This post is by Susan Fairchild, Vice President of Knowledge Management and Chief-of-Staff, New Visions for Public Schools, and Catrin Davies, Associate Director of Institutional Giving, New Visions for Public Schools.
For the last year, we have worked with Principal Stacey King and her staff to re-design the attendance system at New Visions Charter High School for Advanced Math and Science II (AMS II) in the Bronx. This post is the fourth in a multi-part series on attendance systems. At the beginning of the school year, we predicted that AMS II’s graduation rate would be anywhere from 91 to 94 percent based on New Visions’ graduation models. To hit the upper bounds of our predicted graduation rate for Class of 2016 and to sustain high performance in future years, the attendance system at AMS II would need to improve.
But re-designing a high school attendance system is harder than one might think. In part one of this blog series we highlighted the hidden complexity of attendance and the consequences of not managing it well. Next, in part two, we outlined AMS II’s continuous improvement strategy; and, in the third blog post, we documented how improvements to AMS II’s attendance system were the product of common sense changes made by school staff over the last year. In this blog post, we reflect on the costs of absenteeism, the findings from our improvement efforts and how to scale what we’ve learned.
Revisiting The Problem
As we’ve seen, daily attendance is complex. At AMS II, school staff monitor the movements of 546 students, which translates into nearly 100,000 data points over the course of a school year. Even in achieving a 90 percent attendance rate, AMS II will have 10,000 data points (absences) that require immediate administrator attention and response. When this level of complexity is not managed well early on, the rate of erosion in attendance is visible.
One obvious consequence of accumulated absenteeism is the impact on student learning. If students don’t show up to class, they miss valuable instructional time. For every absence, that’s about 6.5 hours of lost instructional time. At AMS II, there were 4,821 total absences by the end of April which translates to 31,227 hours of lost instructional time. A weak attendance system sabotages a school’s instructional strategy and undermines teachers.
Weak systems take real effort to change. Newton’s First Law of Physics--every object will remain at rest or in uniform motion in a straight line unless compelled to change its state by the action of an external force--certainly applies to student absenteeism. In the case of AMS II, Principal King had to apply multiple external forces in order to stop the acceleration of absences in the school. Principal King did this by creating “friction” through a series of attendance interventions that decreased the building momentum of absenteeism across the school. At the same time, the attendance team exerted substantial force to get attendance moving in a new direction.
So, How Did We Do?
Once we understood the attendance problems, efforts to re-design the attendance system went into effect at the end of February and in early March. Our goal was to streamline the attendance recording and outreach system, to enable more time for outreach to families. As seen in Figure 1, outreach activity improved dramatically from a low of 5 percent in January to consistently and reliably contacting parents 80 percent or more of the time in April, May and June.
Figure 1. Student absences vs. parent outreach
Principal King and her staff were aiming for a graduation rate between 91 to 94 percent. As of June their graduation rate reached 94.8 percent and will increase by the end of August. The average attendance across all cohorts for AMS II during school year 2015-2016 was 90.4 percent and we see variability across the four cohorts: 87.1 percent (Cohort 2016), 85.8 percent (Cohort 2017), 94.8 percent (Cohort 2018), and 96.5 percent Cohort 2019 These cohort attendance rates played out differently as the year progressed, as shown in Figure 2.
Figure 2. Student attendance by cohort
Note: This graph includes attendance data for state exit exam (Regents) testing dates which typically happen the last week of school. The previous graph (Figure 1) does not include this week.
The majority of the graduating class (represented by the blue line) entered their senior year already meeting New York City graduation requirements. The attendance system simply was not enough to keep these students in school in May and certainly not in June, a month that is primarily dedicated to high stakes state exam testing. This raises an important question--how do we design senior year to better support college readiness (only 25 percent of AMS II’s graduating class met the college ready thresholds set by City University of New York)?
A second question we are considering is: To what extent does one cohort’s attendance patterns influence another? We ask the question because of the similar attendance trends between cohort 2016 (blue line) and cohort 2017 (orange line).
