Teachers’ Perceptions of the Characteristics and Functions of Feedback in Different Assessment Situations in Secondary Education

Feedback in Secondary Education: Constraints and Patterns

Research indicates that multiple factors may facilitate or constrain feedback practices in secondary education (Contreras-Pérez & Zúñiga González, 2018; Schildkamp et al., 2020). A solid command of disciplinary knowledge (Schildkamp et al., 2020), the co-construction of teacher–student goals (Hargreaves, 2013; Röhl & Gärtner, 2021), the use of problem-posing questions (Furtak et al., 2016), and trust-building that supports students’ receptive uptake of critical feedback (To et al., 2024) contribute to effective classroom feedback. Accordingly, teachers need to create learning environments that foster students’ active participation in feedback (Carless, 2022; Carless & Boud, 2018). Teacher feedback is dialogic (Ajjawi & Boud, 2017) when it enables multiple communicative interactions (Ibarra-Sáiz et al., 2020) and centers students’ agency within those interactions (Nieminen et al., 2022).

Despite the growing visibility of feedback in school discourse and guidance, studies in school settings—particularly in secondary education—remain limited both nationally and internationally (N. E. Winstone et al., 2017). Existing research has largely focused on teachers’ conceptions and practices and on students’ perceptions of feedback (Panadero et al., 2020; Tan et al., 2019; F. Van der Kleij et al., 2017; Vattøy & Smith, 2019). A complementary strand examines classroom barriers to feedback in order to articulate principles or models of good practice, offering guidance that addresses the form and content of feedback alongside the institutional and socio-affective conditions required for effective implementation (Hattie & Timperley, 2007; Panadero & Lipnevich, 2022). Research has also tended to concentrate on mathematics and the natural sciences, including the effects of feedback on learning (Yan et al., 2021) and the strategies employed (Akpinar, 2018), whereas comparatively less attention has been paid to humanities subjects such as history, language and literature, and writing (Klute et al., 2017; Schildkamp et al., 2020).

In Chile, feedback research is concentrated primarily in higher education, with comparatively less work conducted in school settings (Herrera Araya et al., 2023; Quezada Cáceres & Salinas Tapia, 2021). Collectively, Chilean studies describe a predominantly transmissive pattern. For instance, feedback practices in secondary settings have been characterized as grade-oriented, in that “feedback was directed to explain or justify the score and grade obtained by the student” (Contreras-Pérez & Zúñiga González, 2018, p. 17). They have also been described as relying on low-specificity approval and error marking, including “praising or congratulating him or her without being linked to their learning process,” alongside “little presence of suggestions for improvement” (Contreras-Pérez & Zúñiga González, 2018, p. 17). Complementing this, school-based evidence in mathematics suggests that teachers’ declared practices show “feedback oriented toward the evaluative rather than the descriptive” (Muñoz Lira, 2020, p. 1).

Across the articles reviewed, a convergent pattern emerges: in secondary schooling, feedback is often enacted as a largely transmissive and evaluative practice, closely tied to marking, correction, and error identification rather than sustained guidance for improvement (Contreras-Pérez & Zúñiga González, 2018; Muñoz Lira, 2020). Chilean evidence, in particular, depicts feedback that frequently functions to justify grades and signal correctness, with limited specificity and few actionable suggestions, while contextual constraints—especially restricted time and large class sizes—further limit the feasibility of more descriptive, improvement-oriented approaches (Contreras-Pérez & Zúñiga González, 2018; Muñoz Lira, 2020; Quezada Cáceres & Salinas Tapia, 2021).

Broader syntheses caution against treating feedback as a single, uniform intervention. Meta-analytic evidence shows wide variability in effects across feedback forms and conditions, highlighting the need to identify moderators rather than rely on an average effect (Wisniewski et al., 2019). Discourse-oriented work also notes inconsistent usage of the term, which often conflates feedback information with the processes through which learners interpret and use it, complicating cross-study comparisons and strengthening the case for explicit conceptual positioning (N. Winstone et al., 2022). Finally, systematic review evidence suggests that learners’ agentic engagement with feedback remains underexamined, limiting inferences about how students take up and act on feedback in everyday secondary classrooms (N. E. Winstone et al., 2017).

