Enabling automatic retrieval of schemas from long-term memory in English grammar practice

People are increasingly aware that metacognition can help us solve problems more effectively. Scholars believe that students’ metacognition can be facilitated by promoting the interplay between the unlimited long work memory for holding schemas and limited work memory for processing ongoing activities. In this study, a workshop designed teacher mediation method between English language teachers and students is implemented that focuses on instructional talk supported by a visualized thinking framework in learning English grammar. Consequent to student’s automation from retrieving schemas to engage in the self-regulatory process of English grammar problem-solving, this section involves three English language teachers and students in their taught classes in a secondary school in China. The teachers were equipped with an explicit approach to teacher mediation through researcher-designed workshops. As the teacher mediation approach was implemented in English language classrooms, the researchers made meaning of their experiences through the dialogue and interaction between teachers and students by revealing the essence in the mediation process: dialogic assessment, cognitive scheme construction, automated schema retrieval, and metacognitive schema construction. Through teachers’ English pedagogical grammar practice in the three example cases, metacognitive teachers and learners are expected to be cultivated in an interdisciplinary view.

English language listening, writing, and grammar problemsolving on the summative assessment, and memory plays a pivotal role in students' successful performance on the English language test.
Under cognitive load theory, human beings process mental information in two functional areas: working memory and long-term memory (Flavell et al., 1985;Sweller et al., 1998), where limited working memory deals with all conscious activities, and unlimited long-term memory is used to store schemas of varying degrees of automaticity (Paas et al., 2004). A schema can be anything that has been learned and is treated as a single entity, and one of schemas' obvious functions is to provide a mechanism for knowledge organization and storage; schemas thus reduce working memory load. If the learning has occurred over a long period, schemas can incorporate a huge amount of information. Although working memory can process only a limited number of elements at a time (Arbib, 1992;McVee et al., 2005), the size, complexity, and sophistication of elements are not limited (Sweller et al., 1998).

3
Experts are successful not because their reasoning is more sophisticated than that of novices but because they have access to knowledge that is unavailable to others. Conversely, the novice must engage in seemingly less sophisticated reasoning not because it is particularly effective but because they do not have access to knowledge that the experts have. Although schemas are stored in long-term memory, information must be processed in working memory to construct schemas (Chi et al., 1981), and quick recall is needed between the stored schemas in long-term memory and the information relevant to solve a problem (Baddeley, 1992). In other words, automatic retrieval of schemas is a powerful component of knowledge mediation. For example, according to Simon and Gilmartin (1973), an expert chess player can master as many as 100,000 moves through regular practice and make sagacious decisions in a limited time during matches. That is, expert players are no better than others in randomly producing fresh moves: Their expertise lies in their ability to quickly recall the correct moves from their past practice and consequent saved long work memories.
However, unlike an expert chess player who has years of practice to master thousands of chess game moves (Wu, 2015), students have at most three years in high school to complete the designed English language education curriculum along with other subjects. They have limited time to study and save their working memory content knowledge into long-term memory for later retrieval in problem-solving (Globerson, 1983). Therefore, how English language learners construct schemas for effective automatic knowledge retrieval is critical for teachers to help students develop higher level thinking skills in solving grammar problems (McVee et al., 2005). Borjars and Burridge (2019) contended that grammar knowledge is a particular advantage for students who intend to further their English learning; teaching is influenced by a teacher's cognition, beliefs, assumptions, and attitudes about the teaching of grammar. Borg (2011) proposed that professional interventions need to empower language teachers to think more explicitly about, become aware of, and articulate their theories and consolidate others toward developing effective classroom practices, and interactive approaches and methods are crucial to any such interventions (Matkasimova & Makhmudov, 2020). Beeth (1998) described several ways teachers can create environments that foster discussion and negotiation of scientific concepts including encouraging students to talk about and support their ideas.
In this study, we developed a visualized theoretical framework to help English language teachers understand the essence of teacher mediation (Anderson, 2002), which is critical in preparing students to apply metacognitive thinking in their practice of grammar problem-solving. This framework will enhance English pedagogical grammar instruction as a tool to guide students' thinking and develop their metacognitive learning (Walker, 2019). We aimed with this study specifically to answer the following research question: How can English language teachers facilitate students' metacognitive thinking in English grammar teaching and learning?

