Abstract
The Gradual Release of Responsibility (GRR) model for instructional delivery is well supported by research evidence and is often identified as a critical element of instruction for students with learning difficulties. However, there are challenges associated with effectively releasing responsibility to students. This may be especially true during instruction for primary-grade students with or at risk for learning disabilities. In general, educators want to provide a high level of support to young, struggling students in order to ensure that they feel successful and bolster their confidence. However, too much support means that students are not provided with opportunities for independent practice or independent mastery of content being taught. This article describes the GRR model of instruction and provides specific guidance and resources to support special education teachers in implementing the GRR model during small-group reading instruction in primary-grade classrooms, with a particular emphasis on facilitating group and independent practice.
Ms. Banks, a special education teacher, is passionate about providing explicit, systematic reading instruction to her students with reading difficulties. When teaching, she always clearly explains and models each skill. She gives students multiple opportunities to practice skills and carefully monitors their responses, providing corrective feedback when needed. By the end of a lesson, it usually appears as though all of her students have successfully mastered the skill being taught. Yet when she administers curriculum-based measures, it is apparent that some students have not yet achieved mastery. These students’ performance on curriculum-based assessments is often dramatically different from the performance she observes when she is providing instruction. Ms. Banks worries about these students, but she does not know what to do to help them. Why do they demonstrate mastery during lessons but not on individually administered assessments?
One week, while reteaching a phonics lesson on vowel digraphs to a small group of struggling readers, Ms. Banks pays special attention to one student, Sean, who scored very poorly on the curriculum-based assessment she had administered the day before. At first glance, it seems as though Sean is performing with a high degree of accuracy during group practice, as are his peers. However, when Ms. Banks takes a closer look, she realizes something. Sean is not quite responding at the same time as the other students. Maybe even subconsciously, he is waiting for other students to begin responding and then looking to them for clues as to the correct response. As the lesson progresses, Ms. Banks catches herself responding along with students during the group and independent practice parts of her lessons in a way that encourages them to look to her for clues or copy her responses instead of relying on their own knowledge.
It is as though a light has come on for Ms. Banks. In line with evidence-based recommendations, she is giving her students lots of opportunities to respond and receive feedback. But she is not giving them opportunities to respond independently, without her or other students. As a result, she is not giving them opportunities to receive corrective feedback that is specific to their individual needs. With this newfound knowledge, Ms. Banks embarks on a mission to change the way she teaches. In particular, she sets out to refine the group- and independent-practice portions of her lessons.
“… she is giving her students lots of opportunities to respond and receive feedback. But she is not giving them opportunities to respond independently, without her or other students.
The gradual-release-of-responsibility (GRR) model for delivering instruction is well supported by research evidence and is often identified as a crucial element of instruction for students with learning difficulties. However, there are challenges associated with effectively releasing responsibility to students. As researchers and former teachers, we have experienced and observed the challenges that even strong teachers like Ms. Banks face. This article describes the GRR model of instruction and provides specific guidance and resources to support special education teachers in fostering independence within a GRR framework during small-group reading instruction.
What Is the GRR Model?
The GRR model was developed over 35 years ago to describe the process by which teachers can systematically reduce supports provided during explicit instruction and shift the responsibility for learning to students (Pearson & Gallagher, 1983). The first stage in the GRR model is explanation and modeling. In this stage, the teacher clearly describes and demonstrates (in step-by-step fashion, if appropriate) how to perform a skill or use a strategy. Students may be prompted to respond during this stage (e.g., the teacher may prompt students to chorally repeat an important point), but the teacher maintains primary responsibility for demonstrating the content being taught. During Stage 2, guided practice, the responsibility for learning is gradually shifted to the student. This occurs when the teacher provides students with opportunities to respond in the presence of teacher support. During the guided-practice stage, the teacher may respond along with students or use prompting, additional modeling, or another type of scaffolding to support students’ initial efforts to demonstrate their learning. The third stage of Pearson and Gallagher’s original model was independent practice, which was defined as the stage when the student assumes all of the responsibility for learning. During independent practice, students are provided with opportunities to retrieve new knowledge, perform a new skill, or apply a new strategy without assistance from the teacher or peers. These three stages are often summed up using the catchphrase “I do, we do, you do,” coined by Anita Archer (1988).
As the body of research related to instructional practices grew, the GRR model evolved to include a fourth component, collaborative learning, between the guided-practice and independent-practice stages (Fisher & Frey, 2008). In the collaborative learning or group practice stage, students have opportunities to practice and apply their learning while interacting or responding together with their peers, in pairs or small groups. Because less teacher support occurs during group practice, more responsibility for demonstrating learning lies with students. However, students are still able to receive peer support before transitioning to full responsibility or independence. Thus, as shown in Figure 1, the four stages of instruction might be described as “I do” (teacher modeling and demonstration), “we do” (teacher-guided practice), “you all do” (student pair or small-group collaborative practice), “you do” (student independent practice).