Taken together, this evidence suggests a persistent tension between aspirations of feedback as a dialogic, student-engaging practice and how it is often enacted in secondary classrooms under everyday constraints. Crucially, it also indicates that feedback cannot be treated as a uniform construct: “Feedback must be recognized as a complex and differentiated construct that includes many different forms with, at times, quite different effects on student learning” (Wisniewski et al., 2019, p. 213). Moreover, “the term ‘feedback’ is used by different authors to refer to very different representations of the concept” (N. Winstone et al., 2022, p. 213), so interpretation depends on the conceptual lens adopted. As N. Winstone et al. (2022) note, “interpretation . . . is dependent on the assumptions of the writer” (p. 225) and authors should “signal this explicitly” to support clearer communication across studies. Accordingly, the next section outlines the main conceptual perspectives shaping feedback research and clarifies the analytical lens through which teacher feedback is understood in this study.

Conceptual Perspectives on Teacher Feedback

Currently, two theoretical perspectives dominate feedback research (Ajjawi & Boud, 2018). The information-transmission perspective conceives feedback as a monologic practice focused on delivering content to raise academic performance (Ajjawi & Boud, 2017). By contrast, the socioconstructivist perspective understands feedback as a complex social phenomenon (Carless & Winstone, 2023): a socially constructed interaction mediated by dialog (Carless & Boud, 2018) and embedded in the communicative exchanges characteristic of dialogic teaching (Cui & Teo, 2021). From this vantage point, feedback is defined as an interaction between teachers and students (Malecka et al., 2022; Steen-Utheim & Wittek, 2017).

Much of the literature still rests on basic assumptions about feedback (Sutton, 2012). Work aligned with the information-transmission perspective examines the effectiveness of feedback as a cognitive product oriented toward academic performance, drawing on a tradition that models feedback as a linear system (Lam, 2017; Nieminen et al., 2023). This linear conceptualization frames feedback as a structured sequence that constrains learners’ agency to the mechanical attainment of tasks or learning goals according to evaluative–normative criteria (Lam, 2017; Sutton, 2012). In doing so, it downplays the social and interactive nature of feedback its capacity to support meaning-making and the negotiation of understanding in knowledge construction (Lam, 2017). In contrast, the socioconstructivist perspective treats feedback as a privileged space for the co-construction of learning and, therefore, as a dynamic and multi-communicative social phenomenon (Malecka et al., 2022; Steen-Utheim & Wittek, 2017).

This study conceptualizes feedback as an interaction in which information is co-constructed by the participants (teacher and student, or peers) engaged in a learning activity, with the aim of improving performance, consolidating a competence, and/or advancing more complex learning (Lipnevich & Panadero, 2021; N. E. Winstone et al., 2017). From this perspective, feedback generates knowledge that is enacted in interactions between the learner and artifacts, or among learners and artifacts, enabling the interpretation and use of feedback (Nicol, 2021).

Accordingly, effective feedback is organized around three guiding questions—Where am I going? How am I going? and Where to next?—which map onto four feedback levels: task, process, self-regulation and self (Hattie & Timperley, 2007). Evidence shows that targeting the first three levels helps close students’ learning gaps and fosters their active participation (Panadero & Alqassab, 2019).

Teacher feedback is defined as the strategies or actions carried out by the teacher to comment, correct, guide, ask or problematize the performance of students in different assessment situations (Black & Wiliam, 2018; Boud & Molloy, 2013). Considering this delimitation, teacher feedback is categorized into three types. One of them is evaluative feedback, which is conceived as the delivery of a value judgment. This type of feedback can be positive as an act of rewarding or approving the work, or negative, which is linked to punishing or disapproving the learning product. A second type refers to descriptive feedback, which involves a process to identify strengths and weaknesses. This is constituted in for achievement, associated with the need to specify or orient the work, and to improve, associated with the need to construct and design paths for learning. Finally, dialogic feedback refers to, verbal and non-verbal interaction that provides emotional and relational support, maintains a dialog between subjects, provides opportunities for students to express themselves and facilitates the contribution of others to the construction of individual knowledge (Steen-Utheim & Wittek, 2017), linked to self-reflection actions through internal feedback (Nicol, 2021).

