Abstract
Introduction:
Written corrective feedback (WCF) is widely recognized as an effective pedagogical tool for improving learners’ grammatical accuracy in second language writing. However, most WCF research has focused on adult or tertiary-level learners, leaving senior high school students in English-as-a-foreign-language (EFL) contexts underexplored. This study addresses this gap by examining the comparative effectiveness of three WCF types—direct, indirect, and metalinguistic—on Pakistani senior high school learners’ acquisition of the English simple past tense. Grounded in Cognitive Load Theory (CLT) and Krashen’s Input Hypothesis, the study investigates how different feedback types interact with learners’ cognitive processing during writing.
Methods:
A quasi-experimental design was employed with 180 Pakistani senior high school students divided into four groups (direct WCF, indirect WCF, metalinguistic WCF, and control). Over five weeks, participants completed picture-based story-writing tasks (150–180 words) in 30-minute sessions. Learners received feedback according to group assignment, and their performance was measured through pre-test, post-test, and delayed post-test writing tasks. Data were analyzed using non-parametric alternatives to mixed-design ANOVA, including the Kruskal–Wallis and Conover–Inman tests.
Results:
All three WCF types significantly improved learners’ past tense accuracy compared with the control group. Metalinguistic feedback produced the strongest immediate gains in the post-test, while indirect feedback led to superior long-term retention in the delayed post-test. Significant differences were observed across feedback types and testing phases, indicating that feedback effectiveness varies depending on learning timeframe.
Discussion:
The findings demonstrate that feedback type influences both the efficiency and durability of grammatical development. Integrating CLT and the Input Hypothesis, the study suggests that WCF functions as cognitively manageable input (i+1) when aligned with learners’ cognitive readiness. Pedagogically, the results highlight the importance of selecting feedback strategies that match learners’ proficiency and cognitive processing capacity, particularly in secondary-level EFL contexts.
1 Introduction
Acquiring grammatical accuracy is a fundamental challenge in second language (L2) learning, particularly for young learners whose language and cognitive abilities are still developing. Among the various linguistic features of English, the simple past tense poses notable difficulties for non-native learners due to its irregular forms and contextual usage. For senior high school students in non-native English-speaking countries like Pakistan, mastering this tense is critical for effective communication, especially in written narrative tasks. However, despite the emphasis on grammar instruction in Pakistan’s educational system, students often struggle to apply these rules in authentic written communication (Hassan et al., 2021).
WCF can be defined as a written response to a linguistic error in an L2 learner’s text, aiming either to provide the correct form or to indicate where and why the error occurred so the learner can self-correct (Al-Kharusi, 2018). WCF is widely acknowledged as a powerful mechanism for supporting L2 writing development (Rasool et al., 2023). However, much of the existing research has primarily focused on adult learners or university students, with limited attention to senior high school learners who may respond differently to instructional interventions due to developmental and cognitive factors (Moradkhani and Mansouri, 2023).
This study addresses this gap by examining the efficacy of three common WCF types direct feedback, indirect feedback, and metalinguistic feedback on the development of the English simple past tense among Pakistani senior high school learners. The study is grounded in CLT and Input Hypothesis, two well-established frameworks in educational psychology and second language development. CLT highlights the importance of reducing unnecessary cognitive stress and managing the limitations of working memory during instruction (Sweller et al., 1998; Sweller, 2010). In parallel, Krashen’s Input Hypothesis, which posits that language acquisition is enhanced through comprehensible content just beyond the learner’s current skill set (i + 1), provides a theoretical lens for understanding how WCF can either support or hinder learning depending on its delivery (Krashen, 1982, 2025). In the context of this study, i + 1 represents the level of linguistic input that challenges but does not overwhelm learners’ existing grammatical competence. WCF acts as a form of enhanced input that draws learners’ attention to gaps between their current interlanguage (i) and the target grammatical feature (the +1). When feedback is appropriately calibrated, it provides learners with comprehensible yet slightly advanced input that promotes noticing and restructuring of linguistic forms. However, when feedback exceeds the learner’s processing capacity due to complexity, lack of prior knowledge, or poorly timed instruction it may create excessive cognitive load, thereby impeding acquisition. This dynamic connection between i + 1 and CLT forms the conceptual foundation of the present research, which examines how different WCF types balance the provision of new input with the management of cognitive demands.
This study’s relevance is amplified by the Pakistani context, where second language instruction is primarily provided in English amidst a multilingual backdrop. Students often memorize grammatical rules but fail to transfer them effectively into communicative tasks. This study contributes to both theoretical understanding and practical instructional design by targeting the simple past tense, a core grammatical structure frequently used in narration and description. Mastery of the past tense enables learners to articulate sequences of events, reflect on personal experiences, and engage with broader academic writing requirements competencies essential for academic success in an English-medium educational system (Coleman and Capstick, 2012).
Furthermore, the type of feedback provided can significantly influence learners’ noticing, processing, and internalization of grammatical forms, which are crucial for long-term retention and transfer (Bitchener and Knoch, 2008). Focused WCF, particularly when aligned with learners’ cognitive capacities, has been suggested to facilitate higher accuracy gains than unfocused feedback (Li and Roshan, 2019). By testing this claim in the Pakistani high school context, the current research not only fills a demographic and regional gap but also provides actionable insights for ESL/EFL educators working with adolescent learners.
1.1 Research rationale
In Pakistan, Urdu serves as the official language, with English as the primary medium of instruction in higher education. English and Urdu are commonly used in schools, with English being more prevalent in private schools and Urdu in public institutions (Coleman and Capstick, 2012). Both languages exhibit a form of modified diglossia, where each is used in distinct formal and informal contexts. While English holds dominance in sectors like civil and military bureaucracy, law, commerce, media, education, and research, it is often taught with a distinct variation from standard British English. This divergence creates a significant challenge for Pakistani learners, as their exposure to English grammar differs from what they are taught in school (Eslami, 2014).
The discrepancies in language usage affect the writing input of Pakistani students, leading them to be resistant to corrective feedback. This resistance stems from the substantial differences between the English language they are accustomed to and the standard English they are taught, which can influence how feedback is perceived and internalized (Bitchener and Knoch, 2008). In the evolving field of language education in Pakistan, much focus has been placed on the need for effective WCF to address the difficulties students face in acquiring grammar, particularly with respect to the English simple past tense. As educators strive to improve teaching methods, it is essential to comprehend the intricacies of WCF to help students overcome challenges in grammatical acquisition at this critical educational level.
Despite widespread agreement on the benefits of WCF, previous research has shown structural flaws in many studies, such as the absence of a control group (Karim and Nassaji, 2019). Prior studies often failed to compare learners who received feedback with those who did not, limiting the validity of their findings (Storch and Wigglesworth, 2010). To address this limitation, this study includes a control group to compare the effectiveness of different WCF types against the absence of any feedback, ensuring more reliable results and drawing more definitive conclusions about the impact of WCF on past tense acquisition.
1.2 Research objectives and questions
The primary objectives of this study endeavor are twofold,
To examine the effect of written corrective feedback (WCF) on Pakistani senior high school learners’ acquisition of the English simple past tense.
To identify which type of written corrective feedback direct (DWCF), indirect (IWCF), or metalinguistic (MWCF) most effectively facilitates learners’ acquisition of the English simple past tense over time.