Gradual-release-of-responsibility model
There is overwhelming research evidence to support to use of a GRR model during academic instruction in general (e.g., Alfieri et al., 2011; Marin & Halpern, 2011) and reading instruction in particular (e.g., Foorman et al., 2016; Shanahan et al., 2010). A number of systematic reviews of reading intervention research have concluded that students with or at risk for reading disabilities benefit from explicit instruction that involves a systematic GRR approach (e.g., Bryant et al., 2003; Chard et al., 2002; Gersten et al., 2009; Swanson, 1999; Vaughn et al., 2012). However, some research suggests that educators tend to implement the group (“you all do”) and independent (“you do”) practice stages of the GRR model less consistently than they implement the teacher demonstration (“I do”) or teacher-guided practice (“we do”) stage.
In a recent study of intervention implementation fidelity conducted by Hall et al. (2020) that corroborates findings reported by other researchers (e.g., Reutzel et al., 2014), teachers frequently omitted opportunities for students to respond independently during scripted small-group instruction designed to follow a GRR model. On occasion, teachers entirely skipped opportunities for students to respond without teacher support, as a group or as individuals (i.e., they seemed to believe that students had already demonstrated sufficient evidence of mastery during the guided-practice or “we do” stage). On other occasions, teachers who intended to provide opportunities for independent responding were observed to inadvertently respond along with students. For example, teachers often stated that it was “your turn” (i.e., the group’s turn or each student’s turn) to respond but still subtly responded along with the student or students. Students watched the teacher for a verbal or nonverbal cue (e.g., starting the student off by beginning the response, forming their mouth in the shape required to begin a correct response, or silently mouthing the response). Like Ms. Banks, teachers seemed to want to ensure that their students felt successful. However, in their big-hearted efforts to bolster student confidence, they deprived students of opportunities to independently master new learning.
In addition, in the same study of teachers’ intervention implementation fidelity (Hall et al., 2020), teachers sometimes provided corrective feedback following an incorrect response and facilitated student correct responding with teacher scaffolding (all evidence-based practices) but did not subsequently allow students the opportunity to respond correctly independently before moving on. Therefore, students were not able to demonstrate independent mastery of the skill being taught.
Implementing GRR During Whole-Class Reading Instruction
The implementation of GRR may look different in different settings. During whole-class reading instruction in the primary grades, the first two stages of GRR (“I do,” “we do”) might take place as students gather around the teacher on the rug. During the “you all do” stage, students might return to their tables and practice a skill or strategy in pairs and then (“you do”) independently. For example, a teacher targeting narrative comprehension might first read the beginning of a short story to students on the rug, following the read-aloud by demonstrating how to complete a graphic organizer about the characters, setting, and initiating event (“I do”). Next, the teacher could read the beginning of another short story and prompt students to turn and talk to a partner about which story elements belong in the graphic organizer boxes. The teacher would walk around the rug, listening in and providing support; then the teacher would call on pairs of students to share their answers and provide pairs with praise, corrective feedback, and scaffolding so that students can revise incorrect responses (“we do”). Third, the teacher would direct students to go to their desks, read the first section of a short story in pairs, and practice using the graphic organizer with a partner (“you all do”). Finally, students would have the opportunity to practice the skill independently (“you do”), with the teacher continuing to move around the room to provide support if needed.
Implementing GRR During Small-Group Instruction
The GRR model looks a bit different during small-group instruction in the primary grades, when students are usually seated at the “teacher table” throughout a lesson (as is common in Ms. Banks’s special education classroom). In theory, the stages of the GRR model are quite similar during small-group instruction and whole-class instruction:
Step 1 (I do): The teacher explains and models.
Step 2 (we do): The teacher engages students in practicing a skill or strategy while providing a high level of support.
Step 3 (you all do): Students respond in unison.
Step 4 (you do): Students respond without the support of the teacher or their peers.
“It is only when they have opportunities to respond independently—without the benefit of watching how the teacher or their peers move their lips or modulate their voices—that the responsibility for learning is fully transferred to students.
However, in practice, the last two stages do not always unfold in this way. With only four or five students in a small group, it is possible for Ms. Banks and other special education teachers to provide a high level of support to students as they respond not only during guided practice (“we do”), but also during group practice (“you all do”) and independent practice (“you do”)—despite the fact that these stages depend on the teacher providing increasingly reduced levels of support. The “independent practice” or “you do” stage of reading instruction is crucial. It is only when they have opportunities to respond independently—without the benefit of watching how the teacher or their peers move their lips and modulate their voices—that the responsibility for learning is fully transferred to students (e.g., Archer & Hughes, 2011; Pearson & Dole, 1987; Rupley et al., 2009). Independent practice enables confident mastery and prepares students to apply reading skills on their own when they read outside of the small-group setting.