Theoretical framework
Cognitive load theory identifies the interaction between unlimited long-term memory and limited working memory (Arbib, 1992;Baddeley, 1992;Paas et al., 2004;Sweller et al., 1998), where unlimited long-term memory deals with schemas of varying levels of automation. In contrast, limited working memory is associated with processing new information (Baddeley, 1992). A schema is a major learning mechanism that is stored in long-term memory and that relates to some fundamental categories of information and their significances for learning such as generalization and transfer (Baddeley, 1992). Schemas allow us to categorize information based on how we will use it. In this study, we take working memory as consciousness on the assumption that human working memory has limited time and mental capacities (Arbib, 1992). Humans are only conscious of working memory and can monitor its contents, but all other cognitive functioning is hidden in long-term memory and remains unconscious until it can be brought into working memory (Chase & Simon, 1973).
In teacher mediation, teachers guide students to bring knowledge from long-term memory to working memory automatically, thereby freeing their memory capacity to monitor and enhance their response time for ongoing activities (Hui, 2012). Teacher mediation has several avenues of giving students well-organized content knowledge that results in rapid automatic recall (Anderson, 2002). For example, in a well-arranged supermarket, customers easily find what they want without spending much time searching throughout the shelves. In contrast, when knowledge is fragmented, it is less available automatically, and learners must use mental capacity to recall it (Kellogg, 2008), which can interfere with or impede higher level thinking, reasoning, and problem-solving (Hartman, 1998). Figure 1 presents the visualized framework that we proposed for developing teachers' awareness of the interplay between long-term and working memory.
We expected that equipping teachers with this visualized framework in workshops we designed would help teachers understand the avenues of meditation for best helping students learn effectively and thereby facilitating the students' metacognition. Students can construct schemas in short periods of time with strong teacher support (Kellogg, 2008), and Fig. 1 presents a pattern for teacher mediation based on the theoretical framework and our own teaching experiences.
As the figure shows, the process proceeds as follows: (1) dialogic talk between the teacher and students confirms the problem in a joint activity; (2) teachers model conceptual knowledge and strategies to help construct cognitive schema; and (3) teachers use deliberate practice to automate students' schema retrievals. Fourth and finally, students think aloud (Van Someren et al., 1994), and teachers' input on students' reflections create awareness and self-regulation in the students' metacognitive schema construction (Garner & Alexander, 1989). Teachers in this study were empowered by having ownership of the visualized framework as a thinking tool to design instructional activities and engage in instructional practice.

Rationale for the theoretical framework
Previous research has found that automation is important in facilitating metacognition for effective problem-solving, and it is a skill that students can develop with practice and explicit instruction. Gordon and Braun (1983) explored the importance of assembling a set of organized notes in reading comprehension and found that well-organized data charts were powerful in reducing the mental burden of memorizing details. The authors also found that explicit story structure was beneficial for students to understand and recall stories effectively (Gordon & Braun, 1983). These studies support Van Merrienboer and Paas's (1990) conclusion that detailed procedural knowledge is implicit and not easy to verbalize in instructional talk. Those authors investigated visualized worked examples as an alternative way to communicate detailed procedural knowledge and identified that worked examples as concrete schema helped students map their new solutions in problem-solving and shortened the learning process for automation (Van Merrienboer & Paas, 1990).
Various strategies can guide students toward building well-arranged content knowledge (e.g., chart, story structure) for automation. For instance, Shintani et al. (2016) investigated the learning effects of a visual grammar teaching method based on schema theory, and participating students had a positive impression of the instructional method. Although teaching metacognition can enable the learning of specific strategies for facilitating automation, teaching an explicit and overall approach presents a challenge for educators. In a phenomenological study, Guo (2020) investigated three secondary school teachers of mathematics, English, and science and their students in China. The findings highlighted a simple but clear and detailed process (Fig. 2) for teachers in developing students' metacognitive skills in instructional talks for problem-solving: (a) using dialogic talk to assess what they know and where they are in the thinking process; (b) systemically providing the missing knowledge; and (c) monitoring the problem-solving process with students to enable them to develop tools for independent learning (Guo, 2020). The visualized pattern facilitated instructional talk and made the teaching and learning approach explicit and constructive (Guo, 2020).
These findings underscore how automation strategies can enhance metacognition facilitation (e.g., visualizing tools, talk pattern); however, scant attention has been given to supporting and empowering the teacher to know how the schema is constructed and automated in the classroom practice. Therefore, this study aimed to empower teachers to develop students' metacognition with automated schema construction skills using the theoretical framework shown in Fig. 1, which visualized the interaction among dialogic assessment, cognitive schema construction, automated schema retrieval, and metacognitive scheme construction. This framework was created by the researcher and played a significant role in framing and fostering teacher professional learning and classroom implementation sequent to the evaluation of their professional understandings in classroom practice.

