Skip to main content
elementary teacher in lively conversation with students who have hands raised

Curriculum and Instruction

The Six Ts of Effective Elementary Literacy Instruction

What does good elementary literacy instruction look like? In this article, Allington identifies and discusses about these six features: time, texts, teach, talk, tasks, and test. 

On this page:

It seems that, finally, those who make educational policy — at the local, state, and federal levels —  have begun to recognize just how much good teachers matter.

A series of studies have confirmed what was probably obvious from the beginning. Good teachers, effective teachers, matter much more than particular curriculum materials, pedagogical approaches, or “proven programs” (Allington & Johnston, 2001; Darling-Hammond, 1999; Duffy, 1997; Pressley, et al, 2001; Sanders, 1998; Taylor, Pearson, Clark & Walpole, 2000). It has become clearer that investing in effective teaching — whether in hiring decisions or professional development planning is the most “research-based” strategy available. If we are to hope to attain the goal of “no child left behind,” we must focus on creating a substantially larger number of effective, expert teachers.

Good teachers, effective teachers, manage to produce better achievement regardless of which curriculum materials, pedagogical approach, or reading program is selected.

I am not going to attempt to understand why it has taken education so long to recognize what other industries recognized almost from the start — expertise matters. Instead, I am going to describe what the teaching of exemplary elementary teachers looks like and challenge school administrators to examine whether their daily practice and their longer-term planning is designed to foster such teaching. In other words, I believe school administrators should be crafting policies that ensure that more effective teachers are created each year in their schools.

For much of the past decade my colleagues and I at the National Research Center on English Learning and Achievement have been studying some of the best elementary school teachers in the nation (Allington & Johnston, 2002; Pressley, Allington, Wharton-McDonald, Collins-Block & Morrow, 2001). These teachers were selected, primarily, from schools that enrolled substantial numbers of poor children and schools that reflected the racial, ethnic, and linguistic diversity of the nation.

We observed first and fourth-grade teachers from six states (New York, Texas, New Hampshire, California, Wisconsin, New Jersey). In each case we spent at least ten full instructional days, and often more, observing, interviewing, and videotaping in each room. Two books, a number of articles, and related technical reports provide documentary details (the books and articles are cited throughout and the technical reports, along with research summaries, can be found at ).

We studied teachers found to be particularly effective in developing reading and writing proficiency. Over the course of the study, however, it became clear that the teachers we were studying developed academic proficiencies well beyond higher reading and writing achievement test scores (though the evidence we gathered did demonstrate that these teachers did produce significantly better standardized test performances as a matter of course).

The hundreds of days of classroom observation and the hundreds of interviews with teachers and students provide a clear portrayal of what good elementary teaching looks like. Below I sketch six common features — the 6 Ts of effective elementary literacy instruction — that we observed in the exemplary elementary classrooms we studied.


These teachers had a “reading and writing vs. stuff” ratio that was far better balanced than is typically found in elementary classrooms (Allington, 2001).

In other words, these teachers routinely had children actually reading and writing for as much a half of the school day — often around a 50/50 ratio of reading and writing to stuff (stuff is all the others things teachers have children do instead of reading and writing). In typical classrooms, it is not unusual to find that kids read and write for as little as ten percent of the day (30 minutes of reading and writing activity in a 300 minute, or five hour, school day).

In many classrooms, a 90 minute “reading block” produces only 10-15 minutes of actual reading, or less than 20 percent of the allocated reading time is spent reading. Worse, in many classrooms, 20 minutes of actual reading across the school day (Knapp, 1995) is a common event, which includes reading in science, social studies, math, and other subjects. Thus, less than ten percent of the day is actually spent reading and 90 percent or more of the time is spent doing stuff.

The issue is less stuff vs. reading than it is a question of what sorts of and how much of stuff. When stuff dominates instructional time, warning flags should go up.

This is true even when the activity, in some form, has been shown to be useful. Activating students’ background knowledge before reading (Pearson & Fielding, 1991) and generating discussion after reading (Fall, Webb & Chudowsky, 2000) is useful. But three to five minutes of building background knowledge is probably enough; spending most of a 90-minute reading block on building background knowledge seems an unlikely strategy for improving reading proficiencies.

In less-effective classrooms, there is a lot of stuff going on for which no reliable evidence exists to support their use (e.g., test-preparation workbooks, copying vocabulary definitions from a dictionary, completing after-reading comprehension worksheets).