Methodological Approach and Strategy

The study adopts a qualitative approach. This approach seeks to understand participants’ experiences of a specific phenomenon (Flick, 2015) on the premise that social interaction gives meaning to reality (Marshall & Rossman, 2011). Making sense of reality entails interpreting the phenomenon in light of the meanings created by participants (Flick, 2015). The methodological strategy is interview-based research (Kvale, 2011), which provides access to and facilitates the interpretation of the meanings of participants’ experiences as knowledge grounded in lived experience (Brinkmann & Kvale, 2018). Interview research is an interaction between interviewer and interviewee that allows participants to articulate their experiences in their own words and from their own perspective, while the interviewer interprets the meanings conveyed in their responses (Brinkmann, 2018).

Participants

Participants were fifteen teachers of History, Geography, and Social Sciences who taught in the first through fourth years of secondary education. Teachers were recruited using purposive, non-probability sampling appropriate for qualitative research (Ballestín & Fàbregues, 2018) based on: (a) professional profile by teaching experience—novice (0–5 years) and experienced (>5 years)—, given that feedback practices evolve over teachers’ careers (Doornkamp et al., 2022), (b) gender (eight men and seven women), as interactional patterns shape the types of feedback enacted in classrooms (Espinoza & Taut, 2016; Máñez et al., 2024), and (c) school administrative governance: private (n = 5), private-subsidized—with state contribution—(n = 5) and public school (n = 5) because feedback practices vary by students’ socioeconomic characteristics (Agencia de Calidad de la Educación, 2019). Sample characteristics are summarized in Table 1.

OPEN IN VIEWER

Table 1.

Teacher Selection Criteria.

School type	Teacher ID	Gender	Age	Years of experience
Private (fee-paying)	D1	Female	28	4
	D2	Male	29	5
	D3	Female	32	8
	D4	Male	36	12
	D5	Female	33	9
State-subsidized private	D6	Female	29	5
	D7	Male	30	6
	D8	Female	33	9
	D9	Male	30	6
	D10	Female	35	10
Public	D11	Male	31	7
	D12	Female	33	8
	D13	Male	36	12
	D14	Female	35	10
	D15	Female	30	6

A 5–5–5 distribution across school types, with gender balance and representation of professional trajectories (novice and experienced), ensures comparable internal contrasts and sufficient coverage of typical and boundary cases, maximizing the sample’s information power without sacrificing the analytic depth intrinsic to qualitative interviewing (Brinkmann, 2018). This sampling design enables intra and intergroup comparisons, facilitates theoretical iteration during analysis, and supports thick descriptions of the phenomenon, a necessary condition for constructing situated knowledge grounded in teachers’ lived experience.

Procedure, Instrument, and Technique

Individual interviews were conducted by two researchers—PhD candidate in Education and PhD Educational Psychology—whose lines of research focus on feedback, educational assessment, and authentic assessment. In total, one researcher conducted 10 interviews and the other five. Each teacher participated in a one-on-one interview lasting 40 to 60 min via the Zoom videoconferencing platform. Zoom was chosen because the researchers were based in different cities and interviews were scheduled in the evening, outside teachers’ instructional hours, by mutual agreement between researchers and participants. The online platform enabled audio–video recording, which facilitated verbatim transcription given the high audio quality.

Participants were contacted at the start of the study by email that included: (a) the study context, (b) the nature of participation, and (c) an informed-consent invitation to take part in the study. Because both researchers teach at the undergraduate and graduate levels in education, invitations strictly aligned with the predefined selection criteria were extended to teachers who were former students or master’s students enrolled in their course sections.

An interview protocol was developed to elicit teachers’ perceptions of the feedback practices they report implementing across different assessment situations. The protocol covered the following dimensions: meaning, assessment situations, purpose, organization, strategies/actions, influencing factors, and examples of better feedback practices. The question guide, co-constructed by both researchers, drew on deductive categories reported in the feedback literature, specifically, categories concerning teachers’ self-reported feedback practices (Esterhazy et al., 2020) and on common assessment experiences implemented in school classrooms. The detail of the interview script is presented in Table 2.

OPEN IN VIEWER

Table 2.

Components of the Teacher Interview Script About Feedback.