Key terms
Here, we define the key terms that we use in this study:

Visualized thinking framework
In this study, visualized thinking framework is a thinking tool (Sinclair, 2007) that we developed to explain phenomena of metacognition facilitation, draw connections between automated schema retrieval approaches, and predict and evaluate its implementation in the classroom. In Fig. 1, we

Automated schema retrieval
A schema is a cognitive framework or concept that helps organize and interpret information, and according to Flavell et al. (1985), schemas are stored in long-term memory and are unavailable until people bring them into usable working memory. In this study, teacher modeling helped automate students' schema retrieval, and their deliberate practice is one of the steps in Fig. 1 to facilitate metacognition.

Teacher mediation
Vygotsky's central concept of cognitive psychology is mediation (Guo, 2020;Karpov & Haywood, 1998). According to Vygotsky, empowering teachers to mediate and teach metacognitive skills is of enormous significance. We defined teacher mediation in this study as mediation of learning to equip students with the necessary knowledge, skills, and attitudes to be able to mobilize learning in an effective way.

Thinking frame
It is a representation intended to guide the process of thought by supporting, organizing, and catalyzing that process (Perkins, 1986). The thinking frame does not define in advance what answers we will get; rather, it stimulates us to invent answers. We present a thinking frame for metacognition facilitation for this study in Fig. 1.

Definition table
Definition table is a data structure that organizes information into rows and columns that can be used to both store and display data in a structured format. For this study, students used the definition table to differentiate four words for visual functions. These tables help students to organize and differentiate information so they can apply it to particular situations.

Smooth song
Music is one tool to engage students and provide a pathway for connections and deeper understanding. Werner (2018) asserted that songs created for classroom use and designed to target specific linguistic items could harness the memory benefits of music for language learning. These songs are essentially poems that pack much meaning into a few words. In this study, the teacher and the researcher created a smooth song as one of the schema construction strategies in the workshop by adapting the lyrics to talk about the functions of the different time prepositions.

Illustrations
Illustrations such as graphs, charts, maps, drawings, and photographs are often the quickest way to communicate large amounts of complex information that would be complicated to explain in text (Duchastel & Waller, 1979). Illustrations allow students to take shortcuts in interpreting the vast amount of information that is available in our environment. In this study, we designed the fish illustration as one of the schema construction tools to show the relationship of the subject clause and the object clause.

Instructional talks
Discussion-based lessons that develop students' conceptual knowledge and understanding through guided dialogues all students must participate in (Goldsmith, 2013). In the instructional talk, also called dialogic teaching (Lyle, 2008), the teacher listens carefully, makes guesses about the intended meaning, and adjusts responses to assist the student's efforts, in other words engaging in dialogue to make sense of the ongoing activities. In this study, the classroom instructional talk was supported by the visualized theoretical framework to cultivate the metacognitive learners. Wood (1988) differentiates learning into two levels: The initial level is learning about cognitive knowledge such as basic concepts and procedures, and the advanced level is learning how to construct knowledge and thinking. The advanced level of learning comprises competencies that metacognitive learners should hold (Guo, 2020). More precisely, in this study, metacognitive learners know how to plan, monitor, and assess their understanding.

Methods
The primary concern of this qualitative classroom-based research was mediating teachers' use of a visualized theoretical framework to automate their students' retrieval of schemas from long-term memory and evaluating the teachers' use of the framework in classroom practice. Through researcher-designed workshops, we equipped the teachers with an explicit scientific approach to follow in their mediation, and we then observed and recorded how the teachers' professional understanding manifested in their classroom implementations (Creswell & Creswell, 2018). We determined how well the teachers understood the framework and the phenomenon by contrasting the implementation in the classroom with the thinking framework we visualized in Fig. 1.
For this study, we examined three English language teachers and their students in their taught classes in a secondary school in China. We used purposive sampling to select the sample based on their knowledge and experience in teaching. Each teacher had at least 2 years of teaching experience and possessed the required academic qualifications to teach English. The average age of the students was 12.