Extensive reading is critical to the development of reading proficiency (Krashen 2001; Stanovich, 2000). Extensive practice provides the opportunity for students to consolidate the skills and strategies teachers often work so hard to develop. The exemplary elementary teachers we studied recognized this critical aspect of instructional planning. Their students did more guided reading, more independent reading, more social studies and science reading than students in less-effective classrooms.


If children are to read a lot throughout the school day, they will need a rich supply of books they can actually read. This seems a simple statement of fact. But there also exists a large and potent research base supporting supplying children with books of appropriate complexity (Allington, 2001).

Simply put, students need enormous quantities of successful reading to become independent, proficient readers.

By successful reading, I mean reading experiences where students perform with a high level of reading accuracy, fluency, and comprehension. When a nine-year-old misses as few as two or three words in each one hundred running words of a text, the text may be too hard for effective practice. That text may be appropriate for instructional purposes but developing readers need much more high-success reading than they need instructional difficulty reading. It is the high accuracy, fluent, and easily comprehended reading that provides the opportunities to integrate complex skills and strategies into an automatic, independent reading process.

The exemplary teachers we studied too often had to teach against the organizational grain. They rejected district plans that “required” all children be placed in the same textbook or tradebook (and do the same worksheets on the same day). They recognized such schemes for what they are: Truly anti-scientific, non-research-based fads designed more, it seems, as an attempt to exert administrative power than to produce high levels of student achievement.

Unfortunately, these exemplary teachers too often had to spend both their personal time and personal funds to locate and/or purchase the texts needed to effectively teach the children they were assigned. Some were lucky to work in “smart” organizations. These organizations provided a rich and expansive supply of texts that supported children’s learning across the school day (multi-level texts available for social studies and science as well as for reading classes). Organizations that knew that “one-size-fits-all” mandates contradicted virtually everything we have learned about effective teaching.

A primary outcome of these exemplary teachers was the acceleration of literacy development in their lowest-achieving students (Allington & Johnston, 2002; Pressley, et al, 2001). While students of all achievement levels benefited from exemplary teaching, it was the lowest achievers who benefited most.

In these classrooms, lower-achieving students spent their days with books they could successfully read. This has not typically been the case in less effective classrooms (Allington, 1983). In too many schools, the lower-achieving readers receive appropriate reading materials only when they participate in special support instruction (e.g., special education resource rooms, Title 1 in-class support, bilingual education block). In other words, in too many cases the lower-achieving students receive, perhaps, an hour of appropriate instruction each day and four hours of instruction based on grade-level texts they cannot read. No child who spends 80 percent of his instructional time in texts that are inappropriately difficult will make much progress academically.

These exemplary teachers noticed that the highest-achieving students:

  • received a steady diet of “easy” texts — texts they could read accurately, fluently, and with strong comprehension
  • consistently out-gained both the average-achieving students and the lower-achieving students, year after year.

They seemed to notice that motivation for reading was dramatically influenced by student reading success. They acted on these observations by creating multi-level, multi-sourced curriculum that met the needs of the diverse range of students in their classrooms.


Obviously, part of good teaching is planning instructional time allocations and selecting appropriate books. But here I want to focus more on the notion of active instruction  — the modeling and demonstration of the useful strategies that good readers employ.

Much of what many administrators might consider teaching behaviors involves little or no active instruction (NICHD Early Child Care Research Network, in press). Much of what many teachers consider teaching is little more than assignment and assessment. Somewhere along the way, active teaching – explicit explanation, direct teaching — has been lost in the shuffle of thinking about classroom instruction.

These exemplary teachers routinely offered direct, explicit demonstrations of the cognitive strategies used by good readers when they read. In other words, they modeled the thinking that skilled readers engage while they attempt to decode a word, self-monitor for understanding, summarize while reading, or edit when composing. The “watch me” or “let me demonstrate” stance they took seems quite different from the “assign and assess” stance that dominates in less-effective classrooms (e.g., Adams, 1990; Durkin, 1978-79).

The dominance of the “assign and assess” model has been too little written about, but the truth is that “instruction” of this nature is of little benefit to all but the few students who have already acquired a basic understanding of the strategy that is the focus of the lesson.

As Adams (1990) pointed out in her analysis of traditional phonics programs, when teachers assign a worksheet that requires children to fill in the missing vowel, only children who already know the correct vowel response can successfully do the task. And they don’t need the practice activity. Children who do not know which vowel to put in the blank space cannot acquire that knowledge from the worksheet. They need actual teaching. In other words, the missing vowel worksheet is an assessment of who already knows the vowel patterns not an instructional activity that will teach the vowel pattern.