To achieve the research objectives, this study formulates three research questions,
To what extent does written corrective feedback affect Pakistani senior high school students’ acquisition of the English simple past tense?
Which type of written corrective feedback most effectively supports the acquisition of the English simple past tense among Pakistani senior high school learners?
2 Literature review
WCF has long been a central topic in second language development (SLA), especially in relation to written accuracy and grammatical development. While there has been debate over its overall effectiveness (Truscott, 2004; Li and Vuono, 2019), the majority of empirical studies support its potential for promoting L2 learning (Bitchener and Knoch, 2008; Sheen et al., 2009). The current study aligns with this consensus but moves beyond the general efficacy of WCF by investigating the comparative impact of several WCF formats, including direct, indirect, and metalinguistic feedback, on senior high school students’ learning of the English simple past tense learners in Pakistan. Corrective feedback can take several forms, but the three most commonly examined types in written contexts are DWCF, IWCF and MWCF. Understanding how each type functions is critical for evaluating their cognitive demands and potential instructional value.
DWCF involves explicitly pointing out the mistake and giving the accurate form, leaving little to interpretation. Teachers may cross out, replace, or insert words, phrases, or morphemes to show the correct usage (Ferris, 2012; Bitchener, 2012). This type of feedback is often considered suitable for lower proficiency learners as it reduces ambiguity and minimizes cognitive load, making it easier for learners to process the correction (Rasool et al., 2022). However, critics argue that while direct WCF facilitates short-term improvements, it may promote passive learning, where students rely on teacher input rather than developing self-correction skills. From a CLT perspective, direct feedback might reduce extraneous load, but it can also bypass the germane cognitive processing required for long-term acquisition and transfer.
In DWCF signals that a mistake exists without giving the accurate form. Teachers may underline errors, circle problematic segments, or use codes that require learners to identify and correct their mistakes autonomously (Bitchener, 2012; Ferris, 2012; Nassaji, 2016). This form of feedback aligns well with guided discovery learning and encourages deeper cognitive engagement.
Proponents of indirect WCF argue that it promotes learner autonomy and noticing, a key component in Schmidt’s Noticing and Krashen’s Input Hypothesis (Peng et al., 2023). By fostering problem-solving behavior, learners are more likely to internalize grammatical rules, leading to better retention and accuracy over time (Baleghizadeh and Dadashi, 2011; Hosseiny, 2014; Mujtaba et al., 2020). However, if learners lack sufficient metalinguistic knowledge, the intrinsic cognitive load may overwhelm their working memory, particularly in adolescent learners (Moradkhani and Mansouri, 2023).
Metalinguistic WCF offers students with clear information or grammar rules related to their errors. This can take the form of written comments or error codes indicating the type of mistake (e.g., “VT” for verb tense). Students are then anticipated to use this information to self-correct (Alshahrani and Storch, 2014).
Although more time-consuming for instructors, metalinguistic feedback is often found to be particularly beneficial for learners with developing linguistic competence, as it bridges gaps in cognitive ability and supports deeper understanding (Bitchener, 2012). From a CLT perspective, this feedback type increases germane load, promoting schema construction and transfer of grammatical knowledge. However, it must be cautiously implemented to prevent overloading students with excessive extraneous cognitive load, especially when working with complex structures like past tense forms.
2.1 Mixed findings in WCF research
The literature on WCF effectiveness is not without controversy. While several studies report positive effects of WCF on grammatical development (Sheen et al., 2009; Li and Roshan, 2019), others question its long-term utility or suggest potential negative effects when learners become overly reliant on external correction (Li and Vuono, 2019; Truscott, 2004). These inconsistencies may arise from differences in learner profiles, feedback types, instructional contexts, and cognitive readiness.
Furthermore, the target structure plays a decisive role in moderating the effect of WCF. Research traditionally focuses on English articles or plural markers, leaving structures like the simple past tense underexplored. Given its high error frequency, morphological irregularity, and central role in narrative discourse, the simple past tense presents unique challenges for L2 learners and therefore serves as a highly suitable target for focused feedback studies (Kang and Uchikoshi, 2022). There is growing support for differentiated WCF approaches that consider the student’s cognitive capacity, language proficiency, and the complexity of the target structure. While direct feedback may suit learners who benefit from explicit correction, indirect and metalinguistic feedback promote greater learner involvement and cognitive processing factors critical to durable learning.
This study addresses a notable gap by comparing these three feedback types in the Pakistani high school context, focusing on the simple past tense as the target structure. By integrating Cognitive Load Theory into the analysis, it contributes to a more nuanced understanding of how feedback type, learner cognition, and grammatical complexity interact to affect language development. Although CLT provides the principal framework for interpreting the cognitive mechanisms underlying feedback processing, Krashen’s Input Hypothesis offers an important complementary perspective. WCF can be viewed as an instructional technique that supplies learners with modified input slightly beyond their current level of competence (i + 1). Effective feedback therefore depends on its ability to maintain this delicate balance providing sufficient linguistic challenge to stimulate development while avoiding cognitive overload. In this sense, CLT and the Input Hypothesis are mutually reinforcing: the former explains how much information learners can handle at one time, while the latter clarifies what kind of input promotes interlanguage advancement. Integrating these two theories allows the present study to conceptualize WCF as both a cognitive and linguistic scaffold that optimizes learning conditions for adolescent EFL learners.
3 Research method
This study employed a Quasi-experimental research design to explore the effects of different types of WCF on the acquisition of the English simple past tense among senior high school students in Pakistan. The main aim was to investigate the cognitive load associated with each type of feedback and its impact on students’ past tense acquisition over time. This design entails “pretest and posttest data collection” from the control and experimental groups at the start of the term as a pretest and at the conclusion as a posttest. The three types of WCF selected for this study, DWCF, IWCF and MWCF were chosen based on their potential to vary in cognitive load demands, as outlined by CLT. CLT emphasizes the limited capacity of working memory and suggests that instructional strategies should minimize extraneous cognitive load while promoting germane cognitive load to optimize learning outcomes. The process of the study is shown in Table 1.
Table 1
| Time | Procedure | Tasks |
|---|---|---|
| Week 1 | Pretest | Story writing & errors record |
| Week 2 | Treatment 1 | Story writing treated with allocated feedback type |
| Week 3 | Treatment 2 | Story writing treated with allocated feedback type |
| Week 4 | Treatment 3 | Story writing treated with allocated feedback type |
| Week 5 | Immediate post-test | Story writing & errors record |
| Week 9, day 1 | Delayed post-test | Story writing & errors record |
Procedure of the study.
Although CLT emphasizes minimizing extraneous cognitive demands, this study intentionally included an indirect WCF condition to explore how guided problem-solving may enhance germane cognitive processing. While indirect feedback requires learners to self-correct, it simultaneously encourages active noticing, hypothesis testing, and rule internalization all of which contribute to schema construction and long-term retention. From a CLT perspective, the key was not to eliminate cognitive load entirely but to optimize it by shifting mental effort from extraneous to germane processing. To prevent excessive overload, the indirect feedback tasks were carefully scaffolded: students received clear instructions, practiced with teacher support in initial sessions, and were given sufficient time to review their drafts. This ensured that learners could manage the feedback cognitively while still engaging in deeper metalinguistic reflection.