Data collection and analysis
Data collection tools used in this research study were interviews, observations, document analysis, and a reflective journal (Creswell et al., 2007). We collected the data in steps as follows: First, we observed the teachers' current English grammar teaching practice, and then we trained the teachers to use the visualized theoretical framework to enhance their automated schema construction skills in their English grammar instruction practice. We recorded teachers doing exercises aloud and analyzed the data using V-note; Appendix 1 contains an example of an instructional co-designing scene with one of the teachers.
For the study, we supported the three participant teachers in designing grammar instructions on three different topics: four words with visual functions, time prepositions, and object clauses. We used three schema construction strategies in the instructional design to help teachers facilitate classroom talks: a definition table, smooth song, and the fish illustration. Table 1 shows the graphical information of the three teachers and their in-class activities.
In the third step of data collection, we observed the teachers' classroom instruction practices based on their understanding of the workshops, recording the sessions video to capture the dialogue between teachers and students; see Appendix 2 for a picture taken from one of the classroom observations. We transcribed the observation data verbatim in Chinese and then translated these transcriptions into English. We then used thematic analysis to code the collected data according to themes and subthemes (Boyatzis, 1998).
We evaluated how well the teachers were implementing mediation by comparing our observed interactions between teachers and students and with the framework in Fig. 1. We were evaluating the teachers' ability to enhance students' metacognitive schema construction skills in the three grammar problem-solving topics. More importantly, we were also assessing the consistency between the mediation approach we developed with the teachers in the workshop and their classroom implementations.

Ethical consideration
This research study was part of a larger project, and we obtained ethical approval from our academic institution before participant recruitment and data collection. After we obtained ethical approval, we sent participants consent forms that gave information about the study and later collected the signed forms. We also addressed participants using pseudonyms to ensure anonymity and confidentiality (Lodico et al., 2010). For data accuracy, we sent the participating teachers electronic copies of their transcripts for member checking.

Results
Here, we present our findings on the patterns of student-teacher interactions and the correspondence between the visualized framework we co-designed with the teachers to help them facilitate students' metacognitive learning and how the teachers operationalized the framework in the classroom. We designed three cases for analysis: teaching the children to differentiate four English words that articulate visual functions; teaching time prepositions, and teaching object clauses.
We used Fig. 1 to analyze the dialogues and determine the essence of mediation in teaching, and we tabulated the results in Table 2 to correspond to the letters in Fig. 1. For instance, in Summer's class, c in the definition table corresponds with substep c, conceptual knowledge, and substep c correlates with cognitive schema construction B and Automated schema retrieval C in the figure. Summer had the students discuss among themselves to choose the right answer, and then she initiated the classroom conversation with her students to learn what they did and did not know; from there, she presented them with the definition table tool. Below is the transcript of the recorded instructional talk between Summer and the students: In the classroom practice above, knowing the challenges for second language learners in matching the four visual words with the appropriate situations in the text, Summer helped students to construct the schema of the four "visual" words using a categorized table. Then, she followed the predesigned instructional activities from the workshop and led students to retrieve the schema (the table) in deliberate practice. Finally, Summer gave the students an opportunity to think aloud and reflected back to them to further facilitate their self-regulation skills.

Time prepositions: Wen
English teacher Wen was consolidating students' grammar knowledge of the time prepositions learned from a previous class. In the previous class, Wen had informed the students that there are different kinds of time prepositions, but in the present year 7 curriculum, they would only need to learn in, on, and at. The students' task was to know the functions of all three and how to use them appropriately in a sentence, and Wen taught the lesson using a fun, smooth rhyme (co-designed with the researcher). In that lesson, Wen and the students sang together while beating time with their hands, and that lesson continued to the present class to consolidate the students' grammar knowledge of time prepositions in problem-solving. Below is the song Table 2 Relationships of Essence of Mediation, Steps, and Substeps in Fig. 1 Steps Tabulated data across three teachers to show relationships of, Steps, Substeps and Essence of mediation (Summer's class, Wen's class, Chun's class) in Fig. 1 Substeps Essences of mediation

Eliciting students map the Fish illustration into new structures of the two sentences (e.)
To know what they need besides the structure of the object clause (a.)

