Should Reading Be Taught Whole Class or Small Group?

Blast from the Past: This blog first posted on April 28, 2018, and was re-posted on November 23, 2019. It is now being reposted again (and updated) on June 7, 2025. Original content has been adjusted, new content has been added, and the original posting included no research references.

Teacher question 2025: I respect your opinion and would love to get your thoughts on some new information that is out right now. In our community the idea of providing core foundational skills instruction to small groups is being promoted as the “science of reading.” Our schools teach foundational skills on a whole classroom basis and then use tiered intervention time for filling gaps. This has been highly effective. Could you help to clarify this issue?

Original Teacher Question: I was curious what your thoughts are regarding small group instruction in elementary school during the ELA block. I’m unaware of any definitive research on the effect size of small group instruction or the impact it has regarding student achievement in reading. There seems to be a few different schools of thought: direct whole group instruction for all components of reading, shortened whole group reading followed by differentiated small group instruction, whole group instruction followed by student work groups facilitated by teacher walking around. It seems all three could be effective depending on the students, the teacher and rigor of text or content being used.  However, I’m curious if there is a research-based recommendation?  

Shanahan’s response:

Small group teaching is ubiquitous in elementary reading classes. That was true when I myself was being taught to read (65 years ago), though our classes then were so large that “small group” meant groups about the size of today’s typical class.

By the beginning of the 20^th century there was much within-class grouping, but that was due to the pervasive one-room schoolhouse. Those “groups” were the “grades.” My dad, a product of such a school, bragged for years that he “graduated at the top of his class”—meaning the other kid flunked.

When I was first exploring the idea of becoming a teacher, the lore of the time was that reading teachers always had three reading groups: the Robins, Blue Jays, and Crows. As a teacher’s aide it certainly looked that way to me, and I wasn’t surprised during my student teaching when Mr. Krentzin had me take over those leveled reading groups one at a time.

Admittedly, as a primary grade teacher, I always grouped my kids for reading instruction. Studies have long reported that more than 90% of primary grade teachers group for reading, with high incidence in the upper grades as well (e.g., Austin & Morrison, 1963).

Historically, small group reading instruction has been aimed at matching kids with books of certain levels – a practice the effectiveness of which is refuted in my forthcoming book (Shanahan, 2025). As that 2025 letter above illustrates, now small group teaching is as likely to be aimed at phonics – and not as a follow up for the stragglers, but as initial teaching.

Where this valencing of small group literacy teaching will end is anyone’s guess. Kids need instruction in words (phonemic awareness, phonics, spelling, morphology), fluency, comprehension, and writing. Perhaps we could have 3 groups for each. For a 2-hour block that would allow for 10 minutes of teaching of each topic to each group each day. Future research could sort out how long it would take such an approach to drive the average teacher insane!

This blog entry was originally written with a focus on those typical leveled reading groups along with an odd practice that was appearing during the 2010s. Schools were (and are) organizing their schedules around “small group time.” Schools that require a period of small group teaching are mindlessly promoting the value of such teaching — no matter the situation. I was seeing identical lessons taught to one small group after another to accomplish a principal’s or curriculum coordinator’s small group mandate.

I thought this recent shift to foundational skills by small group advocates would require a whole new response from me. But a careful review of the research shows this to be unnecessary. Their use of the term “science of reading” must only refer to their commitment to phonics – and not to any research supporting their small-group approach.

I had largely completed this rewrite when I came across a quote that brought me up short. “Research has shown us that the traditional whole-class phonics lesson is not the way to develop fluent readers – not for kindergartners or, in fact, for any students in any grade level” (Seger & Stukey, no date).

I hunted for that research but, of course, there wasn’t any. People make those kinds of assertions and then cite sources that don’t actually support the idea. In this case, a paper by Nell K. Duke and Anne E. Mesmer (2018-2019) was cited which explicitly recommends teaching phonics whole class with small group follow up as necessary – a completely different idea than what that quote seems to promise.

Does small group teaching advantage kids in learning to read?

John Hattie identified three meta-analyses on small group instruction and reported it to have a medium-to-large effect (.49) on learning. However, few of the studies in those meta-analyses focused on reading, they weren’t always comparing small group teaching to whole class instruction, and some of the reading studies were from back in the day when 45-60 kids in a class was common practice. Several of the studies were based on secondary school and college teaching, too.