3.1 Participants
A total of 180 senior high school students (aged 15–16) participated in the study. The participants were drawn from five intact classes at the same school and shared a similar educational background. Their first languages included Urdu, Siraiki, and Punjabi. To ensure baseline equivalence across groups, participants were not grouped according to intact classroom membership. Instead, they were artificially assigned to four groups (DWCF, IWCF and MWCF, and No Feedback/Control) for the purpose of the research.
Group assignment was based on students’ performance on the grammar section of their most recent final English examination, which included items assessing knowledge of the English simple past tense. Students’ scores were first ranked, and participants were then distributed across the four groups using a matched performance approach, ensuring that each group contained a comparable range and mean level of proficiency. This procedure was intended to maintain homogeneity of overall proficiency across groups rather than to create groups with different performance levels. To statistically verify group equivalence, a preliminary non-parametric analysis was conducted. The results confirmed that there were no significant differences in baseline past tense performance among the four groups (χ2 = 0.71, p = 0.54; see Table 2). This indicates that the groups were homogeneous in terms of initial proficiency, allowing subsequent differences to be attributed to the feedback interventions rather than pre-existing performance differences.
Table 2
| Groups | SS | df | MS | F | Sig. |
|---|---|---|---|---|---|
| Between groups | 5.97 | 3 | 1.92 | 0.71 | 0.54 |
| Within groups | 489.91 | 176 | 2.78 | ||
| Total | 495.89 | 179 |
Non-parametric mixed-design ANOVA results for group equivalence.
3.2 Instruments and data collection
Story writing on a photo set is used in this study to answer the research questions.
3.2.1 Story writing on picture sets
In this study, participants were asked to write stories based on a series of picture prompts at three different stages: the pretest, immediate posttest, and delayed posttest. A total of six unique picture sets were used throughout the process (see Appendix A). For each writing session, the same picture prompt was assigned to all students within that session, regardless of feedback group. Students did not choose or select picture sets themselves. Different picture sets were used across the pretest, treatment sessions, immediate post-test, and delayed post-test to avoid task repetition and practice effects. However, within each session, the picture set was identical for all participants, ensuring uniform task input and comparability of written output across groups. For the pretest, each participant wrote a story using one picture set to assess their use of the past simple tense. Over the following 3 weeks, they engaged in treatment sessions where they composed stories using new picture sets and received corrective feedback ranging from direct and indirect comments to metalinguistic explanations. For both the immediate and delayed posttests, participants wrote new stories based on different picture sets. The primary focus during error analysis was their use of the past simple tense. Each story-writing task was conducted under timed classroom conditions, with a duration of 30 min per session. Participants were instructed to produce a story of approximately 150–180 words based on the assigned picture set. All writing sessions were supervised by the same instructor across all groups to ensure consistency in task administration, instructions, and classroom conditions. The instructor did not provide any additional linguistic assistance beyond the assigned feedback treatment, thereby minimizing instructor-related variability. Each story-writing task was conducted under controlled classroom conditions, with a fixed time limit of 30 min and standardized instructions. These conditions were designed to ensure comparability and consistency of learner output across groups and testing phases, while allowing students to produce extended written language based on the picture prompts without external support. This approach enabled the systematic examination of learners’ written production and their use of the past simple tense across pretest, immediate post-test, and delayed post-test conditions. Importantly, no explicit instruction on the English simple past tense was provided between writing sessions. Apart from the assigned written corrective feedback treatments (DWCF, IWCF and MWCF), students did not receive additional grammar explanations, practice exercises, or instructional input targeting past tense forms during the intervention period. This design ensured that any observed changes in learners’ past tense performance could be attributed to the type of written corrective feedback received, rather than to intervening instructional effects.
3.3 Scoring and reliability
In this study, the “obligatory contexts” for past tense usage were assessed subjectively. All writing samples were assessed by a primary rater, and to ensure the reliability of the scoring process, a subset of the data was double-coded by a second rater. The inter-rater reliability was calculated using the Intraclass Correlation Coefficient (ICC), which yielded a value of 0.61. This indicates a good level of agreement between the two raters, confirming the consistency and robustness of the scoring process. This inter-rater reliability metric ensures that the assessment of past tense usage was both valid and consistent, strengthening the quality of the data. For future studies, we will continue to monitor and report such reliability metrics to maintain the integrity of the evaluation process.
A partial credit scoring system was adopted in this study to reflect the gradual nature of second language development. This method is well-supported by both theoretical perspectives and empirical findings and has been effectively applied in previous research on written corrective feedback (Shintani et al., 2014). In line with this approach, students’ use of the past tense was evaluated using a tiered system: partially accurate attempts, such as “goed” instead of “went,” were awarded 0.5 points, while fully correct forms, like “went” instead of “go,” received one full point (Li and Roshan, 2019). Once all writing samples were assessed, each participant’s overall performance was calculated using the following formula:
The scoring formula for past tense usage relies on the identification of “obligatory contexts,” which refer to situations where the use of the past tense is grammatically required. For instance, if a sentence refers to an event that occurred in the past, it must use the past tense verb form (e.g., “She walked to the store yesterday”). To ensure reproducibility and consistency in scoring, the obligatory contexts were defined as follows:
Obligatory contexts: Any situation where the verb tense is required by the context, such as describing past events, hypothetical situations, or actions that have already occurred.
If a student failed to use the past tense when required (e.g., using the present tense in a context that calls for the past), this was counted as an error.
Regarding the avoidance strategy where a student may choose not to use the past tense entirely—the following protocol was followed:
If a student avoided using the past tense verb form entirely (i.e., no past tense verbs were used in the writing sample), their score was not inflated or deflated due to a zero denominator. Instead, such cases were flagged and treated as a special case. The total number of obligatory contexts was calculated based on the number of sentences where past tense was necessary, and students who did not use any past tense verbs received a score of 0 for that sample.
In cases where students used very few verbs overall (low text production), the score was adjusted to avoid any unfair inflation or deflation. We ensured that the number of obligatory contexts was adequately represented in the text (i.e., the prompts ensured a sufficient number of contexts that necessitated past tense usage). If the number of obligatory contexts was too low, the data was flagged, and those samples were excluded from the analysis if they did not meet a minimum threshold.
For example:
If the prompt required describing past actions in a story or a scenario, a minimum of five past-tense verbs was expected to be used by the student. If the student produced fewer than five verbs, the data for that student was excluded from the analysis, ensuring that all data met a reasonable threshold of language production. As a result, the final dataset included 166 writing samples out of the initial 180 participants. The retained samples were distributed as follows: DWCF (n = 42), IWCF (n = 41), MWCF (n = 43), and No Feedback/Control (n = 40). All subsequent statistical analyses and reported results are based on this final dataset rather than the initial participant enrollment. Group-level exclusions were minimal and proportionally comparable across conditions, minimizing the risk of systematic bias due to data screening.
3.4 Data analysis
The data analysis was conducted using SPSS 26.0 to determine the effectiveness of different WCF types over time. Because the same participants were measured at three stages (pretest, immediate posttest, and delayed posttest), the design included a within-subjects factor (time) and a between-subjects factor (feedback type). However, the Kolmogorov–Smirnov tests indicated that the data violated the assumption of normality across all time points (p < 0.05). Given these violations, a non-parametric equivalent of the mixed-design ANOVA was adopted. Instead of treating all time points simultaneously, Kruskal–Wallis tests were applied separately at each stage to compare the performance distributions across the four feedback groups. This approach is consistent with non-parametric handling of repeated measures when normality is violated (Ostertagová et al., 2014). For pairwise group comparisons, post-hoc comparisons were conducted using the Least Significant Difference (LSD) approach applied to ranked data via the Conover–Inman procedure. This combination of tests ensured a robust analysis of between-group differences without violating assumptions. Although a mixed-design ANOVA would have been ideal under normality conditions, non-parametric treatment across repeated measures is widely accepted in applied linguistics and educational psychology research when data distributions deviate significantly from normality.
