attendance. And that's, you know, when you, I think when you work like this too, when you are sort of a policymaker or politician and you make these proclamations, we're gonna decrease chronic absenteeism or to cut it in half. And then you say, okay, well here's four students. Do that in two weeks. Do that over the course, just two weeks, just 10 days. How hard could it be? Right. [laughter] You see, you see just how hard these things are in reality. Right. So can't just make these proclamations.

0:14:32.9 Andrew Stotz: And is that part of what you're trying to do also in this process is get people to, you know, I mean, obviously what you'd love is to be able to come up with solutions from these four students. But really what you also are identifying is the other side of that. Wow, this is even harder than I thought.

0:14:50.2 John Dues: Yeah. And I, you know, with the people that work in schools, like the folks that are on this improvement team, none of this is gonna be a surprise to them.

0:14:57.3 Andrew Stotz: Yeah, they already know that, I guess.

0:14:58.6 John Dues: They already know.

0:15:00.3 Andrew Stotz: Yeah.

0:15:00.4 John Dues: Because they're on the front lines. I mean, they already know knew this is, you know, multi layered, challenging problem to address. But on the flip side, there were some pretty encouraging signs that getting, we call it getting proximate to students, individualizing some of those interventions and then including the student in that attendance monitoring and the intervention that it, you know, like I said, there were some success stories. So that first student I talked about, he went from 52% attendance to during the intervention period, it jumped up to 70%. So, you know, it's a short time frame, but a positive sign. That second student that I said that, you know, we never even got to do the interview with, he actually got quite a bit worse. So he had something like 37% attendance and that dropped 10% during that two week period. And then that third student that I said was a real success story, she went from 75% attendance to 96% attendance. And 96% attendance is good. And then the last student, she was the one that I said, you know, there was an initial bump and then kind of fell off, due to the flu at the end, she basically stayed the same. Like she was at right around 56% attendance and stayed right in that mid-50s range during the intervention.

0:16:19.8 Andrew Stotz: And just for, if a listener or a viewer has come in just on this episode and they're trying to understand where we're at, it's maybe you could talk a little bit about what you're doing in the sense that someone may look at it and say, wait a minute, this is just some anecdotes and how does that help you? And then on the other hand, one of the lessons in the business world that people sometimes say these days is do things that don't scale, which is counterintuitive. But what they're trying to get you to do is focus in, on getting it right with a small number of people before you then, you know, decide to go to the next level. So just maybe just give a brief of where it's at in this whole process.

0:17:04.8 John Dues: Yeah, I mean, I think. I mean, that's the whole point of the PDSA. The counterintuitive thing is that while you want to improve things for all students, you may start by working with a single student or a single classroom, or in our case, you know, four students. So, you know, the good thing is, is that what you could possibly do out of a cycle like this is, and this is only two weeks. But there's significant learning. And so what you would then do with cycle three, perhaps if the team decided this was what they're gonna do, is make some adaptation to this process, and if you were feeling confident that the adaptation was gonna work, then you could possibly. It's pretty early cycle two to start spreading this real wide. But I mean, you could spread this, you know, if you wanted to, to instead of four students, maybe you wanna try 10 students, something like that. Right. And in that way, and that's the basic idea, is to go from one student to one classroom of students to maybe one grade level of students to a whole school, perhaps if some type of idea is working really well. But the thing is, is that the whole mantra is, I lost my train of thought. Start small, learn fast. That's the mantra. Right.

0:18:23.7 Andrew Stotz: And then the other question I would have is, to what extent is this, you know, just subject to the Hawthorne effect in the sense that we knew in the many years ago that when they increased the intensity of the light, the workers did better, and then when they reduced the intensity of the light, the workers did better. So it was just that somebody was paying attention to the workers and the result was they appreciated that and so they did better.

0:18:53.9 John Dues: Yeah, I mean, that's possible. I mean, I would say...

0:18:57.0 Andrew Stotz: Which of course. Which of course may be the solution anyways. Right? You know, like, geez, if a teacher was... If each student was greeted by a teacher who cared about them and said, I'm so happy to see you, and it's great that you made it on time. Let's get started. I mean that could change the life of some people for sure. Including me.

0:19:13.7 John Dues: Yeah, I mean, I think, you know, like the girl that went from 75% to 96% attendance during the intervention. I wouldn't say, oh my gosh, we're gonna do this school wide now, but what I would say is, wow, that worked for that two week period.

0:19:25.7 Andrew Stotz: We're learning.

0:19:27.5 John Dues: It was relatively easy to do on a daily basis, so let's do more of that. Right. And also another thing you can think about is when you run PDSA cycles and you, let's say you do have really great success on whatever that thing is and you've spread it throughout your system. But now the priorities have changed. There's some other thing that really needs intensive intervention. One thing you can do is sort of as the 10th step in our improvement process is hold the gains. And so you often can then check back in on that thing in a month or two months and just see what the data says. Did it deteriorate because we're not paying as much attention to it or because we put that thing in place and left it in place? Although we're not as hyper focused on it, is it still sort of continuing on in a way that's at least acceptable to us?

0:20:22.3 Andrew Stotz: I just had an idea. Why don't we put a finish line and a countdown clock and a checkered flag and cheer every student as they get over the line?

[laughter]

0:20:35.0 John Dues: Yeah, that would be great. That would be great. The problem is, is the cheering isn't the thing that's preventing them from coming to school.

0:20:45.1 Andrew Stotz: Yeah.

0:20:46.1 John Dues: Maybe like the cherry on top, but you know.