Dimensions	Components or questions
Meaning	What do you understand by feedback?
Assessment situations	Ask the teacher to indicate whether or not they give feedback in the following assessment situations and, if they do, to describe how they do it in:
	1. Presentations (dissertation, forum, dramatizations, etc.)
	2. Written works (essay, report, project, etc.)
	3. Practical work (models, posters, etc.)
	4. Guides (exercises, questions, etc.)
	5. Short activities (from pages of the school textbook or other similar ones)
	6. Tests and exams
	7. Other experiences (if the teacher indicates others that are not covered)
Purpose	What is the purpose or objective of this feedback?
	Does the purpose of feedback change when it is developed or carried out in the classroom? Why?
Organization	When you organize the feedback for some of this/these situation(s), what aspects or criteria do you consider? Why those aspects or criteria?
	When you design the feedback, do you have any kind of pedagogical support?
Strategies or actions	What feedback strategies or actions do you use? Why these strategies or actions?
	What kind of feedback do you give on a daily basis in assessment situations?
Factors	Considering the following factors: time, number of teaching hours, and number of students per course, how do these factors affect your feedback practice?
“Better” feedback	Describe the best feedback practice you have done in the classroom. Why do you consider it the best feedback practice you have done?

Data Analysis

Individual interviews were analyzed using deductive content analysis (Kuckartz & Rädiker, 2023), grounded in an a priori categorical framework derived from the literature on teacher feedback and self-reported practices and operationalized in the semi-structured interview protocol. The categories and operational definitions were:

The analytic procedure comprised: (a) verbatim transcription of the 15 interviews, (b) construction of a coding matrix aligned with the categorical framework, (c) coding of textual fragments (meaningful units—clauses/sentences with complete sense—as the unit of analysis) and (d) thematic interpretation.

Given that the interviews elicit teachers’ perceptions, we analytically distinguished among: (a) beliefs/conceptual frames, (b) declared norms or intentions, (c) reported practices supported by a situated example and (d) perceived evidence of change. These levels were recorded under the attribute “statement type.” The entire coding process was conducted by one of the researchers. To this end, a coding guide grounded in both the interview question guide and the definitions of the categorical framework was used to distinguish and decide how to code segments containing units of meaning that might require prioritization or categorical co-coding.

The coding guide is presented in Table 3.

OPEN IN VIEWER

Table 3.

Coding Guide.

Dimension	Category	Decision rules and co-coding
Characteristics of feedback	Meaning	If the account shows bidirectional interaction involving questions, agreements, or negotiation of criteria, prioritize “dialogic,” co-code with “orientation: feed forward” when actions are projected.
	Assessment situations	Co-code with “strategies/actions” (how feedback is delivered) and “organization” (when and with what support).
	Organization	If the emphasis is on “when” or the “supports,” code here, if it also makes the “purpose” explicit, co-code with “purpose.”
	Strategies and actions	Co-code with “type” (evaluative, descriptive, dialogic) and with “Focus” (task, process, self-regulation) when the agent of the action (teacher or student) is specified.
	Contextual factors	Treat as explanatory context (not as a strategy). Useful for matrices and interpretive discussion.
	Perceived “best practice”	May be co-coded with “strategies,” “organization,” and “functions” when the passage permits.
Functions of feedback	Purpose	If the passage makes the purpose explicit, code here, co-code with “orientation” and “focus” when appropriate.
	Orientation	Mentions of “before or prior,” code as feed up, “during development” as feedback, and “after an activity or grade” as feed forward.
	Focus	Code according to the dominant level (task, process, or self-regulation), if two levels are explicit and distinguishable, allow co-coding.

A subgroup grid by school type (private, state-subsidized, public) was used to track the emergence and stability of indicators. Thematic saturation was declared when no new codes or relevant variations appeared in the final interviews. A 5–5–5 distribution across school types, together with diversity in professional trajectories, increased information power without compromising analytic depth.

A second researcher reviewed selected excerpts to verify categorical coherence. No discrepancies were identified between the operational definitions and the textual segments. The aim was interpretive consistency rather than the estimation of statistical reliability.

NVivo 12 was used in a neutral for corpus management (organization, queries, matrices, reports), analytic quality rests on the methodological logic and documented coding/interpretive decisions (Bezeley & Jackson, 2013).

Finally, results were organized into two dimensions. Characteristics included: meaning, assessment situations, organization, strategies/actions, factors and perceived quality (“better practice”). Functions integrated: purposes, orientation and focus.

Dimension: Characteristics of Feedback

“Better Practice” (Perceived Quality)

Perceived high-quality feedback is associated with procedural sequences featuring explicit criteria, problematizing questions, and active participation: “the best feedback. . . has been in those instances where we set clear milestones, question their decisions and they respond to each other, constructing a space for the active participation of students” (D15).

The findings for this dimension are summarized in Table 4.

OPEN IN VIEWER

Table 4.