4 Results and findings by research questions
This section reports the results and findings for each research question based on the data from the writing tasks.
4.1 Descriptive statistics of four groups
Table 3 presents the descriptive statistics for the pretest, immediate post-test, and delayed post-test scores across the four groups (DWCF, IWCF and MWCF, and No Feedback/Control). Given the skewed nature of the data, we report the Medians and Interquartile Ranges (IQR) for each group at each testing point, instead of Means and Standard Deviations, to better represent the central tendency and dispersion in line with the non-parametric analysis used. Medians represent the middle value of the dataset, ensuring that the scores are not unduly influenced by outliers or skewed data. Interquartile Range (IQR) provides the spread of the middle 50% of the data, showing the variability within each group’s scores.
Table 3
| Group | Pretest (Median, IQR) | Immediate post-test (Median, IQR) | Delayed post-test (Median, IQR) |
|---|---|---|---|
| DWCF | 30.0 (25.0–35.0) | 35.0 (30.0–40.0) | 45.0 (40.0–50.0) |
| IWCF | 30.0 (28.0–32.0) | 40.0 (35.0–45.0) | 50.0 (45.0–55.0) |
| MWCF | 32.0 (30.0–34.0) | 45.0 (40.0–50.0) | 55.0 (50.0–60.0) |
| No Feedback (Control) | 28.0 (25.0–30.0) | 35.0 (30.0–40.0) | 40.0 (35.0–45.0) |
Descriptive statistics of the four groups’ performance on pretest, immediate post- test, and delayed post-test.
As can be observed, the DWCF group showed improvement in all phases, with the Medians increasing from the pretest to the delayed post-test. Similarly, the IWCF and MWCF groups also demonstrated growth in their scores. The No Feedback (Control) group showed relatively lower scores across all stages, with minimal improvement in the delayed post-test.
4.2 Findings of research question 1
The first research question explored whether written corrective feedback affects the past tense acquisition of senior high school learners. To address this question, data were collected and analyzed statistically.
Table 4 presents the results of the Kolmogorov–Smirnov test, which was conducted to assess the normality of the score distributions. The significance values for the pretest, immediate post-test, and delayed post-test scores were all recorded as 0.01. According to the commonly accepted threshold outlined by Dörnyei (2010), a dataset is considered normally distributed when the significance value is equal to or greater than 0.05. However, as indicated in Table 4, all significance values fell below this threshold, suggesting that the data did not meet the assumption of normality. Consequently, the use of parametric tests such as ANOVA was deemed inappropriate. Instead, the study employed the Kruskal–Wallis test, a non-parametric alternative suitable for analyzing data that violate normality assumptions. The results of the Kruskal–Wallis test, including the analysis of variance and mean rank comparisons for the pretest, immediate post-test, and delayed post-test scores, are also reported.
Table 4
| Test scores | Statistic | df | Sig. |
|---|---|---|---|
| Pretest | 0.15 | 179 | 0.01 |
| Immediate post-test | 0.11 | 179 | 0.01 |
| Delayed post-test | 0.15 | 179 | 0.01 |
Kolmogorov–Smirov test for normality.
Tables 5, 6 present the results of the Kruskal–Wallis test, with a significance level of Sig. ≥ 0.05. According to Table 5, the pretest scores showed no statistically significant differences among the various WCF groups, with a Chi-square (χ2) value of 1.59 and p = 0.66. The mean ranks were as follows: 84.08 for the direct WCF group, 97.44 for the indirect WCF group, 91.42 for the metalinguistic explanation group, and 89.06 for the No feedback (Control) group.
Table 5
| Test scores | Pretest | Immediate post-test | Delayed post-test |
|---|---|---|---|
| Chi-Square | 1.59 | 129.45 | 106.06 |
| Df | 3 | 3 | 3 |
| Asymp. Sig. | 0.66 | 0.00* | 0.01 |
Kruskal–Wallis analysis of variance.
Table 6
| Assessment | Ranks | ||
|---|---|---|---|
| Groups | N | Mean rank | |
| Pretest | DWCF | 42 | 84.08 |
| IWCF | 41 | 97.44 | |
| MWCF | 43 | 91.42 | |
| No feedback | 40 | 89.06 | |
| Total | 166 | ||
| Immediate post-test | DWCF | 42 | 127.87 |
| IWCF | 41 | 48.57 | |
| MWCF | 43 | 140.89 | |
| No feedback | 40 | 44.68 | |
| Total | 166 | ||
| Delayed post-test | DWCF | 42 | 94.14 |
| IWCF | 41 | 120.12 | |
| MWCF | 43 | 123.53 | |
| No feedback | 40 | 24.20 | |
| Total | 166 | ||
Kruskal–Wallis mean ranks of pretest, immediate post-test, and delayed post-test.
Scripts were excluded if they failed to meet the minimum threshold of obligatory past tense contexts required for valid scoring.
However, the analysis revealed statistically significant differences in the immediate post-test scores between the different WCF groups, with a Chi-square (χ2) of 129.45 and p = 0.00. The mean ranks for the groups were 127.87 for the direct WCF group, 48.57 for the indirect WCF group, 140.89 for the metalinguistic explanation group, and 44.68 for the No feedback group.
Furthermore, significant differences were observed in the delayed post-test scores across the groups, with a Chi-square (χ2) of 106.06 and p = 0.00. The mean ranks for the groups were 94.14 for the DWCF group, 120.12 for the IWCF group, 123.53 for the MWCF group, and 24.20 for the No feedback group.
In addition to reporting p-values for the Kruskal–Wallis test and pairwise comparisons, we also calculated effect sizes to provide a better understanding of the practical significance of the feedback interventions. For the Kruskal–Wallis test, we computed epsilon-squared (η2), and the values were, DWCF: η2 = 0.978, IWCF: η2 = 0.973, MWCF: η2 = 0.975, No Feedback (Control): η2 = 0.890.
For pairwise comparisons, we calculated r for each comparison, DWCF vs. Control: r = 0.34, IWCF vs. Control: r = 0.28, MWCF vs. Control: r = 0.38.
These effect sizes highlight the practical significance of the feedback interventions. The Kruskal–Wallis η2 values suggest that the feedback types accounted for a large proportion of the variance in post-test scores, with DWCF showing the strongest effect. The pairwise comparison r values indicate moderate to small effect sizes between the experimental groups and the control group, providing additional context for the statistical significance observed.
In order to assess the rate of change in writing performance across the groups, we calculated the gain scores (Posttest - Pretest) for each participant. The Kruskal–Wallis test was then applied to the gain scores to test whether the experimental groups improved significantly more than the control group. The results of the Kruskal–Wallis test showed a statistically significant difference in the gain scores between the groups (χ2 = 30.99, p = 8.53 × 10−7). This indicates that the experimental groups (DWCF, IWCF, and MWCF) improved significantly more than the Control group.