Identify the new knowledge gap (b.)
Providing well-organized principles for problem-solving

(c. and d.)
Retrieve and use the two principles to free capacity for reasoning by using the constructed schema of the two principles (e.)

Practice automation skills in doing (f.) Mapping the two principles into new solutions (e. and f.)
Think of the thinking and teacher reflection (g. and h.) we co-designed to help the students organize their knowledge in preparation for the monitoring process: In "year" in "month" in "season," In "morning" in "afternoon" in "evening." On "day" on "holiday" on "weekday" And on "with specific day's morning, afternoon, and evening." "At" is relatively simple. Just put it before the time of the hour. For classroom practice, Wen gave the following question: The sports meet will begin 9:00 Monday morning.

A. on, in B. at, on C. in, on D. in, at
Wen elicited the instructional talk and identified the grammar topic (time prepositions). Then she held a class discussion with the students for automation skills practice: 10. Student: I can recall the saying in the song about the descriptions "on" and "at." On "day" on "holiday" on "weekday" and "on with specific day 's morning, afternoon, and evening" and "At" is

Teacher: Great. When you first read the questions, you might know that it is a question about time preposition, and then you confirm what problem is to be solved. Then, you recall what knowledge you already have in your memory to match the situations created by the instructional sentences. Then you monitor each blank in the text and the best-fit time preposition utilizing the smooth song. (Teacher's reflection)
Unlike in Summer's class, Wen was consolidating students' grammatical knowledge of the time prepositions learned in a previous lesson to increase the automaticity of retrieving their schemas for time prepositions using the smooth song of time prepositions. Instead, of a single choice problem, in deliberate practice, Wen used cloze test questions, one of the problem-solving question types in the Chinese secondary school curriculum along with single correct answer questions. Cloze test questions generally comprise a cluster of related single choice questions in a story; therefore, they involve higher-order mental processes that require time for processing ongoing new information in the text. Hence, cloze test questions demand students' free capacity for reasoning using automated schema retrieval. Wen led students to recall the smooth song six separate times to model how to shift from schema retrieval to problem-solving.

Object clauses: Chun
Chun was teaching her students the definition and functions of object clauses as well as how to use them in sentences. An object clause is a kind of nominal clause that acts as an object in a subject-subordinate compound sentence; the object clause is the clause after the transitive verb, preposition, or compound predicate (Elsness, 1984). Chun gave the class the following object clause problem: Do you know the meeting? -Tomorrow morning.
A. when they had B. when they are going to have C. when did they have D. when are they going to have Chun then gave the students, divided into groups, time for discussion and walked around listening; sometimes she stopped to provide guidance if students appeared to be struggling. During this classroom observation, Chun first used dialogic assessment to identify the students' knowledge gaps. Chun and the students then had the following recorded dialogue:

Teacher: See the picture in the PowerPoint slides?
Can you see what the picture is illustrating? (Present fish illustration for schema construction)

Teacher: What are the connections between the fish and the object clause? (Using the fish illustration to help make sense of the relationship between the main clause and object clause)
3. Students: The big fish contains small fish; the small fish belongs to the big fish. 4. Teacher: But still, the small fish has its structure by having complete body parts: "head, body, and tail." How does it reflect the object clause, especially the small sentence in the big sentence? (Elicit students' mapping of the fish illustration into new structures of the two sentences) 5. Students: It means that the main clause contains an object clause, but the object clause still has its own subject, predicate verb, and object. (New relationship schema of the two sentences is constructed) 12. Teacher (after the students took some notes): So what can we do now to choose the correct answer? Then see the consistency of tense, "tomorrow morning," so that the object clause is supposed to be in a future tense (Free capacity for reasoning using the constructed schema of the two principles) Teacher: Great. Then which answers can we exclude? Students: A and C. They are past tense. Teacher: So next, what can we do about the other two choices?
When students successfully excluded the wrong choices of A and C, Chun initiated another round of instructional talk by mapping the formed schemas (fish illustration and the two principles) into the reasoning process for the students to select the correct answer from the remaining two choices. The interactive talk aimed to help students practice automation skills in problem-solving, and after the consolidation, the students' and teachers' reflective discussions facilitated metacognitive thinking as follows: Teacher: Who can share the whole thinking flow again with us? Students: Yes. When I see this kind of questions, I will first confirm the problem is about the object clause. Then I will recall what I have about the object clause, such as the big fish small fish structure and the two principles. Then I will monitor it in reasoning to solve the problem. (Think of the thinking) Teacher: Great. Solving a problem very effectively and productively and accurately need lots of input. Essentially, we need have a very well organized structure of the knowledge, and then use it effortlessly whenever the problem comes. Then, you will have the capacity to think and reason in problem solving. (Teacher reflection) As shown in the conversation above, Chun did two rounds of schema constructions using a cyclic design. First, she assessed the students' knowledge of object clauses using the small and big fish illustrations to construct the schema of the relationship between the object clause and the subjective clause. However, when they returned to the problem-solving process, the students raised the need to construct a schema for the two principles in object clause problem-solving. The two rounds of schema constructions demanded that Chun engage the students in the automated skill retrieval process cycle two times to retrieve the thinking framework. Having the students do this work freed the teacher's mental capacity to respond to students' ongoing learning needs in classroom practice.