A third question is: how do we continue to tighten the AMS II attendance system by smoothing out the troughs and peaks that we see in attendance trends? For example, AMS II’s attendance system helped increase the attendance of cohorts 2016 and 2017 from February to April. But, is it possible that efforts to increase the attendance rates of the Juniors and Seniors (see the spike in April) also diverted attention from the two younger cohorts (the green and red lines) during that same period?
Opportunities for Scale
The ripple effects of a weak attendance system may make it the single most important system to get right in a school. In New York City, we need a tool that will build off of what we have learned through this AMS II case study and help us to quickly diagnose attendance systems in other high schools. We are working on a prototype that is based on New Visions’ Attendance Systems Design Framework. Our framework integrates key policy and protocol documentation that was generated by AMS II and analyzed throughout the year as improvements were made. The prototype is divided into five sections: 1) Attendance Trends, 2) Goals of the System, 3) Attendance System Configuration, 4) Evaluation of Effectiveness, and 5) Research on Attendance. Using AMS II as a model in the mock-ups below, we detail the information necessary to fully understand the strengths and weaknesses of a given high school’s attendance system.
The tool provides aggregate data that allows one to see when and where a school’s attendance system might be struggling. For example, in Figure 3, we are looking at trends of different subgroups of students within a cohort who are attending school 90 percent of the time or more over the school year. Interactive visualizations allow us to cut these data any number of ways so that we can pinpoint where to target improvement efforts, whether in the system itself or with particular subset of students.
Figure 3. Attendance trends
Goals of the System
This section examines the school’s policies and protocols that explicitly state the school attendance goals, the teacher attendance goals, and the student attendance goals. In addition, the tool includes descriptions of how attendance expectations are communicated to staff, the type of attendance interventions and thresholds for triggering action. Here, the observer would rate the status of each component against the Attendance System Design Framework.
Figure 4. Goals of the system
Attendance System Configuration
Observations at our schools suggest there are three major categories an attendance system must address: Data Collection Activities, Verification and Compliance Activities, and Intervention and Communication Activities. Yet, in the way each school configures its system, these may look very different. The example below reflects the specific decisions AMS II has made in the design of their system and when these actions occur. Observations reflect potential opportunities to re-think or improve some of those configuration decisions.
Figure 5. Attendance system configuration
This section of the prototype emphasizes the importance of reviewing process data to ensure that interventions are happening as expected. In the figure below, we highlight AMS II’s data monitoring system, which linked the number of parent outreach attempts to number of absences and tracked those data consistently. From April to June, outreach was hitting over 80 percent.
Figure 6. Evaluating effectiveness
Research On Attendance: What We Know
Attendance is one of the more documented components of early warning systems. Schools and their partners should have easy access to broader literature on this topic to inform system improvements. This section of the tool integrates key findings, including the impact of chronic absenteeism, and attendance patterns of dropouts, and what we know about specific intervention strategies.
Figure 7. Research on attendance
In sum, we plan to build off of the important contributions of AMS II to the field in understanding and fine-tuning high school attendance systems. These improvement efforts begin by acknowledging the complexity and fundamental role of attendance systems in schools. AMS II is a high performing school, yet it has taken considerable time and effort to improve their attendance system. With other strong systems in place, Principal King and her staff were able to focus deeply on attendance and tighten a previously slack system by making their process more efficient and without diverting resources from other places. This may not be the case in a different context. That is, if improving attendance in a high performing school is hard, you can bet it will be an even greater challenge in a lower performing school. But, an important first step in this improvement process was Principal King’s willingness to examine the anatomy of her school’s attendance system--for all to see--in a culture of high stakes accountability. What is more, she has opened up their lessons learned to a national audience through this blog series.
The opinions expressed in Learning Deeply are strictly those of the author(s) and do not reflect the opinions or endorsement of Editorial Projects in Education, or any of its publications.