Robert Slavin conducted some best-evidence syntheses of the research on grouping back in the 1980s (with an overall effect size of .32). But at that time, he found no reading studies that compared grouping for reading instruction with whole class teaching (Slavin 1987a; 1987b; 1988).

Not exactly the evidentiary base I was hoping for.

However, even if we were to rely on those meta-analyses, the payoff of grouping for reading instruction had lower effects than was found for the other subjects (only .13), and it mattered a great deal how large the groups were—groups of 5 or larger received little or no learning benefit from within-class grouping (Lou, et al., 1996), meaning that most teachers would require at least 4-6 groups.

There are a couple of sizable and especially pertinent individual studies I think we should consider, however.

For example, Kamil and Rausher (1990) conducted a study in which they compared whole class reading instruction with small group teaching in a large suburban school district. Surprisingly, they found that small groups “were not superior to whole class” teaching in terms of learning. There was just “too much variance within classrooms for the grouping patterns to have much of an impact.”

Even more sobering is a large study of grouping in reading conducted by Sørensen & Hallinan (1986). They found small group teaching more effective than whole class instruction—that is, if one compares 30 minutes of small group teaching versus 30 minutes of whole class teaching, the kids in the small group tend to make larger learning gains.

However, they also found small group teaching provides kids with fewer learning opportunities. Basically, teachers teach more content when working with whole classes.

Comparing equal amounts of small group teaching with whole class teaching might make sense to researchers, but it has little to do with the actual circumstances of classrooms. If a teacher has three groups who each receive 20 minutes of teaching, this should not be compared with 20 minutes of whole class instruction… but with 60 minutes—the time it takes to teach the three groups. It may be fairer to compare it with 65-75 minutes because it takes transition time to manage those groups.

When one compares small group and whole class instruction in this more appropriate way, small group teaching loses any advantage; that is, no differences in average achievement.

Of course, there are no such studies focused on foundational skills instruction. This is a case where it is necessary to generalize from the existing data, which shows no consistent benefit to small group teaching.

In the 2020s, there are studies of the effectiveness of small group decoding interventions, but none that compare them with whole class teaching. Interestingly, those recent studies reveal a decided unevenness of results. Research simply doesn’t find small group teaching to be especially powerful or consistent. It is not a panacea. As one study concluded, “for many students, simply receiving-evidence based reading instruction in a small-group setting is insufficient” (Kulesz, et al., 2024, p. 363). One explanation of this ineffectiveness is the great dosage differences that the approach creates (Piasta, et al. 2024). The more small groups, the less teaching anyone receives.

That unevenness is complicated by the complex nature of literacy. The simple view of reading says reading depends on two components: decoding and language. If we group instruction for one, we must reduce its dosage to accommodate those additional lessons. Or, if we group for one component and maintain its dosage, we must reduce the time accorded to the other, which is a bad deal, too. You need both components. Neglecting one to serve the other many provide a temporary boost in screening test results, but overall, it won’t help kids to become successful readers.

I’m not sure how folks who encourage small group foundational skills teaching want the kids grouped. There are many possibilities. It could be done randomly – reducing the number of kids taught at one time with no alteration of the heterogeneity. More likely the plan is to it separate the better decoders from the strugglers. There are no direct studies of this, but those studies of separating kids based on reading level are informative.

Sørensen and Hallinan (1986) concluded that high achieving groups tended to gain learning opportunities when separated, but that such opportunities were diminished or reduced for the lower achieving groups – an unevenness of group experience still evident in more recent studies (e.g., Kulesz, et al., 2024; Piasta, et al., 2023; Piasta, et al., 2024; Roberts, et al., 2023). That earlier research showed that Black kids and poor kids were more likely to be placed in those low groups and that they ended up with lower rather than greater progress. While there was not any overall or average learning benefit from small group teaching, there was a rotation, ensuring that the most vulnerable kids learned less.

My conclusion from all this is that small group teaching can be beneficial – teaching a lesson to fewer kids, under some circumstances, can improve the impact of the lesson. This advantage is far from certain, however. And, because amount of instruction – dosage – matters, too, we should be concerned about the reduction in teaching that small group approaches usually impose. At best, this results in a tradeoff – the kids seem to learn more in each lesson, but the reduction in curriculum coverage trashes those benefits. However, it is also likely that this tradeoff is an uneven one, an unfair one – conferring a small learning benefit to the most advantaged kids while delivering more substantial social and academic harm to the most disadvantaged.