Likewise, when assigned a story to read, with questions presented at the end to answer (Durkin, 1978), children who have already the developed appropriate strategy to use while reading can respond correctly, but those who have not developed the strategy cannot. And these latter children cannot acquire the strategy from the end-of-story questions. They would need someone to actually teach the strategy to them — someone who would model and demonstrate the strategy to use (Duffy, 1998).

These exemplary teachers seemed to realize that most commercial instructional packages provide no useful information on direct and explicit skills and strategy instruction. In other words, they realized that the scripts that one typically finds in commercial packages offer teachers a “definitional” model. Students are taught that the main idea in a text is the author’s most important idea about a topic. They offer little in the way of helping children develop useful strategies for determining the relative importance of the various ideas an author might present on a topic.

Thus, these teachers took on the responsibility of crafting explicit demonstrations of skill and strategy use.

For example, they might demonstrate the use of the deletion strategy when teaching summarization. They might show how to list the various ideas an author presents in a persuasive paragraph through a line-by-line analysis — a “watch me do this” lesson. Then they might demonstrate through a think-aloud process the strategy of deleting redundant, trivial, and subordinate information until they have arrived at the summary statement.

These teachers offer useful strategy models — decoding strategies, composing strategies, self-regulating strategies – as separate lessons to the whole class, to targeted small groups, and to individual students in side-by-side instruction. In fact, it is this literal plethora of instructional activity that truly sets these teachers apart and explains much of their effectiveness with lower-achieving students (Taylor, et al, 2000).

We have a wealth of studies demonstrating the power of active teaching, especially for children who struggle to learn to read and write. But for children to come to own the powerful strategies being presented they must have enormous successful practice in using the strategies independently — extensive successful reading experiences.

The instructional environment must also foster independent strategy transfer and use. A real concern is that when instruction becomes too explicit, too much of the time, children never acquire the independent strategy transfer and use. Use of a strategy in a highly structured, teacher-directed setting is not the same as knowing how and when to profitably and successfully use the strategy when reading independently. Thus, expert teaching requires knowing not only how to teach strategies explicitly, but also how to foster transfer from the structured practice activities to independent use while engaged in reading. It is this transfer problem that makes scripted instructional material problematic.


Like the Teach component, classroom talk is under-researched. We saw fundamental differences in the nature of the classroom talk in the exemplary teacher classrooms and the talk typically reported in classroom observational studies. First, we observed these teachers fostering much more student talk — teacher-student, student-student — than has been previously reported. In other words, these exemplary teachers encouraged, modeled, and supported lots of talk across the school day. This talk was purposeful talk though, not simply chatter. This talk was problem-posing, problem-solving talk related to curricular topics (Allington & Johnston, 2002; Johnston, Woodisde-Jiron & Day, 2001).

It wasn’t just more talk but a different sort of talk than is commonly heard in classrooms. We described this difference as “more conversational than interrogational.” Much previous work has well-documented the interrogational nature of most classroom talk. Teachers pose questions, children respond, teacher verifies or corrects. That is the dominant pattern observed in study after study, grade after grade (Cazden, 1988; Nystrand, 1997).

The classroom talk we observed was more often of a conversational nature than an interrogational nature. In other words, teachers and students discussed ideas, concepts, hypotheses, strategies, and responses with others. The questions teachers posed were more “open” questions, where multiple responses would be appropriate. For instance, consider the difference between the three after-reading questions below:

  1. So, where were the children going after all?
  2. So, what other story have we read that had an ending like this one?
  3. Has anyone had a problem with a pet like the boy in the story?

Responses to Q1 are strictly limited to a single “correct” response as dictated by the story content. But Q2 and Q3 offer the opportunity for multiple “correct” responses. In addition, while a response to Q1 leads only to a “Right” or “Wrong” teacher reply, Q2 and Q3 lead to follow-up teacher queries along the lines of, “Explain how the endings are similar” and “Tell us more about how your pet problem was like the problem in the story.” While Q1 offers an assessment of appropriate strategy use, Q2 and Q3 offer the opportunity to examine the thinking — the strategy in use — and the opportunity for instruction. Q1 assesses recall; Q2 and Q3 assess a broader understanding and help make children’s thinking visible.

