As 2015 comes to a close and students part from their classrooms to celebrate the holidays, I wonder how much of what they learned over the past year lingers. Do they look forward to returning?
Copyright © Ghica Popa
Indeed, this year marks the seventh year since the Common Core State Standards initiative launched. Looking back, it’s become clear that this universal framework of math and reading benchmarks has some pernicious bugs.
Denouncers of the Common Core often reduce it to the “One size fits all” epithet — all students are subject to a set of uniform standards. I prefer Peter Greene’s version, “All must fit one size,” which tells it like it is: a program bent on imposing a rigid criterion to measure academic success in America.
The Common Core asserts a top-down, quantitative model governed by big data. Policy makers determined that student and teacher performance is best represented by large datasets from which they can draw broad conclusions. And the engine that churns the data? Standardized tests. There’s a whole host of them, from the more familiar SAT and ACT to the NAEP, SBAC, PARCC, and the [insert valid regex match]. They assess a wide range of factors and serve many purposes, and many have a tremendous impact on teachers’ salaries and school funding.
As a result of standardized test scores’ influence on teachers’ livelihoods, educators are incentivized to teach to the test — that is, prioritize higher test scores over a substantial mastery of the material.
We have operated schools as if they were industrial factories, with teaching and learning practices that mimic assembly-line and batch-processing manufacturing.
In addition, many teachers are subject to an obscure evaluation algorithm. In New York, teacher effectiveness is calculated in a statistical model that takes 32 variables into account, many of which seem entirely irrelevant to the teacher’s actual performance. Many perfectly capable teachers are poorly evaluated by this model, hurting their chances for tenure.
A statistical model the school system uses in calculating the effectiveness of teachers. (New York Times)
But perhaps the most significant flaw with this system is the pervasive belief that every student must reach the same level. Students have vastly different resources than one another given their socioeconomic status and location. And every individual has a unique set of strengths and weaknesses, learns at a distinct pace, and is biased toward particular interests and passions. Common Core rejects heterogeneity.
It also marginalizes certain types of students. According to Pat Wingert the Common Core is “tough on kids who are still learning English.” Katharine Beals argues that the standards are “tough on kids with special needs.”
The data tells the same story.
It is doubtful that even the most ardent Common Core supporter will be satisfied if the best CCSS can offer—after all of the debate, the costs in tax revenue, and blood, sweat, and tears going into implementation—is a three point NAEP gain.
Empirical analysis of NAEP performance across states implementing the Common Core shows its ineffectiveness, according to a 2014 Brown Center Report on American Education. It’s evident that the Common Core “will have little to no impact on student achievement.”
Supporters of the Common Core argue that strong, effective implementation of the standards will sweep away such skepticism by producing lasting, significant gains in student learning. So far, at least—and it is admittedly the early innings of a long ballgame—there are no signs of such an impressive accomplishment.
With dismal prospects for the future of the Common Core, where can we look to for effective education reform in America?
All signs for student success point at personalized education. And it’s a simple premise, really. One size only fits one.
When learning a new sport, would you prefer a slew of lectures followed by an examination; or rather, a coach who guides you one-on-one, correcting your mistakes, leading by example, and challenging you individually to improve? I’d be hard-pressed to choose the former.
Why should the classroom be any different? Imagine an environment where students have the agency to set their own goals rather than chase an uncompromising standard. The curriculum is tailored to the individual, adjusting based on their interests, strengths, and weaknesses. And it’s meaningful because the student takes charge of their learning — it’s no longer about the NAEP or the PARCC; it’s about the student reaching personal goals at their own pace.
Personalization redefines the role of a teacher. Rather than looking at how their students perform overall against a desired statistic, teachers will instead zone in on each student individually, determining where they are, where they need to be, and how to get them there. This also simplifies teacher assessment: look at how the teacher facilitated each student’s path to accomplishing their academic goals.
Research also shows the promise of personalized learning. Look to Bloom’s classic 2 sigma learning studies, where students who were tutored at a one-to-one ratio achieved scores two standard deviations higher than those who had learned in a traditional school setting of a 30-to–1 student-to-teacher ratio (Bloom, 1984).
And in 2014, ninth-grade STEM21 students in 12 urban schools in Massachusetts and Connecticut who engaged in a personalized, online learning curriculum with in-person support showed significantly increased achievement compared with students who didn’t take part in the program.
But a human-centric personalization classroom is extremely demanding, logistically. There are simply not enough resources and time for a single teacher to engage one-on-one with each student in the classroom, following a distinct curriculum for each individual. Just picture the monumental burden that would place on a teacher.
Imagine an intelligent machine that can accurately assess a student’s strengths and weaknesses. It knows what the student loves and what they hate. It knows what time of the day they are most distracted, what kind of learning modules engage them the most, and how quickly they learn. It knows just about every idiosyncrasy that defines the learner. It gamifies the material so that the student has fun learning in byte-sized programs. And it follows the student outside of the classroom, finding relevant context in their experiences throughout life to incorporate into their curriculum.
This AI would redefine a teacher’s role even further. Since the AI facilitates the student’s curriculum and learning, the teacher can focus on ensuring that each student is achieving mastery of their goals. The AI can spotlight students where human assistance is prudent and exactly where they need help.
As great as this AI sounds, unfortunately, I don’t think we can expect this system within the next decade. Education is a challenging technical problem to solve — many subject areas are difficult to quantify and assess computationally. For example, The E.T.S’s automated reader that grades essays draws a great deal of criticism for its inability to fact-check statements and its tendency to reward verbosity. But as advances continue to be made in machine learning, computer scientists will surely overcome these hurdles.
With education, the stakes are extremely high. We can’t compromise on the future of America’s students and teachers. Not only is this is a problem that must be solved — it must be solved correctly. I look forward to a holiday season where students return home and the classroom goes with them; the joy of learning persists in their experiences throughout life.