Summary of Findings: Characteristics of Feedback.

Category	Reported finding	Device/modality	Type of feedback
Meaning	Pedagogical process to guide improvement, part of formative assessment, clarifies gaps and positions students’ step by step goals and criteria (D1, D9).	—	Descriptive
Assessment situations	Presentations. The rubric functions as an “alignment device,” makes criteria explicit and anchors improvement actions (D13).	Rubrics + oral comments.	Descriptive
	Written assignments. Imperative cues (“complete,” “improve,” “elaborate”) and error identification (D6).	Brief written comment.	Evaluative
	Guided activities/practical work. Group oral feedback predominates, broad coverage but low specificity (D7).	Group oral feedback without specific guidance for improvement.	Descriptive–dialogic
	Tests. Use of selected-response items to model and detect errors, review focused on content (DSE8).	Guided review + modeling.	Evaluative
Organization	(a) Disciplinary/goal criteria (D6), (b) cross-cutting criteria (collaboration/communication; D7), (c) contextualization with positive reinforcement (D14). Modeling and “guided speech” activate participation (DE5).	Rubrics, prompts, and verbal scaffolding	Descriptive
Strategies/actions	Modeling and guided questioning; oral as the main channel due to immediacy (DE5, DE11).	Immediate oral, secondary written.	Descriptive
Factors	Time, instructional hours, and class size determine modality, timeliness, and depth (D10, D13).	—	—
“Better practice”	Sequences with explicit criteria, problematizing questions, and active participation (D15).	Milestones + guided peer discussion.	Dialogic

Dimension: Functions of Feedback

Focus

Teachers indicate that task-focused feedback predominates in tests and brief activities. Process-focused feedback becomes more salient in written assignments and presentations through elaborative comments and modeling. Finally, a self-regulation focus emerges when questions invite students to justify decisions and plan next steps: “we elicit ideas and promote participation. . . where they are the ones answering” (DE5).

The findings for this dimension are summarized in Table 5.

OPEN IN VIEWER

Table 5.

Summary of Findings: Functions of Feedback.

Category	What the results show	Where it is identified
Purpose (accompaniment and use of error)	Support learning processes and use error as a didactic resource, the teacher does not possess all the answers.	Cross-cutting, adjusted to students’ formative trajectories.
Variability of purpose	The purpose of feedback varies across situations and adapts to needs, interactions and levels of achievement.	Changes across assessment situations and by level of achievement.
Feed up (explicit criteria)	Criteria are made explicit through rubrics that align expectations and goals.	Presentations, at the outset/before tasks.
Feedback (contrast and error)	Performance is contrasted with criteria, responses are reviewed, errors are detected and explained.	Tests and written assignments.
Feed forward (next steps)	Concrete directions for improvement for the next cycle or submission.	Written assignments and presentations.
Task focus	Correction of the product in brief activities and tests.	Tests, short-answer activities.
Process focus	Elaborative comments and modeling (guided speech).	Written assignments and presentations.
Self-regulation focus	Questions that invite justification of decisions and planning of next steps (emergent).	When metacognitive questioning occurs in class.

Characteristics of Feedback: From Perceptions to Self-Reported Practice Patterns

The characteristics of feedback, as perceived by teachers regarding their declared practices, align with the literature on the relationship between meanings (Deneen & Boud, 2014; Fulmer et al., 2015) and feedback practices (Lee et al., 2020; Schildkamp et al., 2020). Teachers understand feedback as a pedagogical process that is part of formative assessment and enables the identification of errors and strengths. These aspects become a reference framework that potentially shapes the feedback practices they report (Black & Wiliam, 2018; Yates & Johnston, 2018).

Based on teachers’ self-reports, feedback is centered on what the teacher does and says (Mandouit, 2020). However, beyond this declarative alignment, tensions emerge between the conceptual framework and the reported enactment: in classroom contexts, emphasis frequently shifts from descriptive to evaluative feedback, which limits the depth of exchanges and the joint construction of meaning a phenomenon already noted in studies contrasting dialogic approaches with predominantly descriptive uses (Contreras-Pérez & Zúñiga-González, 2017; Quezada Cáceres & Salinas Tapia, 2021).