The nature of classroom talk is complicated and too little understood. While there is evidence that more “thoughtful” classroom talk leads to improved reading comprehension (Fall, et al, 2000; Johnston et al, 2001; Nystrand, 1997), especially in high-poverty schools (Knapp, 1995), we still have few interventions available that focus on helping teachers develop the instructional expertise to create such classrooms and few of the packaged programs offer teachers any support along this line. True conversation cannot be scripted or packaged. The classroom talk we observed was highly personalized and focused on a targeted reply to student responses. Teacher expertise was the key, not a scripted, teacher-proof, instructional product.


Another characteristic of these exemplary teacher classrooms was the greater use of longer assignments and reduced emphasis on filling the day with multiple, shorter tasks. In these classrooms, students often worked on a writing task for ten days or more. They read whole books, completed individual and small group research projects, and worked on tasks that integrated several content areas (reading, writing, and social studies).

The work these children in these classrooms completed was more substantive, more challenging, and required more self-regulation than the work that has been more commonly observed in elementary classrooms. We observed far less of the low-level worksheet-type tasks and a greater reliance on more complex tasks across the school day and across subject matter. Perhaps because of the nature of this work, students seemed more often engaged and less often off-task than other researchers reported.

Relatedly, the tasks assigned often involved choice — student choice. We described the instructional environment as one of “managed choice.” Students did not have an unlimited range of task or topic choices, but it was less common to find every students doing the same task and more common to observe students working on similar but different tasks. For instance, in a fourth-grade unit on insects, each child caught and brought that insect to class. They then sketched the insect using magnifying glasses to discover detail. These sketches were then labeled for body parts (thorax, abdomen, antennae, etc.). Students also observed the insect in its natural environment and jotted field notes about observed behaviors and habits. They wrote a short description based on these notes and constructed a model of the insect from craft materials. Finally, they presented their insect to classmates and then posted their sketches, models, and descriptions on the classroom wall where classmates could review and study the insect projects.

Choice of this sort has been documented to lead to greater student ownership of the work and greater engagement with the work (Turner, 1995). A related characteristic is that such an array of student work makes it more difficult for students (and perhaps teachers) to rank student work from best to worst. Low-achieving students may have selected one of the more interesting insects to research and display. Peers see the new information on an interesting bug rather than seeing the same insect worksheet they just completed.


Finally, these exemplary teachers evaluated student work based more on effort and improvement than simply on achievement status. This focus meant that all students had a chance at earning good grades, regardless of their achievement levels. This creates an instructional environment quite different from one where grades are awarded based primarily on achievement status. In those cases, the high-achieving students do not typically have to work very hard to earn good grades. Lower-achieving students often have no real chance to earn a good grade regardless of their effort or improvement.

Achievement-based grading — where the best performances get the best grades — operates to foster classrooms where no one works very hard. The higher-achieving students don’t have to put forth much effort to rank well and the lower-achieving students soon realize that even working hard doesn’t produce performances that compare well to those of higher-achieving students. Hard work gets you a C, if you are a lucky low-achiever, in an achievement-based grading scheme.

The complexity, though, of effort and improvement grading lies in the fact that teachers must truly know each student well in order to assign grades. They have to be able to recognize growth and to track or estimate the student effort involved. The exemplary teachers often used a rubric-based evaluation scheme to assign grades. Improvement was noted based on where students started and where they ended up rather than on simply the latter.

One other aspect of the improvement and effort evaluation model is that it shifted much of the responsibility for earning grades over to the students. Students could not assign bad grades to “unluckiness” if only because the evaluation scheme was rather transparent to them. The rubrics provided the information needed to improve their grade.

The fourth-grade exemplary teachers we studied did acknowledge that the effort and improvement grading scheme required careful explanation to parents because most were more familiar with achievement-based grading. However, none of the teachers reported much parental resistance, perhaps because these teachers were typically able to describe in substantive detail just what a child needed to do to achieve a better grade.

I must also note that we observed almost no test-preparation activity in these classrooms. None of the teachers relied on the increasingly popular commercial test preparation materials (e.g., workbooks, software). Instead, these teachers believed that good instruction, rich instruction, would lead to enhanced test performances. The data bore out their beliefs. It was in the less-effective teachers’ classrooms that we observed as part of our sub-study that we found much test preparation activity. It seems that less-effective teachers truly don’t know what to do and, as a result, drift towards the use of packaged test-preparation activities in the hopes that such activities will make up for less-effective teaching throughout the year.