Discussion
In this study, we provided English language teachers with a visualized theoretical framework to help them understand the essence of teacher mediation (Anderson, 2002), which is critical for preparing students to apply metacognitive thinking in their practice of grammar problem-solving (Pintrich, 2002). We grounded the visualized framework in cognitive load theory (Arbib, 1992;Baddeley, 1992;Paas et al., 2004;Sweller et al., 1998), highlighting the teacher's mediation of automated schema retrieval between the student's limited working memory and unlimited long-term memory. We presumed that with strong teacher support, students could acquire both schema construction and automated retrieval skills in a short time. The framework we visualized outlined four steps as a thinking frame for teachers' pedagogical grammar practices to facilitate students' metacognition: dialogic assessment, cognitive scheme construction, automated schema retrieval, and metacognitive schema construction.
The findings highlight that automating schema retrieval is a powerful technique in a teacher's mediation in deliberate classroom practice. As designed in the framework we visualized, teachers needed two prior steps to prepare students for schema retrieval: first, to conduct dialogic assessment of what students need, and second, to construct cognitive schemas with students by supplying what they needed systematically. Empowered by these two preparation steps, teachers and students were expected to co-experience the thinking journey toward increasing the automaticity of students' schema retrieval. The final step in constructing metacognitive schema was shifting students' thinking to a higher level with their thinking aloud and the teachers' reflections.
In our findings, the three English language teachers effectively brought students' knowledge from long-term memory to working memory automatically, thereby increasing their work memory capacity to monitor and enhance their response time for the ongoing activities (Ericsson & Kintsch, 1995). The visualized framework we equipped the teachers served as a thinking tool to develop the students' metacognition by having them practice automated schema retrieval; the thinking tool supported the teachers in designing instructional practices and activities in a cyclical rather than linear process: Teachers can always start the grammar instructions with a dialogic assessment and then naturally come into the second step of schema construction, and although the assumption is that they would proceed to the third step, they found in this study that they could stop and step back when a new conceptual or procedural gap emerged. The thinking framework empowered the teachers to insert new episodes as part of the bigger sequence of the thinking journey. The inserted episode constructs the needed schemas for problem-solving using the four substeps (dialogic assessment, cognitive scheme construction, automated schema retrieval, and metacognitive schema construction).
For example, Chun initiated a dialogic assessment to learn what students did and did not know about object clauses. After identifying the students' learning needs, she constructed cognitive schemas that illustrated the big and small fish conceptual relationships; the aim was to prepare students to engage in problem-solving that would free capacity for self-regulated monitoring. Once the schemas were constructed, Chun and the students engaged in the deliberate practice to raise the automaticity of the schema retrieval.
However, as the thinking in classroom instructional talk evolved, Chun perceived the need to construct a schema for the two principles of object clause problem-solving. Therefore, she paused the current activities and started a new cycle of the thinking framework to confirm the recent problems and to systematically give students knowledge using the four substeps. The findings suggest that the thinking frame itself can be a schema that frees teachers' mental capacity to interplay students' knowing and thinking. The formed schema would be expected to increase the automaticity in teaching practice to attend to students' changing needs (Kotovsky, et al., 1985). In turn, students are likely to be more engaged when they know they can introduce changes if they need to without causing chaos for the teacher.
Grammar skill is a particular advantage for students who wish to pursue advanced English learning (Matkasimova & Makhmudov, 2020). In this study, grammar as the structure and system of a language is treated as a schema to guide and free students' thinking in problem-solving. Working memory can process only a limited number of elements at a time, but the size, complexity, and sophistication of features are not limited (Sweller et al., 1998). The framework we visualized for this study theoretically and practically helps teachers in grammar instructional design and pedagogical practice to mediate students' thinking and learning toward independence.
In our findings, the teachers' instructions guided students to intentionally construct well-organized grammar structures with unlimited size and information. The teachers used three different schema construction strategies to help students organize the grammar structures: definition table, smooth song, and fish illustrations. According to Perkins (1986), teachers can help students become better thinkers by purposely using designed frames to organize their thinking, and the three schema construction strategies the teachers used here can scaffold students to think with more depth and structure (Reiser, 2004). Examples abound of effective thinking strategies for educational contexts such as brainstorming, diagram, figure, flowchart, music, and tables. The findings in this study suggest that it is important to select the right strategy and tool for the particular activity and to scaffold the students' use of the strategy and tool carefully.
Using the effective strategies and tools allows students to effortlessly categorize substructures into central systems in the students' ongoing independent practice. We also encourage teachers to constantly consolidate the grammar structures in deliberate practice toward automating students' schema retrieval. Regular repeated practice would increase students' capacity to engage in phrasing for communication and problem-solving and help them be aware that cognitive development is modifiable (Vygotsky, 1978). That is, knowledge in grammar and language learning is accumulative: "I am not an expert now, but I can be more of an expert with effort later."

