Toward a sociocultural approach to feedback provision in L2 writing classrooms: the alignment of dynamic assessment and teacher error feedback

Abstract

Feedback is a key component of teaching second language writing. As the field of second language writing instruction has recently placed greater emphasis on sociocultural considerations, issues related to the nature, negotiation, and delivery of feedback need to be re-examined. The present paper sought to provide a sociocultural basis for such a re-examination by integrating Dynamic Assessment (DA) into the revising stage of the writing process in L2 writing classes. The paper discussed how such integration might help improve the nature of teacher error feedback and its delivery in L2 writing classes.

Background

Feedback is widely seen as crucial for encouraging and consolidating learning (Anderson, 1982; Brophy, 1981; Vygotsky, 1978). It is also regarded as an essential factor in the writing context where the importance of feedback was highlighted with the development of “the process approach” to writing instruction. The process approach places a high value on teacher-student encounters around text issues and encourages teachers to shift the focus away from the final written product in an effort to support writers through multiple drafts by providing them with feedback during the process of writing. Its importance is also acknowledged by genre-oriented approaches, where great emphasis is placed on sociocultural theories of scaffolded instruction and learning as a social practice. Here classroom situations and interactions, if pitched appropriately and meaningfully at learners’ zones of proximal development, can help in diverse ways to scaffold people’s acquisition of text forms, composing processes, and purposeful social interactions through writing in the second language.

As sociocultural considerations have recently become the focus of attention in the field of second language writing instruction, issues related to the nature, form, and value of feedback have had to be re-examined. The present paper provides a basis for such a re-examination by seeking to incorporate Dynamic Assessment (henceforth, DA) concepts and strategies into teacher feedback which is the most widely adopted feedback technique in second language writing classrooms. To this end, we will first explain why DA forms the basis for our re-examination by tracing its genesis to Vygotsky’s sociocultural theory (SCT) of mind and by exploring its relevance to some key concepts of SCT. We will then try to integrate DA with teacher feedback in second language writing classes by investigating the potential benefits of DA in promoting the most dominant feedback type. In so doing, we will focus on how adopting a DA approach to feedback provision helps change the nature of teacher error feedback and reshape its negotiation and delivery.

Why dynamic assessment?

Genesis of DA in Vygotsky’s work

Dorfler et al. (2009) have defined DA as “one approach to gaining insight into the current level of competence as well as into how this competence can be influenced by specific educational interventions”. DA is basically an approach which understands individual differences and their implications for instruction and embeds intervention within the assessment procedure by including appropriate forms of mediation that are sensitive to the individual’s (or in some cases a group’s) current abilities and the assessment of the effects of that mediation on subsequent performance with the aim of promoting learner development. From this definition it is clear that DA considers abilities to be “malleable and flexible rather than fixed” (Sternberg & Grigorenko, 2002) and focuses “on modifiability and on producing suggestions for interventions that appear successful in facilitating improved learner performance” (Lidz, 1991).

It is also clear that DA has certain definitive characteristics that are rooted in Vygotsky’s sociocultural theory of mind. The most definitive feature of DA is its interactive nature. In the DA model the examiner is an active part of the assessment and functions as an assessment tool, responding to observations and inferences about the learner and functioning in a way to reveal learning processes and to facilitate change. This closely parallels how sociocultural theory sees learning, including language learning. According to SCT, learning is dialogically based; that is, acquisition occurs in rather than as a result of interaction. From this perspective, then, L2 learning cannot be treated as a purely individual-based process but rather as one shared between the individual and other persons. Dialogic interaction enables an expert (such as a teacher) to create a context in which novices can participate actively in their own learning and in which the expert can fine-tune the support that the novices are given (Anton, 1999). In particular, dialogic discourse demonstrates what a learner can and cannot do with assistance. The second definitive feature of DA is the focus on learning processes, usually meta-cognitive processes, rather than the products of learning. The interaction between the examiner and the learner reveals how the student engages in the problem-solving process and promotes inferences about mental processes involved in task engagement. This feature also has its roots inVygotsky’s observation that a body can show what it is only in movement (Lidz and Gindis, 2003). Vygotsky believed that if we want to understand learning and development, we have to focus on the process instead of the product. According to Lantolf and Thorne (2006), Vygotsky argued that “the only appropriate way of understanding and explaining … forms of human mental functioning is by studying the process, not the outcome of development”. The third characteristic of DA is its perspective on assessment and the kind of assessment information it yields. DA embraces the assumptions that any assessment must include a specification of the degree to which the testee's performance is modifiable and that assessment must include an active teaching process aimed at modifying cognitive functioning. The extent of modifiability is an indicator of the person's potential learning capacity in future. As such, DA is intended to produce change in the examinee—within the testing situation—and assesses the implementation of learned strategies and cognitive principles in progressively more difficult problems. The individual’s level of performance after mediation points to his/her ability to benefit from mediation and provides more accurate indications about future treatment procedures and the prognosis of academic success. In this regard, the DA procedure is very similar to the diagnostic processes developed by Vygotsky (1978) for assessment of the ZPD. Vygotsky argued that interpreting static test results (e.g., IQ) as a reflection of the child’s abilities is misleading, because, in reality, the child’s performance reflects his or her entire socioeducational history. Moreover, given that the natural form of learning is collaborative, the assessment situation also should be constructed around collaborative interaction. DA seems to have caught this proposition when it places great emphasis on the unification of assessment and instruction. In DA, assessment and instruction are a single activity that seeks to simultaneously diagnose and promote learner development by offering learners mediation. Mediation is provided during the assessment procedure and is intended to bring to light underlying problems and help learners overcome them and move to the next level of competence.