It can be observed that the pretest mean ranks showed little difference across the groups. However, for both the immediate and delayed post-test scores, there was a noticeable difference, with the experimental groups (DWCF, IWCF, and MWCF) showing significant improvements in mean scores. In contrast, the control group, which received no feedback, did not show any improvement in their writing scores. Instead, the delayed post-test scores of the control group revealed a slight decline.
The Kruskal–Wallis test confirms these significant results. In relation to the first research question, three types of WCF (DWCF, IWCF, and MWCF) significantly enhanced the development of past tense usage among senior high school learners. In contrast, the control group, which received no feedback, did not demonstrate any improvement in their writing, and their scores even declined according to the delayed post-test.
To further address the first research question, Figure 1 presents a comparison of the mean ranks, clearly showing a difference in mean scores. MWCF and DWCF, followed by IWCF, led to noticeable improvements, while the No feedback group showed no improvement. These findings reveal that different types of WCF contributed to writing score improvements, whereas the control group, lacking feedback, did not improve and instead showed a decline in scores on the delayed post-test.
Figure 1
Performance trajectories of the four groups across pre-test, immediate post-test, and delayed post-test (median scores).
Table 7 presents the pretest performance of all four groups, indicating that there were no statistically significant differences in the mean scores between any of the group pairs.
Table 7
| (i) Participants | (j) Participants | Mean difference (i–j) | Asymp. Sig. |
|---|---|---|---|
| DWCF | IWCF | −1.28 | 0.19 |
| MWCF | −0.65 | 0.50 | |
| No feedback | −0.72 | 0.46 | |
| IWCF | DWCF | 1.28 | 0.19 |
| MWCF | 0.63 | 0.51 | |
| No feedback | 0.56 | 0.56 | |
| MWCF | DWCF | 0.65 | 0.50 |
| IWCF | −0.63 | 0.51 | |
| No feedback | −0.06 | 0.94 | |
| No feedback | DWCF | 0.72 | 0.46 |
| IWCF | −0.56 | 0.56 | |
| MWCF | 0.06 | 0.94 |
Post-hoc comparison test on the performance of the four groups on pretest.
4.3 Findings of research question 2
To further understand the specific areas and extent of the differences among the groups, an additional analysis was conducted to determine which type of WCF most effectively supports learners’ acquisition of the past tense. To achieve this, a pairwise comparison approach was adopted. This method involves testing whether any two given groups denoted as groups i and j (where i, j = 1, 2, …, k) originate from the same distribution or differ significantly. Given that the Kruskal-Wallis test had already identified overall group differences, a post hoc analysis was necessary to pinpoint where those differences lay. For this purpose, the Conover-Inman procedure was applied, which is a widely accepted follow-up test to the Kruskal–Wallis analysis. This technique operates by applying the Least Significant Difference (LSD) method to ranked data (Conover, 1999). Two groups are considered significantly different if the absolute difference between their mean ranks exceeds a calculated critical value, determined by the chosen level of significance (α). This approach enables a more precise identification of which feedback types lead to statistically meaningful improvements in learners’ grammatical performance.
Table 8 presents the results of multiple comparisons among the four groups on the immediate post-test, showing a statistically significant difference in their performance between each pair of the various groups except between IWCF and no feedback. This indicates that the effect of MWCF and DWCF type has no significant difference in learners’ past tense acquisition.
Table 8
| (i) Participants | (j) Participants | Mean difference (i–j) | Asymp. Sig. |
|---|---|---|---|
| DWCF | IWCF | 20.04 | 0.01 |
| MWCF | −2.85 | 0.05 | |
| No feedback | 21.23 | 0.01 | |
| IWCF | DWCF | −20.04 | 0.01 |
| MWCF | −22.90 | 0.01 | |
| No feedback | 1.18 | 0.41 | |
| MWCF | DWCF | 2.85 | 0.05 |
| IWCF | 22.90 | 0.01 | |
| No feedback | 24.08 | 0.01 | |
| No feedback | DWCF | −21.23 | 0.01 |
| IWCF | −1.18 | 0.41 | |
| MWCF | −24.08 | 0.01 |
Post-hoc comparison of the performance of the four groups on immediate post- test.
The Post-hoc comparison demonstrated that students who were given metalinguistic explanations outperformed the other three groups. Their performance was followed by those who received DWCF, then by those who received indirect feedback. The control group, which did not receive any feedback on their writing, showed the lowest performance. According to immediate post-test scores, IWCF and No feedback (control group) showed least improvement in past tense acquisition. It can be concluded that the participants did not benefit from IWCF, just like no feedback.
Table 9 presents the results of multiple comparisons of the performance of the four groups on the delayed post-test. The data indicates a statistically significant difference between each pair of groups, except for the comparison between IWCF and metalinguistic explanation. The findings suggest that, when assessed on the delayed post-test, the effects of IWCF and MWCF are similarly effective in improving students’ development of the past simple tense among senior high school learners.
Table 9
| (i) participants | (j) participants | Mean difference (i–j) | Asymp. Sig. |
|---|---|---|---|
| DWCF | IWCF | −4.1489 | 0.005 |
| MWCF | −4.7400 | 0.001 | |
| No feedback | 19.5422 | 0.001 | |
| IWCF | DWCF | 4.1489 | 0.005 |
| MWCF | −0.5911 | 0.683 | |
| No feedback | 23.6911 | 0.001 | |
| MWCF | DWCF | 4.7400 | 0.001 |
| IWCF | 0.5911 | 0.683 | |
| No feedback | 24.2822 | 0.001 | |
| No feedback | DWCF | −19.5422 | 0.001 |
| IWCF | −23.6911 | 0.001 | |
| MWCF | −24.2822 | 0.001 |
Post-hoc comparison of the performance of the Four Groups on Delayed Post-test.
The post-hoc analysis showed that students who received MWCF or IWCF performed better than those who received direct feedback, as well as those in the control group who received no feedback on their writing. Looking at the immediate post-test scores, metalinguistic explanation was identified as the most effective form of WCF for improving students’ acquisition of the past simple tense, followed by DWCF, then IWCF, with the control group showing the least improvement. In the delayed post-test, the combination of MWCF and IWCF proved to be the most effective in enhancing students’ understanding of the past simple tense, followed by direct feedback, with the control group again being the least effective. These results highlight a shift in the ranking of feedback effectiveness compared to earlier findings.
Regarding the second research question, MWCF emerged as the most effective form of WCF in improving students’ writing performance, followed by DWCF and IWCF. However, the delayed post-test analysis revealed that both IWCF and MWCF helped students acquire the past simple tense, although the degree of their impact varied. It can be noticed that delayed post-test scores of the IWCF group showed significant improvement as compared to immediate post-test scores, which suggests the participants became more independent for long-term learning than DWCF.
5 Discussion
The study adopted a non-parametric alternative to the mixed-design ANOVA due to the non-normal distribution of scores across time points. While some studies have employed repeated-measures ANOVA to analyze similar datasets, the use of Kruskal–Wallis and Conover–Inman post-hoc tests is statistically appropriate for ordinal or non-normally distributed data and has been widely used in SLA research (Ostertagová et al., 2014). The results should therefore be interpreted as rank-based comparisons rather than parametric mean differences, preserving analytical validity while avoiding assumption violations.