GRR Within Different Components of Reading Instruction
As is represented in Table 1, the GRR model is applicable when teachers are targeting foundational reading skills (e.g., introducing a new letter-sound or practicing previously introduced letter-sounds) as well as when they are teaching more complex skills (e.g., teaching students to make an inference, identify a main idea, or read with prosody). Here are the steps that Ms. Banks would want to follow during a small-group phonics lesson with first-grade students:
Examples of GRR Stages Within Components of Small-Group Reading Instruction
Step 1 (I do): Introduce a new grapheme–phoneme correspondence. Show students a flashcard representing the spelling pattern, then turn it over to show an illustration of a key word that uses the spelling pattern (e.g., an illustration of a giraffe to help students remember the sound made by the letter combination gi). Model saying the sound associated with the spelling pattern.
Step 2 (we do): Show students the flashcard and prompt them to respond chorally with the correct sound for the spelling pattern on the card, responding along with students. This will give students one more opportunity to watch their teacher’s mouth and listen to their teacher’s expert pronunciation of the sound, even as they take on some responsibility for responding themselves. Provide scaffolding and corrective feedback when appropriate (followed by an opportunity for students to respond again, correctly, with teacher support).
Step 3 (you all do): Provide students with an opportunity to chorally respond with the sound for the spelling pattern. Intentionally do not respond with students. Listen and provide scaffolding and corrective feedback when appropriate (followed by an opportunity for students to respond again, correctly, without teacher support).
Step 4 (you do): Provide opportunities for independent responding, having students take individual turns pronouncing the sound. Make sure you do not give students “clues” to the correct response (e.g., by shaping your mouth in the way you would when beginning to pronounce the sound). Provide corrective feedback and opportunities to respond again, correctly, without teacher or peer support, when appropriate.
Here are the same steps within the GRR model that Ms. Banks could use to teach an example sentence-reading fluency lesson to her first-grade students:
Step 1 (“I do”): Model reading the first sentence in a list of decodable sentences, making your voice go up dramatically because the sentence ends in a question mark.
Step 2 (“we do”): Ask students to notice the punctuation mark at the end of the next sentence; prompt them to read it chorally along with you, making their voices do what the punctuation mark tells them to do. Provide corrective feedback and opportunities to respond again correctly.
Step 3 (“you all do”): Ask students to read the next sentence chorally without you. Intentionally do not respond with students. Listen carefully, stopping students to provide corrective feedback/opportunities to respond again correctly).
Step 4 (“you do”): Call on individual students to read a sentence independently. Intentionally do not respond with students. Provide corrective feedback and opportunities to respond again, correctly, when appropriate.
GRR During Small-Group Instruction: Frequently Asked Questions
How do I know when to release responsibility? How might the steps change depending on the levels of understanding that my students demonstrate?
Although the GRR model is described as having four stages, these four stages do not always occur linearly. Often, students receiving a small-group reading intervention will need to move back and forth between stages as they develop mastery (Fisher, 2008). It may be necessary for the teacher to repeat and perhaps ultimately to skip stages as students develop their understanding and mastery of given skills (Fisher, 2018).
For example, if Ms. Banks’s students were successful at reading decodable sentences during the group- and independent-practice stages on Monday, then on Tuesday Ms. Banks might decide to skip explanation and modeling and jump right to guided or group practice. However, if Ms. Banks realizes during group practice that her students are having some difficulties remembering the necessary grapheme–phoneme correspondences or blending sounds to read words, then she might go back to the “I do” stage and progress through the stages again until students are able to demonstrate independent mastery. Importantly, when it is necessary to backtrack, it may be necessary to go back and model only a specific part of the skill that students were not able to master independently.
As students develop mastery of a skill that is revisited during the course of the school year, the amount of time the teacher spends explaining and modeling (“I do”) before releasing students to engage in guided, group, and independent practice (“we do,” “you all do” and “you do”) will decrease (Archer & Hughes, 2011). Figure 2 represents the first lesson plan in a series of lesson plans targeting phonological awareness by teaching kindergarteners at risk for reading difficulties to “Put It Together!” or to blend parts of words (i.e., compound words, syllables, onsets/rimes, and ultimately phonemes) to make whole words. In the first lesson, the teacher devotes significant time to explaining and modeling (“I do”; cf. “Teach the Routine,” “Teach the Concepts,” and “Demonstrate” in Figure 2). Students spend less time engaged in guided and group practice (“we do” and “you all do”), and the only opportunities for independent practice (“Time to Shine”) are not truly independent; the teacher is only observing individual students in the context of group practice responding. However, when the same activity is taught the next week (see Figure 3), there is no scripted “I do” portion of instruction. The teacher launches right into guided practice (“we do”) and spends significant time engaging students in group practice (“you all do”) and independent practice (“you do,” or “Time to Shine”). These lesson plans are from a kindergarten version (still under development; Denton et al., 2020) of the Reading RULES! program developed by researchers at the Children’s Learning Institute at the University of Texas Health Science Center at Houston (Denton et al., 2018), which has been shown to positively impact students’ development of word reading, decoding, and reading fluency (Solari et al., 2018).