Linking DA to key concepts of SCT

Grounded in Vygotskyan theory of development, DA closely links to several SCT concepts, in particular the concepts of mediation, ZPD, and scaffolding. First, SCT postulates that the adult human mind is mediated; that is, through social activity genetically endowed capacities are modified and reorganized into higher mental activity forms, which allows individuals to exercise conscious control over such mental activities as voluntary attention, logical reasoning, planning, and problem-solving. Taking this postulation into account, DA techniques provide learners with a “mediated learning experience” (Lidz, 1991) in which, through social interaction, experiences are filtered, focused, and interpreted as needed by the learner. In mediated learning experiences the examiner may guide learners in highlighting important content, making connections, setting goals, planning, regulating and controlling behavior, etc. In sum, the examiner provides the learner with a scaffold that may allow the learner to improve subsequent unassisted performance. For the examiner to be able to give successful mediation, s/he should take note of “contingent responsivity” (Lidz, 1991) which refers to the ability of the mediator to respond to the learner’s behavior timely and appropriately. This means teachers need to balance the giving and withholding of guidance and assistance in accordance with students' progression through a task (Aljaafreh and Lantolf, 1994). DA studies show that verbalization and elaborated feedback are two of the most powerful elements of mediated learning experiences.

The second SCT concept that is closely related to DA is the concept of ZPD. The ZPD is ‘the distance between the actual developmental level as determined by independent problem solving and the level of potential development as determined through problem solving under adult guidance or in collaboration with more capable peers’ (Vygotsky, 1978). The ZPD is considered as an emergent and open-ended trait of the learner that unfolds through interaction and expands the potential for learning by providing opportunities which are not anticipated in the first place (Wells, 1998). Applied to language learning, the concept of the ZPD brings together all of the relevant pieces of the language learning situation including ‘the teacher, the learner, their social and cultural history, their goals and motives, as well as the resources available to them, including those that are dialogically constructed together’ (Aljaafreh and Lantolf, 1994). In short, from this perspective, learning principally takes place within the learner’s ZPD. Vygotsky (Minick, 1987) attempted to relate the ZPD and assessment by calling for a change from symptomatic assessment focusing on characteristic behavior of a particular stage in development to diagnostic assessment focusing on understanding behaviors and developing recommendations to foster development. Vygotsky hypothesized that intervention would benefit learners with a broad ZPD, that is, with a high degree of readiness. For Vygotsky, the application of ZPD to assessment was a way to provide a more complete picture of the learner’s actual stage of development and of the proximal phase. Vygotsky was mostly concerned with qualitative assessment of psychological processes and the dynamics of their development. The fundamental link between the notion of ZPD and DA procedures is established by the idea that assessment that is entirely based on what the child is able to do without help ignores important differences in mental functioning that come to light when the child’s interaction with an expert is analyzed (Minick, 1987). Central to DA is the analysis of the learner’s performance in social interaction with the assessor. Thus, DA targets what individuals are able to do in cooperation with others rather than what they can do alone (Sternberg and Grigorenko, 2002). Furthermore, DA is not a standalone activity carried out in isolation from other pedagogical activities. It is instead an ongoing, development-oriented process of collaborative engagement that reveals the underlying causes of learners’ performance problems and helps learners overcome those problems. In other words, DA does not differentiate instructional activities from assessment activities because every mediator–learner interaction encompasses both types of activities. Instead, DA sessions vary according to learner development so that over time learners engage in increasingly complex tasks with less mediation, which explains why DA has replaced the drive toward standardization and measurement that characterizes many forms of assessment with descriptive and qualitative profiles of learner development.

