KIPP Empower Academy opened in 2010 with an instructional model focused on small-group instruction, a strategy that had produced strong results at its principal’s previous school in New York City. With over 90% of KIPP Empower’s students qualifying under federal government rules for free or reduced-price lunch and 10% qualifying for special education, the team implemented a blended-learning model in which students rotated between small-group and personalized online-learning stations.
The kindergartners began with a 90-minute reading block, in which one-third of the class met in small groups with the lead teacher, another third worked in small groups with an intervention teacher, and the other third worked independently on individual computers. Every 30 minutes, the groups rotated to the next station. The children continued through the day with roughly the same rotational design for writing, math, and science.
KIPP Empower’s theory for providing students with small-group instruction was that it would be “right-sized” for their level and that, in a smaller group, the students, many of whom were learning English as a second language, would have more opportunities to speak and participate.
Whereas 61% of KIPP Empower’s students were “below basic” level on the Benchmark STEP Test in the fall of 2011, a full 91% were at the “proficient or advanced” level by the spring of 2012. The following school year, in 2012–13, the students continued to produce results that were as astounding as the gap they had faced. At the time, California’s Academic Performance Index (API) rated schools’ academic growth based on results from statewide testing, with 1000 as the highest score and 800 as the targeted score. KIPP Empower tested at 991.
According to the school’s website at the time, “Though many students at KIPP Empower Academy entered kindergarten without basic letter and number recognition skills, by the end of the year, 98% were reading and performing math at or above the national average.” Not only that, but many students were reading at a “2.5” grade level and performing math at almost the 3rd-grade level.
Yet in Mike Schmoker and Timothy Shanahan’s recent piece in Education Week, “Small-Group Reading Instruction Is Not as Effective as You Think,” the small-group instruction in the school was “squandering learning time and reducing the odds that the most vulnerable students learn to read, speak, and write effectively.”
So, who is right?
This seems like a classic case of drawing a broad conclusion at too high a level—and ignoring anomalies on both sides to understand the circumstances under which one approach or another may work better.
To be fair, Schmoker and Shanahan do admit that “small groups can be effective in certain circumstances” later on in their piece—but then say that “any advantage is wiped out by the model’s drastic reduction in the amount of instructional dosage.” And it’s certainly possible that KIPP Empower’s students experienced gains despite the small-group instruction. But either way, it stands out as an anomaly to what they’re suggesting.
What are the possible things Schmoker and Shanahan could be overlooking?
For starters, their comparison to their whole-class instruction scenario is one in which students are, in their words, engaging in “low-value or ‘cut, color, paste’ activities and worksheets. They keep students busy but do little or nothing to improve their reading. There are no substitutes for an explicit, outcome-driven lesson replete with multiple cycles of guided practice, checks for understanding, and targeted reteaching.”
It’s hard to disagree with them here.
As Adam Porsch, a senior advisor at Amira Learning, an AI-based program designed to help students learn to read, said, “Overreliance on small-group instruction without clear instructional purpose or sufficient dosage is a problem.”
Perhaps then the problem is less the time in small groups and more what one does during individualized practice. Is the work in those time blocks purposeful and designed to reinforce—or fill in the gaps—to what’s being learned from the teacher?
As Mark Angel, the founder and CEO of Amira, told me, “The essence of the argument is coherence. If teachers do instruction from the core and then Supplemental Instruction and Practice, then there are random exercises in unrelated skill building, the sum is LESS than the parts.”
In other words, the supplemental instruction should focus on ensuring students master the specific knowledge and skills of the lesson. Allowing students to do this at an online-learning station, where they receive real-time, specific, actionable feedback, can be a much higher-value-added activity than the cut-and-paste worksheets on which Schmoker and Shanahan base their arguments. What’s more, with a robust digital program, teachers can receive data on students’ performance to understand where students are and aren’t struggling, allowing them to give specific students more attention and support—differentiation that’s almost impossible with a whole-class setup.
Could this deliberate practice—which, along with response to intervention, is in education researcher John Hattie’s top-20 list of effective factors related to student achievement—be done for homework rather than use up class time?
Theoretically, yes, but we know that students often don’t do homework and that digital programs are frequently underutilized unless a classroom makes them a core part of the learning design. Given that we’d never have students skip deliberate practice on the skills they’re seeking to master in music, sports, or math, using a Station Rotation model embeds deliberate practice in the classroom so the teacher can ensure students are doing the work.
What’s more, as Angel told me, “The research shows that differentiation must be based on supplemental instruction and practice aligning with the core instruction. The core may prescribe work on Skill X, but if a student doesn’t know the prerequisite skills for Skill X, the Tier 1 instruction will fail.”
This is where a digital program in a Station Rotation can be a game-changer. And it’s the flaw behind a blind focus on instructional dosage independent of student mastery.
As Angel continued, “The job of supplemental instruction is to differentiate by enabling the student to work on the foundational pre-reqs. Similarly, after instruction on Skill X, practice needs to center on consolidating the instruction on that skill, not a skill that won’t be taught for three weeks down the road. The goal is Tier 1 (teach to the grade level scope and sequence), Tier 2 (scaffold students who need additional assistance so that they can do the Tier 1 work, and Practice the Skill (all students should independently exercise the skill just taught).”
On top of that, digital programs that are aligned with the teacher’s and school’s instructional goals can also deliver direct instruction—meaning that time need not be further “wasted” on low-value activities or on losing out on valuable instructional minutes to the exclusion of phonics or knowledge-building.
Either way, it’s significantly better than aimless play that has nothing to do with core instruction—the counterexample on which Schmoker and Shanahan base part of their argument for whole-group instruction.
Indeed, independent evaluations by Columbia University and the Utah State Board of Education found that students using Amira regularly gained reading skills equivalent to five months of additional instruction, with effect sizes of 0.45. These results, while not fully randomized controlled trials, suggest the power of a robust digital program like Amira, coupled with small-group instruction that allows teachers to create a more active learning environment for early literacy skills, yielding gains on par with one-on-one human support.
As Julie Young, the founder of the Florida Virtual School, who started her career as an elementary school teacher, told me, “I don’t see this as an either-or issue. Effective literacy instruction needs both whole-class and small-group approaches. Whole-class lessons create shared understanding, build vocabulary, and strengthen a sense of community. Small groups allow teachers to focus on individual needs and give students the confidence to take risks and grow.”
This isn’t an argument for divvying students up based on flawed notions of leveled reading. It is an argument for increasing active learning, ensuring students master the skills they’re working on, and providing the direct instruction they need.