5.1 The effectiveness of WCF on targeted structure
This study primarily aimed to examine the impact of WCF on senior high school students’ learning of the English simple past tense. The results indicated that students in all three experimental groups who received WCF showed significant improvements in their post-test scores compared to their pre-test performance. While no notable differences were observed in pre-test scores (χ2 = 1.591, p = 0.661), a substantial effect was found during the post-test phase (χ2 = 129.45, p < 0.001), emphasizing the effectiveness of the feedback intervention. These findings align with previous research, which suggests that systematic and pedagogically consistent WCF can enhance learners’ understanding of grammatical structures, such as the past tense (e.g., Bitchener, 2012; Mao et al., 2024). The results support the idea that feedback on written language production positively contributes to grammatical development, particularly in acquiring and accurately using past tense forms.
However, the degree of effectiveness appears to vary depending on the type of feedback given. Research by Ferris (2012), Niu and You (2020), and Patra et al. (2022) has shown that while WCF generally benefits second language development, its impact depends on the type of feedback, learner characteristics, and the instructional context. These findings highlight the importance of customizing feedback to suit learners’ cognitive and linguistic needs to maximize learning outcomes.
The study’s results reinforce prior research that indicates strategic and consistent use of WCF leads to significant improvements in past tense acquisition, offering learners an effective tool to enhance their language proficiency and expression (Wondim et al., 2024). These results can be understood in the context of CLT, which posits that the human cognitive system has a limited capacity to process information effectively at any given moment (Sweller, 1988). The CLT is highly applicable in the context of language learning and English simple past tense development. According to the results, WCF helped students manage the cognitive load by providing focused and precise information, which in turn reduces the strain on their working memory. Upon receiving WCF, students were furnished with targeted correction feedback about errors in the use of past tense in their work. This detailed and explicit feedback enabled learners to focus on a single area of their language production, thus preventing cognitive overload. The prompt and unambiguous feedback explicitly targeted mistakes specifically associated with the English past tense, by the CLT focus on effectively handling the intricacy of learning tasks.
For instance, in a pretest writing task, excerpt 5 from a participant,
[My father eat a lot yesterday].
After getting a MWCF from the teacher through the treatment process, the participant’s understanding of the second form of the verbs was improved, such as,
Excerpt 5.1 [I ate ice cream with my grandfather].
Excerpt 5.2 [we enjoyed a lot].
This study specifically targeted a single grammatical structure, “past tense,” as the target structure. For several reasons, students frequently make blunders or errors in the simple past tense, such as L1 interference, overgeneralization, fossilization, and cognitive load. Many researchers have attempted to explore the most frequent errors learners make while writing. Bitchener et al. (2005) has studied the most frequent errors and revealed that three linguistic categories are prevalent in students’ writing. The research revealed that the students found past tense errors (11.96% of all errors) as the second most challenging category. Some common reasons contributing to these errors are worth consideration. Firstly, the past tense in English might be structured and used differently from the student’s original tongue. The student’s native language (L1) may have different rules and structures for forming past tenses. This can lead to interference when applying these rules in the target language.
Secondly, students who only occasionally interact with English outside the classroom may need to be more familiar with properly using past tense forms. Thirdly, students occasionally may oversimplify a rule they have learned. For instance, they could wrongly terminate irregular verbs with ordinary past tenses. Irregular verbs in English have unusual past tenses that do not conform to standard patterns.
Excerpt 7 [I sleeped late last night].
Excerpt 8 [we goed to the zoo].
Excerpt 9 [my sister cutted her hand].
The use of “sleeped” instead of “slept,” “goed” instead of “went” and “cutted” instead of “cut” showed a lack of understanding of past tense irregular forms. These examples displayed that the participants faced problems with irregular verb forms as a major issue.
The study emphasized empowering learners to participate actively in their past- tense development. Educators can facilitate this by encouraging self-correction, reflection on feedback, and independent practice. The study’s findings offered proof of the effectiveness of various WCF approaches in correcting past tense errors, illuminating the best methods for boosting grammatical skills. The observed variations in learners’ progress across feedback types can be explained by how each form of WCF contributes to grammar learning within the framework of CLT. MWCF, which provides explicit grammatical explanations, promotes rule awareness and conscious understanding of verb inflection patterns essential for mastering the irregular and regular forms of the English past tense. This type of feedback fosters explicit knowledge while maintaining cognitive manageability through guided explanation, aligning with the concept of germane cognitive load. In contrast, DWCF supplies the correct form immediately, minimizing extraneous cognitive load by reducing the mental effort required to locate and apply rules. This supports accuracy-focused learning among students with limited grammatical competence. IWCF, which only signals the presence of an error, requires learners to engage in self-correction and hypothesis testing, stimulating deeper processing and noticing mechanisms central to long-term language development. While this approach imposes higher cognitive demands, it strengthens learners’ ability to internalize grammatical rules over time. Thus, each WCF type contributes differently to past tense acquisition: MWCF enhances explicit rule learning, direct feedback improves immediate accuracy, and indirect feedback promotes durable grammatical development through self-regulated learning.
5.2 The efficacy of WCF types on immediate posttest
The second objective of this study was to determine which type of written corrective feedback (WCF) is most effective in enhancing senior high school students’ development of the past simple tense. The findings indicated that metalinguistic explanation was the most effective, with students in this group achieving the highest mean score (M = 76.74, SD = 5.64) in the post-test.
This finding supported the research by Fotos (1994), who also found the MWCF compelling. The immediate post-test results showed that the group receiving metalinguistic explanations achieved the most substantial improvement in scores, followed by the DWCF group, which recorded a mean score of 73.89 (SD = 7.05).
Here is an example of a teacher providing a MWCF of the past tense verb error of the participant:
Jumped shouted
Excerpt 11 [I jump on the sofa, and my mother shout at me yesterday].
Use “ed” after the verb to describe an action that has happened already.
Here is another example of a teacher providing a metalinguistic explanation of the past tense irregular verb error of the participant:
fought broke
Excerpt 11 [Last day my sister fight with me because I break her doll].
Some verb forms do not end on “ed” and do not follow the usual rules of grammar such as buy (bought), catch(caught).
Followed by the effective results of metalinguistic feedback on learners’ writing proficiency, the direct feedback was beneficial one. The following is an example of a teacher providing DWCF on the past tense verb error of the participant:
played
Excerpt 12 [I play with my brother yesterday].
Here is an example of a teacher providing DWCF on the past tense irregular verb error of the participant:
broke
Excerpt 15 [I break the window of the police car].
Within the framework of senior high school students’ development of the past simple tense, incorporating written corrective feedback and applying metalinguistic explication emerged as the most effective strategies. In addition to addressing specific errors, these methods fostered a profound understanding of tense structures.
CLT is extensively used in educational research to explain why certain teaching methods may not be effective. The central idea of this theory is that some instructional strategies create unnecessary mental demand on students, overloading their working memory. CLT offers a framework for understanding the cognitive processes involved in learning, emphasizing how learners absorb, store, and retrieve information. A key component of this theory is the concept of working memory, which has a finite capacity to temporarily hold and process information during learning. Because working memory can only handle a limited amount of information simultaneously, instructional designs should focus on reducing cognitive overload. CLT distinguishes among three types of cognitive load: intrinsic, extraneous, and germane. Each type interacts with working memory in different ways.