The third notion crucial to the Vygotskian framework that is linked to DA is the concept of scaffolding which refers to a ‘situation where a knowledgeable participant can create supportive conditions in which the novice can participate and extend his or her current skills and knowledge to higher levels of competence’ (Donato, 1994). Through scaffolding, learners are able to use the target language with assistance from teachers or peers in the classroom to produce what they would not yet be able to do on their own (Sheen, 2010). Scaffolding- a joint process constructed on the basis of the learner’s need- is a function of the collaboration of both the learner and the expert operating within the learner’s ZPD. In such conditions, help is generated as a joint effort and through the supportive condition created in social interaction by the novice and the expert (Donato, 1994). Scaffolding is intended to provide a model of gradual release which is based on the idea that students require intensive instructional support when learning important skills and strategies that are challenging and in the zone of proximal development. The gradual release of responsibility model requires variable amounts of assistance. In the first stage, the teacher has high responsibility for modeling and explaining the learning task. In the second stage, the teacher and student share responsibility for learning. The student practices or approximates the task, and the teacher gives constructive feedback. When students are ready for the third and final stage, they take on all or nearly all of the responsibility for the work. DA is very much in line with the notion of scaffolding. In DA the underlying assumption is that learners need to be scaffolded and supported in their complex task of learning as they interact with the teacher or peers. The idea of scaffolding favoured by DA differs from the idea of helping the learner in a unidirectional way as characteristic of traditional teaching–learning activities; it is a collaborative process that involves negotiation of meaning between teacher and learner about expectations and how best to improve performance and is intended to provide guided support to the less knowledgeable partner (the novice) as s/he collaborates with a more knowledgeable partner (the expert). Scaffolding during DA uses insights about a learner's current understanding to alter the course of instruction and thus support the development of greater competence.

Having linked DA to the key concepts of SCT, we can conclude that the goal of DA is to measure, intervene, and modify behavior and to document the process of learning. Activity on the part of the examiner and the learner, and modifiability of behavior, are crucial in DA (Lidz 1991). Thus, DA is above all social, interactive, and qualitative (Lidz and Elliott, 2000). As such, DA has all the hallmarks of Vygotskyan theory of development and could be called the pedagogical instantiation of SCT, so it might benefit those teaching approaches that are grounded in SCT. Genre-oriented approaches to second language writing instruction, where learning as a social practice and sociocultural theories of scaffolded instruction are important, are no exception. While DA, once incorporated into various stages of the writing process, seems to be a promising approach to the teaching of writing in general, the focus of the present paper will be potential benefits DA provides writing instructors at the revising stage of the writing process. These include new insights into the nature, value, and form of feedback and ways of reshaping feedback practice, to which attention will now be turned.

DA changes the nature of error feedback in writing classrooms

There is no denying the fact that good feedback can significantly improve learning processes and outcomes. In writing classes, the ways teachers choose to express their feedback can affect both students’ reactions to it and the extent to which they use it in their revisions, and may have a significant impact on writing development. Therefore, the crux of the matter is how to deliver feedback correctly. DA seems to capture and integrate features of good feedback and define prescriptions relating to its appropriate delivery. In DA approaches, good feedback is facilitative in nature and tends to improve students’ learning of a particular skill or content area. DA exerts influence on features of feedback in several ways.

Feedback in DA is timely and task-level

In DA, feedback is intended to provide specific and often real-time information to the student about a particular response to a problem or task, and it may additionally take into account the student’s current understanding and ability level. For instance, a struggling student may require greater support and structure from a feedback message than a proficient student.

Feedback in DA, is an in-process and interactive endeavour

According to this view, feedback does not take place after learning has occurred but during the learning process. Feedback is thus similar to mediation, which is meant to support learner development and is an important component of the mediator-learner collaborative activity that occurs during the assessment itself. For this to happen requires that the mediation be sensitive to the ZPD. Thus, the mediator has to pay attention to how the learner responds to attempts at mediation and he or she must be prepared to adjust mediation accordingly.

Feedback in DA is scaffolded

Scaffolded feedback within DA a) motivates the learner’s interest related to the task, (b) simplifies the task to make it more manageable and achievable, (c) provides some direction to help the learner focus on achieving the goal, (d) clearly indicates the differences between the learner’s work and the standard or desired solution, (e) reduces frustration and risk, and (f) models and clearly defines the expectations (goals) of the activity to be performed. Scaffolded feedback, for instance, in a writing classroom may include models, text types, cues, revision prompts, hints, possible solutions to revision problems, as well as direct instruction (Hartman, 2002).

