In 2017, I shared an idea for a post with my small circle of confidants. The response was immediate and unanimous, “Don’t go there, Charlie.” Last year, I shared the same idea with a wider circle of family and colleagues and again, “Don’t go there, Charlie.” But with the world already turned upside down in 2020, let’s go there.
“All Children Can Learn” has been a mantra in K-12 education for as long as I can remember. Some trace the concept, if not the slogan, back to the 1979 article “Effective Schools for the Urban Poor” by Ronald Edmonds. (I’ll circle back to that article later.) Versions of the slogan are evident in the names given to the last two reauthorizations of the Elementary and Secondary Education Act of 1965: No Child Left Behind in 2001 and the Every Student Succeeds Act in 2015. Those monikers stand in contrast to the name given to the 1994 reauthorization, the Improving America’s Schools Act.
Given its role in shaping the assessment and education policy and practices that have driven our profession for the past two decades, perhaps it is time to take a closer look at how the slogan has affected assessment and accountability practices, and at the same time consider what assessment and accountability results tell us about the effectiveness of the slogan as a driver of education policy.
In the spirit of the sense of urgency implied in the distinction between “Black lives matter” and “All lives matter,” the February 2017 Kappan shifted the focus from “All student learning matters” in an issue titled “Black student learning matters.” Can we do the same for assessment and accountability and ask what our policies and practices really mean with regard to Black students, students with disabilities, English language learners, and students with economic disadvantages?
All Means All – ternative Facts?
Questions about the usefulness of “all children can learn” as a driver of education policy are not new. In an October 2003 Educational Leadership issue on Teaching All Children, Denis Doyle asked
What are we to make of the claim that all children can learn? Is it just one more slogan? As an empirical statement, it either means too much or too little. Of course, all children can learn, but the tough questions are, what can they learn, when can they learn it, and what format is best suited to learning.
In assessment and accountability, we apply the slogan “all children can learn” very literally in policies that call for all students to be proficient, all students to be college-and-career ready, and for all students to demonstrate what they know and can do on state assessments administered to all students. In practice, however, the implementation of those “all students” policies is more nominal than effective (in every sense of the word). There was NEVER actually a requirement in NCLB that all students be proficient by 2014. Under ESSA, states have set different long-term goals (and longer-term goals) for different subgroups of students. Even the highest-performing states begin each school year knowing that only half of their students will be classified as college-and-career ready or on-track to college-and-career readiness. In the lowest-performing districts and schools, educators begin each school year knowing that only 1 out of 10 students will be classified as on-track at the end of the year.
It should come as no surprise, therefore, when
- Absolutely nobody inside or outside of education believes that the goal of 100% of students proficient or college-ready is attainable or enforceable.
- Urban school districts in a state demand “bonus points” in accountability for helping their student perform above the lowest achievement level.
- Although urban superintendents are often staunch supporters of high standards for schools and students, one urban superintendent gains the support of the media, several former state department of education officials, and ultimately the state legislature by claiming that it is unfair to hold their students to the same graduation requirements as other students in the state?
- Advocacy groups for students with disabilities and English language learners support rigorous standards as a lever to improve instruction but balk when those standards are applied to individual students.
“What can they learn, when can they learn it, and what format is best suited to learning.”
One Size Fits All -most No One
One paradox or consequence (intended or unintended, I’m not sure) of education policy based on “all students can learn” has been the shift from the inappropriate practice of tracking students based on subgroup membership to the also inappropriate practice (I’ll hold back on calling it equally inappropriate) of placing all students in a single track; that is, expecting all students to meet the same set of academic standards in multiple content areas every year from the third grade through high school, and as stated above to demonstrate achievement of those standards on the same test. Although there is arguably a core set of knowledge and skills that all students should acquire in critical content areas, where that set falls within students’ K-12 experience and when/how to assess students is still open to debate and empirical investigation.
Another offspring of the “one-size-fits-all” approach is the concept that if we design assessments and educational experiences to meet the needs of the most affected or underserved students then all students will be better served. One example is the cartoon show below that my dear friend and colleague, Michael Hock, would share when discussing universal design for learning (UDL). Another example came in a recent NPR interview in which a public school official wondered whether designing programs to ensure that schools better meet the needs of Black males would result in improvements that benefit all students.
On paper, there is a lot to gain from the ideas expressed in both examples. In practice, however, we know from experience that the flexibility and personalization that are the hallmark of both examples will evaporate when a state develops top-down guidance on how to implement policies and practices at the district and school level. Districts and schools interpret guidance and examples as “this is the way to do it” or “this is the way the state wants us to do it.”