Teachers also report primarily implementing immediate oral feedback because it is feasible and timely, turning to written feedback in more limited situations. This pattern is consistent with reports on the utility of timely feedback, but also with the need to sustain a dialogic approach so that timeliness translates into use rather than mere exposure to information (F. M. Van der Kleij et al., 2019). In our case, the oral channel enhances immediacy, although without recording supports (e.g. follow-up prompts or micro-tasks) it is difficult to trace feed-forward progression across iterations. Even so, when describing their “better feedback,” a dialogic form emerges that allows students’ participation in the construction of collective knowledge (Steen-Utheim & Wittek, 2017) and fosters self-reflection through internal feedback (Nicol, 2021). This finding is consistent with other research on teacher feedback in schools (Elliott et al., 2020), since teachers associate “good” feedback with a dialogic model for its development (F. M. Van der Kleij et al., 2019).

Teachers indicate that across different assessment situations they use directed or open questions at varying levels of cognitive complexity as feedback strategies or actions. This resembles findings reported elsewhere (Furtak et al., 2016). However, studies also indicate that the effect of questions as a feedback strategy is limited when compared with more elaborated feedback for example, providing a detailed explanation (F. M. Van der Kleij et al., 2015).

The results also highlight the careful attention given to tone and formulation of comments that teachers report using to support students’ motivation and learning trajectories an essential aspect for avoiding inhibitory effects and promoting active reception of information. Such relational care fosters conditions for dialog but does not guarantee it: when time is scarce or class sizes are large, interaction is simplified and drifts toward general indications, which affects students’ understanding and application of information to improve their learning trajectories (F. Van der Kleij et al., 2017; F. Van Der Kleij & Adie, 2020).

Contextual factors—available time, workload, number of students, and institutional conditions—emerge as decisive modulators of the modality (oral/written), timing (immediate/delayed), and depth (general/specific) of feedback. This helps explain why practices that teachers value as “better” are not sustained on a daily basis. The literature has emphasized that, by optimizing conditions and sequences, feedback can become a more effective activity (Panadero et al., 2020; F. M. Van der Kleij & Lipnevich, 2021). We interpret that the challenge is not only technical (what to say or which tool to use), but also one of design and structuring: how to foster moments, supports, and interactions so that evaluative dialog occurs systematically and iteratively in learning-oriented improvement cycles.

Functions of Feedback: Purpose, Orientation and Focus

The function of feedback that predominates in the interviews is formative, in the sense of accompanying learning, providing guidance, and using error as a resource, which is consistent with definitions that emphasize guiding learning rather than merely justifying a judgment or grade (Ruiz-Primo & Brookhart, 2018). Nonetheless, the move from “indicating what happened” to “specifying how to proceed” is irregular: feed up is usually better established (criteria are made explicit and known), whereas feed forward becomes contingent on the situation and working conditions, generating gaps between performance evaluation and the planning concrete next steps.

Regarding whether the purpose of feedback changes across assessment situations, teachers report that it is flexible and adapts to students’ learning trajectories. In this sense, feedback is adaptive yet remains teacher-centered, grounded in what the teacher does and says (Klute et al., 2017; Lee et al., 2020; Mandouit, 2020), oriented toward improvement but without consistently making explicit to students the concrete ways to improve. In practice, this slippage can bring feedback closer to evaluative uses, particularly when it functions to legitimize grades rather than to support feed forward through actionable guidance.

Three ways of organizing feedback were also identified: (a) criteria about content composed of possible student errors, (b) needs for improving learning and mastery of the content under study, and (c) criteria about transversal aspects of performance with an emphasis on values or attitudes for practical work, alongside the contextualization of evaluative criteria to generate improvement actions, positive reinforcement, and identification of strengths. Consequently, the ways of organizing feedback through instruments emphasize criteria oriented toward “next time” that is, there is an absence of feedback before and during assessment, with attention centered on strengthening the subsequent task (Hattie et al., 2021). This progression indicates weak internal coherence in the feedback cycle and suggests that its enactment depends on the nature of the task.

Disaggregating the foci, there is a predominance of task-level feedback, particularly in brief activities and tests where attention centers on product correction, alongside a more situated use of process-level feedback in written work and presentations through elaborative comments and modeling via guided talk. A focus on self-regulation emerges when metacognitive questions are posed in class, for example when students are asked to justify decisions or plan concrete next steps, indicating a still incipient potential to promote student agency beyond product correction. Taken together, the results depict feedback that moves from the initial alignment of expectations through feed up toward decision-making for improvement through feed forward yet still concentrates its efforts in specific moments and formats. This opens space to strengthen dialogic instances and metacognitive scaffolding more systematically, enabling students to use criteria and errors as levers for learning.