Feedback in DA is indirect

Indirect feedback occurs when the teacher indicates in some way that an error has been made but does not provide the correct form, leaving the student to solve the problem that has been called to his or her attention. Researchers have suggested that indirect error feedback is generally preferable because it forces students to engage in “guided learning and problem-solving” (Lalande, 1982) and helps them build skills as “independent self-editors” (Bates et al., 1993).

DA reshapes the negotiation and delivery of teacher error feedback

DA can help reshape teacher feedback practice by introducing a new approach to the teaching of writing in general and feedback provision in particular. We propose a DA-based approach to teaching writing that comprises three stages, namely orientation, execution, and control. At the orientation stage, students are stimulated to think about the writing process and the writing topic given. At this stage, mediation is aimed at the whole class. Mediation offered can be in the form of thinking prompts that are intended to help students generate ideas that they can use later in their writing. At the execution stage, students organize, plan, and write their first drafts. In this practice phase, the teacher might carry on instructional conversations with individual learners and offer mediation when necessary. At the control stage, - we’d also call it the DA feedback session- the teacher uses mediational strategies targeted at the whole class (Group DA) or at individual language learners in order to stimulate students to revise and edit their writing. In so doing, the teacher replaces traditional feedback with a meditational strategy called “talkback”, to use Lillis’s (2001) words, which refers to the exploratory talk aimed at “opening up discussion and … moving away from a tutor-directed talking space” (Lillis, 2001). Talkback, in DA feedback sessions, is intended to help the teacher and the students go through a 3-stage process of error treatment. The three stages, which occur in the order mentioned, are Identification of Trouble Source, Initiation of Repair, and Making Actual Repair. At the initial stage, the teacher and the students cooperate at the whole class in order to identify and locate potential sources of error in the texts the students have produced. It is the responsibility of the teacher to assure students that errors are expected and accepted as a part of their learning to write. By adopting a “correction-free approach” (Truscott 1996), the teacher helps students develop positive attitudes toward error correction At this stage, the instructor adopts a dialogical approach and, by means of conversational scaffolds, tries to enable students to spot their errors. In so doing, the instructor needs to consider the features of the specific genre in which the students have produced their texts and the types of errors students most likely commit while writing in that particular genre. The instructor need to observe that, for pedagogical purposes, student attention should first be drawn to treatable errors and then to untreatable ones because the former occur in a patterned, rule-governed way, whereas the latter occur in a haphazard way and there is no set of rules students can consult to avoid or fix those types of errors. At the second stage, the tutor negotiates with the tutees as to how they can fix the errors they have identified in the texts. He should note that the ways he conveys his praise or criticisms, and how he phrases his suggestions, are central to effective feedback; they represent key interpersonal resources for negotiating judgments and evaluations of student writing (Hyland and Hyland, 2001). It is suggested that the instructor simplify the task at hand, keep the learner on track, and draw students’ attention to key aspects of the task or its solution. At the same time, he should try to make the learners self-sufficient in managing the task by controlling only those elements of the task that are initially beyond the learners’ capacity and permitting them to concentrate upon and fix those elements that are within their range of competence. Teacher correction at this stage is indirect; that is, the teacher indicates the presence of an error without supplying the correct form or uses an error-coding system to signal the general category of an error. The teacher takes on some responsibility for correcting but places the burden of correction on the learner and leaves it up to the individual student to make the actual correction. There is evidence to suggest that prodding the learner to self-correct is effective in promoting acquisition (e.g., Lyster, 2004; Ferris, 2006). At the third stage, the tutees are supposed to enact feedback and adopt revisions in their subsequent drafts. The tutor is not actively engaged in this activity as long as the students are able to function independently. If, for some reason, the students fail to enact the feedback they have received at the previous stage, the instructor may give them implicit and, if necessary, explicit revision prompts and models a possible solution to the problem posed by the task. And while the students are engaged in the solo behaviour of making actual repair, the instructor monitors them looking for two signs of students’ appropriating the negotiated feedback: (a) indications that the students are processing the talkback as revealed in the student’s corrections of errors; and (b) indications that the students are in fact using the talkback to draw correct inferences as to how to treat different kinds of errors, even those which have not been drawn to their attention through the whole process.