One danger in reducing complex activity to a bulleted list of key features is that such deconstructing risks oversimplifying the true complexity of the expert activity. Such seems the case here. While the six Ts offer a shorthand, of sorts, for describing exemplary elementary grades teaching, they also oversimplify the complex nature of good teaching. For instance, the six Ts actually operate interactively. That is, it doesn’t seem likely that we could choose a single T to isolate and attempt to develop teaching that reflects that T alone.

If we want, for instance, to substantially enhance the volume of reading that children do (and I would argue that is but one absolutely necessary modification needed if we hope to enhance reading proficiency), it would seem important that children had books they could read and choices as to which books they would read. Likewise, crafting a supportive conversational environment where students talk to their teacher and to their peers about the books they are reading would be an important component for sustaining increased reading. And, of course, adding active teaching of useful reading strategies would expand the array of books that children could read. Finally, shifting the evaluation to an effort and improvement scheme would foster enhanced motivation for reading.

In other words, creating and supporting exemplary teaching of the sort we observed is complicated. But it seems unfortunate that so many of the exemplary teachers we studied were forced to “teach against the organizational grain.” These teachers had to simply reject school and district plans that put 25 copies of the same reader, tradebook, textbook, or workbook in every child’s desk. They had to reject scripted lessons and pacing schedules and grading schemes that presented a “one-size-fits-all” model for instruction. They too often had to search out appropriate instructional texts and materials on their own because the school or district provided lots of copies of but one text and often that text was neither of appropriate difficulty for most students and rarely offered accurate and engaging information of the sort that might work to entice students into sustained and effortful study. Worse, in too many cases, these teachers were forced to spend their own funds to purchase the materials they needed to teach the students they were assigned.

Good teaching, exemplary teaching, should not be so hard to accomplish. Schools and school districts must assume more responsibility in crafting instructional and curricular support so that exemplary teaching becomes more common and requires far less effort. Good teaching should not have to work against the organizational grain.

In closing, I will note that few of these exemplary teachers much credited their school districts for the development of their expertness. Some pointed to administrators who allowed them to experiment, encouraged them to “break the mold” and told them not worry about test scores or following the organizational plan. But most credited other exemplary teachers for supporting and encouraging them to become better teachers and to assume greater professional responsibility for the success of their students. These teachers seemed to understand that personal professional responsibility rested on the fact that they chose how to teach, what to teach, and with what sorts of curricular materials and tasks. They rejected the low-autonomy/high accountability models that seem increasingly popular with advocates of “proven programs” (Day, 2001).

Instead, these teachers elected a high autonomy/high accountability model. They seemed to feel no particular pressure from state testing schemes, perhaps because their students performed so well. At the same time, because they were the architects of the instruction offered in their classrooms, they also reported a greater sense of personal professional responsibility for student outcomes. In other words, these teachers accepted the professional responsibility for developing high levels of reading proficiency, but insisted on the autonomy to act on their expertise (McGill-Franzen, 2000).

Educational leaders might do well to consider the nature of the instruction these teachers offered. They might do well to ask whether current school policies seem likely to foster this sort of teaching. They might ponder how the organizational plan, including the professional development opportunities and the curricular schemes, currently work to foster or undermine the emergence of exemplary elementary classroom teaching.

In the end, enhanced reading proficiency rests largely on the capacity of classroom teachers to provide expert, exemplary reading instruction. Our study of these exemplary teachers suggests that such teaching cannot be packaged. Exemplary teaching is responsive to children’s needs not regurgitation of a common script. In the end, it will become clearer that there are no “proven programs,” just schools where we find more expert teachers – teachers who need no script to tell them what to do. The question for the education profession — teachers, principals, professors, and policy makers — is: Are we creating schools where every year every teacher becomes more expert?

Note: The author would like to thank the following people for their assistance in the project: Peter Johnston and Michael Pressley, co-principal investigators for the fourth and first grade teacher studies, respectively; the classroom teachers who allowed us to study their teaching practices, and the other researchers on the project, including Elizabeth Asbury, Kim Baker, Kim Boothroyd, Greg Brooks, Melissa Cedeno, Cathy Collins Block, John Cronin, Jeni Pollack Day, Gay Ivey, Haley Woodside-Jiron, Susan Layden, Anne McGill-Franzen, Lesley Morrow, Steven Powers, Jean Veltema, and Ruth Wharton-McDonald. The paper is based on research supported in part under Research and Development Centers Program (award number R305A6005) as administered by the Office of Educational Research and Improvement, US Department of Education. However, the contents do not necessarily represent the positions or policies of the sponsor.

Publication Date
For any reprint requests, please contact the author or publisher listed.