A relevant point is the apparent contradiction between deficits highlighted by some studies, such as limited assessment competencies or predominantly corrective uses (Gysling, 2017; Herrera Araya et al., 2023), and the capacity that teachers in our study report to design sequences with explicit criteria, problematizing questions, and improvement guidance. The key interpretive issue is not a lack of knowledge but conditions of implementation. As temporal pressure and class size increase, feedback becomes more general and less dialogic, even when teachers are aware of more productive practices. In this regard, connecting feed up, feedback, and feed forward requires organizational scaffolding that protects time for specifying how to proceed and for making concrete next steps visible and usable (Brooks et al., 2019).

From a practice-oriented perspective, the findings point to several feasible actions for shifting feedback away from evaluative correction and toward dialogic use and developing self-regulation, even under constraints of time and class size. Rubrics can be used not only to make criteria explicit through feed up, but also to structure brief dialogic exchanges by prompting criterion-referenced questions that require justification. In addition, modeling and guided questioning can be organized into short, repeatable classroom routines that move from shared criteria, to an example or model, to a problematizing question, to student justification, and then to a prompt about what to do next, so that immediacy is translated into use rather than mere exposure. Teachers can also embed milestones and structured peer discussion within presentations or writing tasks in order to redistribute participation and make feedback more collective and iterative. Finally, when error identification is necessary, as in tests, teachers can prompt students to explain why an option is incorrect and to identify a specific next step, thereby strengthening the still incipient focus on self-regulation reported in the interviews.

Implications of the Study

Using the interview guide made it possible to investigate teachers’ perceptions of their feedback practices according to the characteristics and functions they report. The categories employed shed light on how teachers’ self-reported feedback practices are useful for identifying strengths and opportunities for improvement to enhance classroom feedback.

This is relevant because the findings indicate that feedback goes beyond the mere transmission of information. It should be understood as a dialogic practice that integrates feedup-feedback-feedforward in iterative cycles, attentive to the contextual conditions of the school environment.

The need to integrate participation and self-regulation as focal points for feedback suggests explicitly incorporating frameworks that connect feedback with regulation of learning (goals, monitoring, and planning), thereby avoiding a “product-correction” bias.

The internal coherence of the feedback cycle, given the relative strength of feed up and the contingency of feed forward, indicates that theory should emphasize “sequencing and continuity” as quality criteria for feedback, not only clarity and specificity. Therefore, there is a need—and a recommendation—to advance teacher professional learning through feedback literacy initiatives (Carless & Winstone, 2023; Gravett & Carless, 2024) designed for the school classroom.

Limitations and Future Research

The limitations of this study are primarily methodological. Because the evidence relies on teachers’ self-reported perceptions elicited through semi-structured interviews, the findings should be interpreted as accounts of how feedback is understood and reported across assessment situations rather than as direct evidence of classroom enactment or impact. Although the interview guide aligned with the categorical framework proved consistent for analyzing reported characteristics and functions of feedback, it remains necessary to complement this strategy with (a) analysis of written feedback artifacts, (b) observations of practice, (c) interactional evidence of feedback exchanges (teacher–student and peer-to-peer), and (d) students’ perspectives (“learner voice”) on how feedback is understood, taken up and used, in order to better connect reported practice with perceived effectiveness. Additionally, the recruitment pathway, where invitations were extended to teachers who were former students or master’s students enrolled in the researchers’ course sections, may have shaped how practices were represented. Finally, while a second researcher reviewed selected excerpts to verify categorical coherence, the full coding process was conducted by one researcher, thus, the analytic strategy prioritized interpretive consistency over formal estimation of inter-coder reliability.

Likewise, this study highlights substantive aspects of teacher feedback that warrant further inquiry. Future research should examine how assessment instruments function as scaffolds for feedback processes, how evaluative criteria concerning content, transversal performance, and contextualization shape feedback characteristics and functions and enable learning improvement, and how these elements are integrated into feedback processes across assessment situations. It is also important to investigate how different assessment situations foster students’ active participation in feedback and influence the feedback’s focus (task, process, self-regulation) and orientation (feed up, feedback, feed forward), using designs that incorporate more “encompassing” evidence to problematize and strengthen the salience of the findings, including the use of reporting guidance such as COREQ.