Intrinsic load is linked to the complexity of the material itself. The more difficult the content, the more working memory is required to process it. Extraneous load arises from the way information is presented. Poorly structured or cluttered instructional materials can introduce unnecessary processing demands, reducing the capacity available for meaningful learning. Germane load refers to the cognitive effort learners invest in understanding and integrating new knowledge. Unlike extraneous load, germane load supports learning and contributes to the formation of long-term memory. CLT emphasizes that for learning to be effective, the overall cognitive load must be carefully managed so that it does not exceed the capacity of learner’s working memory. Overloading working memory can result in cognitive overload, which impairs the student’s ability to absorb and retain information (Wu, 2025).
To address this, CLT offers several instructional design principles. These include presenting information in a clear and organized manner, eliminating irrelevant content, and incorporating support mechanisms such as scaffolding or step-by-step guidance. By doing so, educators can enhance learners’ ability to process information in working memory and facilitate its transfer to long-term memory, where it becomes available for future use.
5.3 The efficacy of WCF types on delayed post-test
The delayed post-test results revealed a shift in performance, with the IWCF group showing a similar improvement trend to that of the metalinguistic explanation group. In fact, the IWCF group (M = 80.11, SD = 5.65) outperformed the direct WCF group in the delayed post-test. These findings underscore the notable effectiveness of in DWCF in enhancing past tense acquisition among senior high school students, particularly over time. These results are in the line with some previous research works as (Ghandi and Maghsoudi (2014), Baleghizadeh and Dadashi (2011), Hosseiny (2014), Mujtaba et al. (2020).
Indirect or Implicit feedback also has the benefit of pushing learners to produce the correct form if they can (Pawlak, 2014). This is consistent with the Input Hypothesis (Swain and Lapkin, 2002), which states that input must be added to input exposure for learning to occur. Here, the role of teachers, who should offer support as needed and then step back by employing the scaffolding approach, is vital. The learners’ language development is facilitated when they engage in meaningful input and receive feedback on their production (Febrian and Nurcholis, 2025). Based on the input function, Kang (2020) claimed that “it is likely that learners notice their linguistic limitations as they engage in writing and that such noticing prepares them to use model texts as a resource to resolve their problems.” Many effects can be observed when considering the effects of indirect WCF on the input hypothesis and the scaffolding approach.
Firstly, IWCF prompted senior high school learners to become more aware of past tense errors in their writing. By indirectly signaling errors without explicit corrections, learners are encouraged to analyze their writing more critically and notice areas that may require improvement, such as the wrong use of irregular verbs (Almanea, 2025). This heightened error awareness aligns with the scaffolding approach, where learners were supported in identifying past tense errors and making appropriate revisions.
Secondly, IWCF promoted self-reflection among senior high school learners. By subtly indicating errors, learners were encouraged to reflect on their writing and consider potential revisions independently. This self-reflective process fostered metacognitive skills, allowing learners to evaluate their language input and make informed decisions about improving their past tense development.
Thirdly, IWCF provided scaffolding support for senior high school learners. Rather than providing explicit corrections, indirect feedback acted as a supportive tool that encouraged learners to participate actively in error detection and correction. This scaffolding approach aligned with the notion that learners benefit from guided assistance as they gradually develop their language skills and avoid past tense errors.
Fourthly, IWCF facilitated incremental learning and language development. By engaging in meaningful input and receiving subtle feedback, learners had the opportunity to refine their past tense acquisition gradually. This incremental learning approach promoted a growth mindset, as learners understood that improvement occurs through ongoing practice, reflection, and feedback.
Lastly, IWCF encouraged senior high school learners to compose writing tasks. By providing feedback that subtly indicates the presence of past tense errors, learners were prompted to engage in the writing process and generate their language production. This input-oriented approach aligns with the input hypothesis, emphasizing the importance of learners actively using the past tense for errorless writing. The effectiveness of IWCF concerning the Input Hypothesis and scaffolding approach varied depending on learners’ past tense understanding, their ability to interpret and apply indirect feedback, and the quality of the feedback provided (Lalande, 1982).
6 Implications of the study
The findings of this study carry significant theoretical, pedagogical, and methodological implications for second language acquisition research and classroom practice.
6.1 Theoretical implications
From a theoretical standpoint, this study extends the application of CLT and Krashen’s Input Hypothesis to the domain of WCF. The results demonstrate that the effectiveness of feedback depends on its ability to balance cognitive manageability (as emphasized in CLT) with linguistic advancement (as conceptualized by i + 1). When feedback provides comprehensible yet slightly challenging input, learners engage in meaningful noticing and restructuring of their interlanguage without experiencing cognitive overload. This integration offers a more nuanced theoretical model for understanding WCF as both a cognitive scaffold and a linguistic input enhancer. It highlights that effective feedback should not only provide linguistic correction but also optimize learners’ cognitive processing capacity ensuring that new input remains within the learner’s zone of proximal development and fosters durable grammatical development.
6.2 Pedagogical implications
Pedagogically, the findings underscore the importance of tailoring feedback strategies to learners’ cognitive readiness and linguistic proficiency. Direct feedback appears particularly beneficial for beginners as it minimizes extraneous load, whereas metalinguistic and indirect feedback encourage deeper engagement and long-term retention through active problem-solving. Teachers should therefore adopt a differentiated feedback approach, aligning the type of WCF with students’ developmental level, task complexity, and learning goals.
Moreover, feedback should be designed to provide i + 1-level linguistic input sufficiently challenging to promote progress, yet not so demanding as to overwhelm working memory. This has direct implications for curriculum designers and teacher training programs, emphasizing the need to integrate principles of CLT and Input Hypothesis into practical WCF pedagogy to enhance grammatical accuracy and learner autonomy.
6.3 Methodological implications
Methodologically, this study contributes to the ongoing refinement of research designs in applied linguistics. By employing a robust quasi-experimental design with clearly defined research questions, random group assignment, and repeated measures, the study strengthens both internal and external validity. The incorporation of non-parametric alternatives to mixed-design ANOVA ensures analytical rigor under non-normal data conditions, providing a replicable model for future SLA research. Furthermore, the preliminary pilot testing and use of a partial credit scoring system enhanced the reliability and sensitivity of the data analysis. Future studies may extend this framework by incorporating additional linguistic structures, longer treatment durations, and qualitative data (e.g., learner reflections or cognitive load self-reports) to better capture how WCF interacts with learners’ cognitive and affective processes over time.
7 Limitations
This study has several limitations. It focused on low-intermediate senior high school learners, which may restrict the generalizability of results to other age groups or proficiency levels. Future studies could include more diverse participants to assess the broader applicability of WCF. The research also examined only the English simple past tense, while language learning involves multiple grammatical areas. Exploring other features in future work would provide a more comprehensive understanding. Lastly, the short study duration limits insight into long-term effects. A longitudinal design could better assess the sustained impact of WCF on grammar development and writing accuracy.
8 Conclusion
This study explored the impact of WCF on the development of the English simple past tense among senior high school students in Pakistan, using CLT as a guiding framework. The findings demonstrate that all forms of DWCF, IWCF, and MWCF, significantly improved students’ past tense acquisition, with metalinguistic explanations emerging as the most effective strategy for long-term learning. The results underscore the importance of aligning feedback types with learners’ cognitive capacities to reduce extraneous cognitive load and enhance germane processing, which is crucial for durable learning.
From a practical standpoint, this study provides valuable insights for ESL/EFL educators, suggesting that feedback should be tailored to students’ cognitive development. DWCF may be beneficial for beginners, while MWCF and IWCF strategies offer deeper cognitive engagement and long-term retention. Educators should aim to strike a balance between providing explicit corrections and encouraging self-reflection and error analysis, which are vital for fostering learner autonomy and deeper understanding.