I don’t think we should prohibit small group or even individual teaching because sometimes we can gain well-targeted learning benefits from them. But I would use these tools strategically. I would never organize my instruction with the purpose of providing daily small group time, per se. In fact, I would try to minimize my reliance on small group teaching whenever possible.

Instead of encouraging more and more small group teaching, districts should provide professional development in the most effective ways to deliver whole class instruction. This should include attention to the nature and use of seatwork. Seatwork is needed in both situations. With small groups, the emphasis is typically on keeping kids busy while the teacher works with other groups (e.g., “shut up sheets”). However, for whole class teaching the assignments can be more instructional since the teacher is able to circulate among the kids while they work giving support and additional guidance.

My watchword is “never do with a small group what you could do just as well with whole class teaching.”

References

Austin, M. C., & Morrison, C. The First R: The Harvard report on reading in elementary schools. New York, NY: Macmillan, 1963.

Duke, N., & Mesmer, H. A. E. (2018-2019). Phonics faux pas: Avoiding instructional missteps in teaching letter-sound relationships. American Educator. https://www.aft.org/ae/winter2018-2019/duke_mesmer 

Hattie, J. (2012). Visible learning for teachers: Maximizing impact on learning. London: Routledge.

Kamil, M. L., & Rauscher, W. C. Effects of grouping and difficulty of materials on reading achievement. National Reading Conference Yearbook, 39, (1990): 121-127.

Kulesz, P. A., Roberts, G. J., Francis, D. J., Cirino, P., Walczak, M., & Vaughn, S. (2024). Latent profiles as predictors of response to instruction for students with reading difficulties. Journal of Educational Psychology, 116(3), 363-376. https://doi.org/10.1037/edu0000832\

Lou, Y., Abrami, P. C., Spence, J. C., Poulsen, C., Chambers, B., & d’Apollonia, S. (1996). Within-class grouping: A meta-analysis. Review of Educational Research, 66(4), 423–458. https://doi.org/10.2307/1170650

Piasta, S. B., Hudson, A., Sayers, R., Logan, J. A. R., Lewis, K., Zettler-Greeley, C., & Bailet, L. L. (2024). Small-group emergent literacy intervention dosage in preschool: Patterns and predictors. Journal of Early Intervention, 46(1), 39. https://doi.org/10.1177/10538151231155411

Piasta, S. B., Logan, J. A. R., Zettler-Greeley, C., Bailet, L. L., Lewis, K., & Thomas, L. J. G. (2023). Small-group, emergent literacy intervention under two implementation models: Intent-to-treat and dosage effects for preschoolers at risk for reading difficulties. Journal of Learning Disabilities, 56(3), 225-240. https://doi.org/10.1177/00222194221079355

Roberts, G. J., Cote, B., Mehmedovic, S., Lerner, J., McCreadie, K., & Strain, P. (2023). Integrating behavior support into a reading intervention for fourth-grade students with reading difficulties and inattention. Journal of Behavioral Education, 32(2), 277–299. https://doi.org/10.1007/s10864-021-09457-y

Seger, W., & Stukey, M. R. (no date). Foundational skills instruction: Whole group? Small group? What’s best? Collaborative Classroom. https://www.collaborativeclassroom.org/blog/foundational-skills-instruction-whole-group-small-group-whats-best/

Shanahan, T. (2025). Leveled readers, leveled lives. Cambridge, MA: Harvard Education Press.

Slavin, R. E. (1987a). Ability grouping and student achievement in elementary schools: A best-evidence synthesis. Review of Educational Research, 57(3), 293-336. https://doi.org/10.2307/1170460

Slavin, R. E. (1987b). Grouping for instruction in the elementary school.” Educational Psychologist, 22(2), 109-127. https://doi.org/10.1207/s15326985ep2202_2

Slavin, R. E. (1988). Synthesis of research on grouping in elementary and secondary schools. Educational Leadership, 46(1), 67-77.

Sørensen, A. B., & Hallinan, M. T. (1986). Effects of ability grouping on growth in academic achievement. American Educational Research Journal, 23(4), 519-542.  https://doi.org/10.2307/1163088

 

LISTEN TO MORE: Shanahan On Literacy Podcast

 Click here to see the original posting the 64 comments it elicited.