Conclusion and implications
This study provides a visualized theoretical framework as a thinking tool for English teachers to follow when designing and conducting grammar instructional activities for classroom practice. With the thinking frame, we visualized a detailed four-step cyclical procedure for teachers to follow to guide students' metacognitive learning and automated schema construction in English grammar practice: (1) make dialogic talk to confirm the problem in joint activities; (2) give conceptual knowledge, scientific, definitions and strategies in a systematic way to help construct cognitive schemas; (3) automate the schemas' retrieval through modeling and deliberate practice; and (4) emphasize reflection and facilitation of metacognitive schema construction to invoke awareness and self-regulation.
As Heap (1990, p. 43) noted, "If some activity is important in our lives, then knowing how it is organized may make a difference to how we act." Based on seeing the need for automated schema construction if students are to be effective problem-solvers, we designed the framework for this study to help teachers prepare instructional activities that would actively engage students in ongoing learning. The thinking framework equips teachers with a tool by organizing the theory and unfolding it into explicit and practical steps. In terms of implications for teaching, the thinking framework we provide would be of immense value in language teacher education and preparation programs. The three cases we reported demonstrated successful outcomes from implementing the four steps, and sometimes repeating the steps in the framework for automation of students' schema retrieval for developing their metacognition.
The scaffolding provided by our thinking framework enables teachers to understand the essence of transforming instruction from constructing to consolidating knowledge. The goal of the framework is not to restrict teachers to specific steps in specific order but to serve as a thinking tool to make effective teaching flexible and equip students with a tool increase the automaticity in their independent applications. The findings of this study demonstrate the effectiveness of our visualized thinking framework for informing classroom practice but at the same time raise questions for application to wider subjects and contexts.
Although our findings are valuable, there is a need for continuing research into and development of effective instructional methods in parts of English language teaching other than grammar such as writing, listening, and pronunciation. In addition, we only investigated English grammar teaching in a few classrooms in one secondary school in China, and thus, the implications of our findings might be limited in other similar settings. However, the theoretical framework we visualized for this study enhanced the teachers' English pedagogical grammar practice and their skills at facilitating students' metacognitive learning. We expect that by using the evidence here for the value of our framework as a thinking tool for understanding the essence of teaching and learning English grammar, future scholars can produce outputs that guide metacognitive teaching and learning over a broad range of disciplines. We hope this study will stimulate interest among language educators and serve as a start for further research and development of similar programs in future.
Funding Open Access funding enabled and organized by CAUL and its Member Institutions. The authors have received no outside funding for this research.

Data availability
The database is available for this research. Code availability: V-note.

Conflict of interest
There is no conflict of interest in this study.

Ethical approval The researchers' institutions granted ethics approval.
Consent to participate Yes.

Consent for publication Yes.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visithttp:// creat iveco mmons. org/ licen ses/ by/4. 0/