The 3-stage process of error treatment discussed above has certain characteristics that make it mesh well with feedback provision in writing classrooms. First, it seems to have neatly encapsulated the principles of scaffolding formulated by Van Lier (1996): It creates a safe but challenging environment in which errors are expected and accepted as part of the learning process (the principle of Contextual support); it is based on mutual engagement and support (the principle of Intersubjectivity); in DA sessions, communication between participants is not forced, but flow in a natural way (the principle of Flow); the scaffolded assistance during DA depends on learners’ reactions: elements can be added, changed, deleted, repeated, etc. (the principle of Contingency); and the ZPD in DA includes repeated occurrences over time of a complex of actions, keeping a balance between routine and variation (the principle of Continuity) and closes when learner is ready to undertake similar tasks without help (the principle of Handover). Second, it helps create an appropriate context in which there is a match between teachers’ intentions and students’ performance. This, in turn, helps remove the causes of student inattention to teacher error feedback. Several reasons for student inattention and for the mismatch between teachers’ intentions and students’ performance have been put forth in the literature. In the first place, teacher feedback is oftentimes imprecise and full of vague comments that are incomprehensible to the students. By adopting a dialogical approach, the aforementioned DA feedback model keeps the teacher and the students actively involved in an ongoing negotiation of meanings and steers students through an evaluative process using a series of modifiable prompts that take into account students’ developmental readiness to enact the type of feedback and revision called for. And thus it helps improve clarity of teachers’ comments and minimize student inattention. One more reason for students’ inattention to teacher’s feedback has to do with their fear of being corrected. Some students may feel so threatened by the teachers’ comments that they develop a strategy of avoidance; they do not pay attention to the comments they feel uncomfortable with in order not to lose self-esteem in the process. One cause of this might be the teacher’s paying too much attention to errors. The 3-stage approach proves to be useful in alleviating this problem. It makes the teacher conscious of the potential feedback has for helping to create a supportive teaching environment and aware of the need for care when constructing his or her comments. It also reminds him or her that writing is very personal and that students’ motivation and self-confidence as writers may be damaged if they receive too much criticism (e.g., Connors and Lunsford, 1993) and that praising what a student does well is important, particularly for less able writers, and using praise reinforces good writing and fosters students’ self-esteem. It also requires that the teacher take the sting from some of his or her suggestions for improvement that may carry an implied criticism by adopting an “internal persuasiveness discourse”. Internal persuasive discourse that is “half-ours and half someone else’s” (Bakhtin, 1981) is contrasted with authoritative discourse, which is the more alien and static language of authority, and helps the teacher tone down the force of his comments and controls frustration. Thus, through controlling frustration, where the tutor provides the tutees with reassurance, the model minimizes student intimidation.

Conclusion

The current attention to feedback in education as a whole, and in the writing classroom in particular, is a positive development. But it is also the case that here, as in all areas of instructional innovation, the need for effective teacher professional development is greater than generally recognized and certainly greater than is currently available. Effective giving of feedback is a highly skilled teaching process, and we need to develop strategies and materials for helping teachers to acquire this skill. In this paper, we have attempted to show that DA promotes teacher feedback practice by exploring how it changes the nature and form of feedback and how it reshapes the negotiation and delivery of feedback in writing classrooms. As our discussion has shown, feedback within DA is a meditational strategy at the teacher’s disposal at the revising stage of the writing process to provide students with support tailored to their ZPDs. We have also introduced a DA- assisted commenting approach that meshes well with genre-oriented approaches to writing instruction where learning as a sociocultural activity is important. The alignment of DA and feedback provision has two main outcomes. On one hand, the picture that emerges of the ability of the learner in a helpful non-threatening DA environment is quite rich and seems to improve our understanding of what a learner is able to do in a particular learning/testing situation. On the other hand, from the DA experience one can derive “suggestions for intervention that appear successful in facilitating improved learner performance” (Lidz, 1991).

With these assumptions in mind, the time is therefore ripe for the adoption of a new approach to feedback provision in writing classrooms that is grounded in DA. However, it should be noted that unless these claims and assumptions have been empirically tested in a range of contexts, they remain little more than conjecture.

Authors’ information

Parva Panahi is currently doing her Ph.D. at Islamic Azad University, Tabriz branch. Her research interests include second language writing assessment, interactive assessment, and Dynamic Assessment.

Parviz Birjandi is a full professor holding an M.A in applied linguistics from the Colorado State University and a Ph.D in English education from the University of Colorado. His areas of interest include first language acquisition, and issues in language testing and assessment.

Behrooz Azabdaftari is a full professor of applied linguistics currently teaching at the Islamic University, Tabriz branch. His research interests include issues in first language acquisition, SLA, and Vygotsky’s theory of development.