The last thing that we want at this point in time is to assume that a practice or program that works for students with disabilities or Black males or any other group of students is the best approach for all students. As stated in the Kappan article, Teaching students of color: Looking race in the face, “Programs aimed at meeting the needs of ‘all students’ …don’t necessarily address the needs of black youth challenged by poverty.” (p. 24). Or as Taylor Swift put it in cardigan, “A friend to all is a friend to none.”
Stand and Deliver
What does “all children can learn” and assessment/accountability policy look like at the teacher and student level?
Throughout the ten-year run of the New England Common Assessment Program (NECAP), we had an ongoing discussion regarding whether and how student demographic information should be included with assessment results on individual student roster reports prepared for classroom teachers. There was little debate about the potential value to teachers of aggregate subgroup reports, but the benefit of providing demographic information for individual students was less clear. The issue took on added significance as technology made it practical to deliver those reports via an app that enabled teachers to easily filter and sort their rosters on factors included on the reports.
The catalyst for this conversation was the reporting of an individual student’s free-and-reduced price lunch (FRL) classification, which was the proxy used to define the economically disadvantaged subgroup. Based on interpretation of USDA regulations (yes, USDA, not USED) it was not clear whether it was permitted to share that student-level information with classroom teachers.
Our discussion quickly shifted, however, to broader questions such as
- Why report student demographic information to teachers?
- What do we expect or want them to do with it to improve instruction?
A compelling reason offered in favor of providing the data, particularly during the ill-conceived and ill-fated Obama-Duncan-Biden educator evaluation era, was that if teachers are being held accountable for subgroup performance they have a right and a need to know which students are in each subgroup. It also became obvious that with the possible exception of the FRL classification, teachers already knew which students belonged to each accountability subgroup. None of that, however, addressed the question of how teachers should use the information to improve instruction, student learning, and ultimately, student performance.
In an attempt to answer that question, we laid out the following scenario as a thought experiment:
In a fourth-grade class of 18 students, there are 6 students who are not performing well in class (and on the state assessment) in mathematics. Will access to demographic information improve the teacher’s ability to plan instruction for each of those students?
Will/should mathematics instruction be different for a female/Black/non-economically-disadvantaged student and a male/White/economically-disadvantage student? The short, but not particularly helpful, response is maybe, maybe not. The more nuanced response is that the teacher’s instructional decisions for individual students should be based on a much deeper and richer understanding of the student than coarse demographic labels can provide. A worst-case scenario is that teacher decisions for individual students are inordinately affected by stereotypes associated with those demographic labels or perhaps by group-level performance –which can be our seemingly objective, quantitative version of stereotypes.
I recently re-watched Stand and Deliver on Netflix. The first takeaway, of course, is that ETS, its people, and by extension, our profession is the manifestation of pure evil – I was almost convinced this time. The more important takeaway is the importance of the relationships between teachers, students, and families; a point that has been central to virtually every, movie, book, podcast, or anecdote about teaching, and even shows up in research. As with large-scale assessment results themselves, it’s is difficult for me to envision large-scale data of any kind providing teachers with information about individual students that they shouldn’t already know.
We Can and Must Educate All Children
I first read Ronald Edmonds, Effective Schools for the Urban Poor, in the spring of 1981. In that first education course I took during my senior year at Harvard we also read Inequality by Jencks et al. and several publications related to the Coleman Report. Like the Bloom article that I wrote about in my last post, Edmonds described conditions (or in Bloom’s terminology, alterable variables) that lessened or weakened the correlation between family background and student achievement. Edmonds also offered his set of answers to the questions that Doyle posed a quarter-century later about what students can and need to learn and the conditions best suited to learning.
My belief in 1981 and my belief today, is that schools, particularly public schools, can make a difference. I also adhere to the belief that schools can overcome, or provide opportunities for students to overcome, a significant amount of inequality in students’ circumstances and inequity in society – however, certainly not all of it, and certainly not on their own.
As Jim Lovell (or Tom Hanks, if you prefer) said about the Apollo 11 landing on the moon, “From now on we live in a world where man has walked on the Moon. It’s not a miracle; we just decided to go.” I have argued for years that the problems we are trying to solve in education are much more complex than simply landing a person on the moon, but they can be solved. We just have to decide that it is important enough to do so. I urge you to read the entire Edmonds article, but I leave you with these two quotes:
There has never been a time in the life of the American public school when we have not known all we needed to in order to teach all those whom we chose to teach.
It seems to me, therefore, that what is left of this discussion are three declarative statements: (a) We can, whenever and wherever we choose, successfully teach all children whose schooling is of interest to us; (b) We already know more than we need to do that; and (c) Whether or not we do it must finally depend on how we feel about the fact that we haven’t so far.