Initial “Put It Together” phonological awareness lesson (less independence)

Subsequent “Put It Together” phonological awareness lesson (more independence)
What if my students within a small group have varying levels of understanding?
It is common for primary-grade students with reading disabilities to be at different levels or stages of understanding (Shanahan et al., 2010). Teachers often worry about unintentionally leaving the most struggling students behind while the rest of the group learns and grows. One way to alleviate this concern is to provide students who demonstrate a lesser degree of understanding with additional opportunities to respond and receive corrective feedback during the independent-practice stage. For example, if Daniela is having a challenging time remembering short vowel sounds while the other students in her small group have demonstrated mastery of these grapheme–phoneme correspondences, it may be necessary for Ms. Banks to conduct independent practice in such a way that she is able to call on Daniela to respond more often than she calls on Daniela’s peers, thus allowing Daniela much-needed additional practice with vowel sounds.
Another way to address differing levels of student understanding is to pull an individual student or subset of the students who are at a similar level of understanding and provide extra practice using the GRR model during a different time during the day. Once again, this will allow individual students to have more opportunities to respond and receive specific feedback from the teacher. As few as 5 minutes per day of one-on-one reteaching and practice, provided in addition to the small-group lesson, may help accelerate a student’s progress.
How can I keep students engaged during the explanation and modeling stage of instruction?
Teachers should be encouraged to spend a lot more time on the guided-, group-, and independent-practice stages of instruction than on explanation and modeling (Vaughn et al., 2012). Teacher talk is generally not very engaging for students, so it is best to keep it to a minimum. But it is possible to keep students engaged during the explanation-and-modeling stage of the GRR model by encouraging them to respond even while the teacher maintains primary responsibility for demonstrating a new skill or strategy. Ms. Banks might ask her students to chorally repeat an important word or phrase (e.g., “What’s my word?”) or to mirror her actions as she demonstrates. If her students are learning a strategy that has multiple steps, Ms. Banks can prompt students to turn to a partner and name each step after it has been described or modeled.
How can I support students’ memory for skills they learned previously?
It is crucially important to incorporate ongoing distributed practice into GRR lessons, so that students can practice newly learned skills along with skills they have previously learned (Dunlosky et al., 2014). The grapheme–phoneme correspondence lesson described earlier (i.e., “gi” sounds like /ji/) could be followed by a brief cumulative practice session in which Ms. Banks would mix the flashcard for the newly taught sound in with a few cards for previously learned sounds and have students practice all of them. First, students could practice by responding in unison; next, Ms. Banks would ask each student to name a few cards from the stack independently to check their independent mastery of the skills. During the lesson focused on reading fluently by attending to punctuation, Ms. Banks might provide cumulative practice by having students read sentences that end in periods as well as those that end with question marks. Students could do this chorally and then independently.
Conclusion
The GRR model describes the process by which teachers gradually transfer the responsibility for learning (i.e., demonstrating new knowledge or performing new skills) to students during explicit instruction. Key components include explanation and modeling, guided practice, group practice, and independent practice (“I do, we do, you all do, you do”). Using this model during whole-class reading instruction facilitates student learning and improves student outcomes. Research also strongly supports using the GRR model to teach students with reading difficulties during small-group reading instruction. However, it is easy to subconsciously omit opportunities for independent practice during small-group reading instruction. Teachers are likely to see tremendous benefits for student learning when they incorporate all stages of the GRR model into their small-group reading instruction in order to enhance student success.
Ms. Banks knows that a change in her approach to group and independent practice is vital for the success of her students. She reaches out to her district for support, and it is able to provide trainings on an evidence-based reading program that emphasizes the use of GRR during small-group reading instruction.
Now, Ms. Banks implements the GRR model during her small-group reading instruction with ease. She is able to seamlessly transition from “I do” to “we do” to “you all do” to “you do” throughout her lessons, and she knows when to skip or repeat a stage depending on the level of understanding that her students demonstrate. Although she still occasionally has to remind herself not to respond along with her students when it is their turn to practice skills independently, she mostly keeps her lips tightly shut while still nodding and encouraging silently. Her students have benefited tremendously from her change in instructional approach: She has seen improvements in their progress-monitoring scores and in their overall abilities as readers.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported in part by the Institute of Education Sciences, U.S. Department of Education, through Grant R305A180094 to the University of Texas Health Science Center at Houston. The opinions expressed are those of the authors and do not represent views of the Institute or the U.S. Department of Education.