0:20:49.4 Andrew Stotz: Exactly.

0:20:51.6 John Dues: So, so it's tough. Yeah.

0:20:52.2 Andrew Stotz: Great, great example of where superficial things that outsiders see may not really connect with the real trouble that they're struggling with. Yep.

0:21:01.5 John Dues: And remember, the student's not the only part of the process. The people, the staff that were doing these interventions, they reported that this was like a really powerful experience for them. Most of them haven't been through a PDSA cycle before.

0:21:18.6 Andrew Stotz: Right.

0:21:18.6 John Dues: So this.

0:21:20.3 Andrew Stotz: Maybe I was, slogans and exhortations is what I was just talking about.

0:21:24.8 John Dues: Right, right, right. Yeah. But they felt pretty confident that doing more of that thing was a good idea coming out of cycle one and then this second cycle. Now, with that being said, one of the things that we've also discovered is that there's some required intervening that has to happen as kids meet certain thresholds that are required by the state of Ohio. And so I think I mentioned this before, that one of the things that we're doing as a part of the project was mapping out the intervention process in place at each campus. So while the benchmarks are the same for when you need to do that, the how and the who and the when, that there's variation in that. So right now we're taking like each of the campuses mapped out their intervention process for attendance, and we're taking a look at that. And so what we're actually gonna do in the next cycle is work to put together a more standardized approach to these intervention teams. So, and that... And you get that from the team, you know, you're sort of talking about, you know, as we get to the Act.

0:22:38.4 John Dues: I think I've talked about this before. There's sort of the three A's. You can adopt this into your system, write it down in a manual or whatever, you can adapt it, change it a little bit, or if it's really not going well, then you abandon it. So in this case, there's an adaptation where we're gonna sort of shift gears and work on this process mapping and get that put in place 'cause the team feels like that's the highest lever, next thing to do, basically.

0:23:06.5 Andrew Stotz: And is that. Have you already mapped out PDSA 3 now? And you're in that process? Is that. And that's what we're gonna talk about next one, or where are you at with that?

0:23:15.5 John Dues: Yeah, so I think looking at the process maps would be interesting because you can sort of assume that things happen relatively similarly, you know, especially in a small school system like ours. But you're gonna see that there's... And it's not right or wrong necessarily, but there may be a sort of like, again, a better way to do things because there are many... There's sometimes many people involved, many processes, many forms. And so, yeah, we're gonna look and see, basically, is one more efficient than the other, is one more effective than the other, that type of thing.

0:23:51.4 Andrew Stotz: Okay.

0:23:52.3 John Dues: Yeah. So a lot of learning, you know, I mean, I think, and you know, again, this takes time. But, and we're still sort of in that "define the problem" step. But again, the fact that we're studying the problem, the fact that we're talking to kids as a part of it, we're actually learning from data coming from what's happening on the ground. The ultimate solution or set of solutions we come up with are gonna be much more robust, you know, and they're gonna be much more sort of durable into the future, anti-fragile you know, into the future because of this work versus, you know, the typical fly by night. Let's have a pizza party on Friday to encourage kids to come to school, that just... There may be an initial bump, but it's just, this has no durability into the future. Yeah.

0:24:44.8 Andrew Stotz: Great, great. Well, I like that. And I got excited 'cause I thought, oh, maybe we can come up with some incentive or something. But what I see is that the challenge is how do you make it durable? So I like that word. And this was a good discussion on that.

0:25:01.8 John Dues: Yeah. Well, I was just gonna say the other point I would make is, you know, sometimes you can plan, plan, plan, but that, you know, at any point of an improvement process, you can use a PDSA cycle, start running an experiment and start gathering information right away. Just do something, you know, thoughtful, but do something. Don't plan in a room when you can go out and get actual, run actual experiments and get back real data.

0:25:27.5 Andrew Stotz: All right, and one other question I had that just came to my mind is what is the value of doing this in such a structured way as the PDSA versus Oh, come on, John, we're testing things all the time, you know, and let's say that to some extent they are, right? We all are testing every day. Why is it important that it's done in a structured way, in a documented way?

0:25:50.8 John Dues: Well, I mean, one, you have a historical record. Two, you know, we had four. So there's one PDSA cycle, but actually there's four different mini experiments happening, one at each campus. And if you didn't write it down like the guy does that's designing these, every one of the experiments would have been different. And he really thinks through step by step. Okay, put this number here. How are we gonna define that? There's a validated list of definitions of that type of attendance issue that day. Because if you don't have all that stuff, then there's no way to analyze it after the fact, or at least it's a lot less, a lot less efficient.

0:26:28.3 Andrew Stotz: And the learning could be lost too.

0:26:30.1 John Dues: Learning could be lost. And you know, I would venture to guess that, you know, if you run a sort of an experiment haphazardly, especially if there's multiple locations, the people sort of, their definition of the thing that's being measured is gonna vary too.

0:26:45.0 Andrew Stotz: Yeah.

0:26:45.5 John Dues: Almost every time. Almost every time, even subtly.

0:26:50.9 Andrew Stotz: All right, well, that was a great discussion. And on behalf of everyone at the Deming Institute. I wanna thank you again for this discussion and for listeners, remember to go to deming.org to continue your journey. You can find John's book, Win-Win, W. Edwards Deming, The System of Profound Knowledge and The Science of Improving Schools on Amazon. This is your host, Andrew Stotz, and I'll leave you with one of my favorite quotes from Dr. Deming. People are entitled to joy in work.