In terms of educational psychology, this research contributes to the application of CLT in second language development, particularly in writing. By demonstrating how different WCF types interact with cognitive load, this study advances our understanding of how instructional strategies can be designed to optimize cognitive resources and facilitate language development. It provides a theoretical extension of CLT, offering a clearer framework for understanding how feedback can enhance both short-term accuracy and long-term retention in L2 writing tasks.
Ultimately, this research highlights the critical role of cognitive load in shaping the effectiveness of written corrective feedback. Future studies should explore how cognitive load considerations can be integrated into broader instructional practices, addressing various linguistic structures and learner demographics for more comprehensive insights into language development.
Statements
Data availability statement
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material.
Ethics statement
The studies involving humans were approved by Emerson University Multan. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants’ legal guardians/next of kin.
Author contributions
UR: Formal analysis, Investigation, Methodology, Writing – original draft. JQ: Conceptualization, Supervision, Writing – review & editing. HC: Data curation, Resources, Software, Writing – original draft.
Funding
The author(s) declared that financial support was not received for this work and/or its publication.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/feduc.2026.1695913/full#supplementary-material
References
1
Al-KharusiF. M. N. (2018). The practice of teachers’ written corrective feedback as perceived by EFL teachers and supervisors. Oman: Sultan Qaboos University.
2
AlmaneaM. (2025). Less is NOT more for learners: EFL learners' preferences and perceptions of teachers' written corrective feedback. Acta Psychol.255:104926. doi: 10.1016/j.actpsy.2025.104926,
3
AlshahraniA.StorchN. (2014). Investigating teachers’ written corrective feedback practices in a saudi EFL context: how do they align with their beliefs, institutional guidelines, and students’ preferences?Aust. Rev. Appl. Linguist.37, 101–122. doi: 10.1075/aral.37.2.02als
4
BaleghizadehS.DadashiM. (2011). The effect of direct and IWCF on students’ spelling errors. Profile Issues Teach. Prof. Dev.13, 129–137.
5
BitchenerJ. (2012). A reflection on “the language learning potential” of written CF. J. Second. Lang. Writ.21, 348–363. doi: 10.1016/j.jslw.2012.09.006
6
BitchenerJ.KnochU. (2008). The value of written corrective feedback for migrant and international students. Lang. Teach. Res.12, 409–431. doi: 10.1177/1362168808089924
7
BitchenerJ.YoungS.CameronD. (2005). The effect of different types of corrective feedback on ESL student writing. J. Second. Lang. Writ.14, 191–205. doi: 10.1016/j.jslw.2005.08.001
8
ColemanH.CapstickA. (2012). Language in education in Pakistan: recommendations for policy and practice. Islamabad: British Council.
9
ConoverW. J. (1999). Practical nonparametric statistics, vol. 350. Hoboken, NJ: John Wiley & sons.
10
DörnyeiZ.2010 “Researching motivation: from integrativeness to the ideal L2 self,” In HunstonS.OakeyD. (eds) Introducing applied linguistics: concepts and skills374–83London: Routledge
11
EslamiE. (2014). The effects of direct and IWCF Techniques on EFL students’ writing. Procedia – Soc. Behav. Sci.98, 445–452. doi: 10.1016/j.sbspro.2014.03.438
12
FebrianR. A.NurcholisI. A. (2025). The efficacy of written corrective feedback in improving English foreign language university students ‘writing. Jurnal Review Pendidikan dan Pengajaran8, 2931–2939. doi: 10.31004/jrpp.v8i1.43109
13
FerrisD. (2012). “Does error feedback help student writers? New evidence on the short- and long-term effects of written error correction” in Feedback in second language writing (Cambridge: Cambridge University Press), 81–104.
14
FotosS. S. (1994). Integrating grammar instruction and communicative language use through grammar consciousness-raising tasks. TESOL Q.28, 323–351. doi: 10.2307/3587436
15
GhandiM.MaghsoudiM. (2014). The effect of direct and IWCF on Iranian EFL learners’ spelling errors. Engl. Lang. Teach.7, 53–61. doi: 10.5539/elt.v7n8p53
16
HassanM. U.MalghaniM.ZaiK. A. (2021). Exploring undergraduate EFL learners’ beliefs about WCF and their impact on improving writing skills in Pakistani universities. Humanit. Soc. Sci. Rev.9, 1430–1436. doi: 10.18510/hssr.2021.93142
17
HosseinyM. (2014). The role of direct and indirect written corrective feedback in improving Iranian EFL students’ writing skill. Procedia. Soc. Behav. Sci.98, 668–674. doi: 10.1016/j.sbspro.2014.03.466
18
KangE. Y. (2020). Using model texts as a form of feedback in L2 writing. System89:102196. doi: 10.1016/j.system.2019.102196
19
KangH. S.UchikoshiY. (2022). Child second language development of English tense and aspect: the role of narrative organization. Appl. Psycholinguist.43, 785–804. doi: 10.1017/s0142716422000145
20
KarimK.NassajiH. (2019). The effects of written corrective feedback. Instructed Second Language Acquisition3, 28–52. doi: 10.1558/isla.37949
21
KrashenS. (1982). Principles and practice in second language acquisition.
22
KrashenS. (2025). Reading and writing, they do different things. TESOL Bangladesh J.2, 1–8. doi: 10.69907/tbj.v2i1.98
23
LalandeJ. F. (1982). Reducing composition errors: an experiment. Mod. Lang. J.66, 140–149. doi: 10.1111/j.1540-4781.1982.tb06973.x
24
LiS.RoshanS. (2019). The associations between working memory and the effects of four different types of written corrective feedback. J. Second. Lang. Writ.45, 1–15. doi: 10.1016/j.jslw.2019.03.003
25
LiS.VuonoA. (2019). Twenty-five years of research on oral and written corrective feedback in system. System84, 93–109. doi: 10.1016/j.system.2019.05.006
26
MaoZ.LeeI.LiS. (2024). Written corrective feedback in second language writing: a synthesis of naturalistic classroom studies. Lang. Teach.57, 449–477. doi: 10.1017/S0261444823000393
27
MoradkhaniS.MansouriB. (2023). Impact of a research-informed intervention on L2 teachers’ WCF cognitions and practices: a sociocultural perspective. System115:103052. doi: 10.1016/j.system.2023.103052
28
MujtabaS. M.ParkashR.NawazM. W. (2020). Do indirect coded corrective feedback and teachers short affective comments improve the writing performance and learners uptake?Read. Writ. Q.36, 34–47. doi: 10.1080/10573569.2019.1616638
29
NassajiH. (2016). Anniversary article interactional feedback in second language teaching and learning: a synthesis and analysis of current research. Lang. Teach. Res.20, 535–562. doi: 10.1177/1362168816644940
30
NiuR.YouX. (2020). Effects of IWCF with and without written language on L2 written accuracy: a multitask intervention study. Asia Pac. Educ. Res.29, 343–351. doi: 10.1007/s40299-019-00488-8
31
OstertagováE.OstertagO.KováčJ. (2014). Methodology and application of the Kruskal-Wallis test. Appl. Mech. Mater.611, 115–120. doi: 10.4028/www.scientific.net/AMM.611.115
32
PatraI.AlazemiA.Al-JamalD.GheisariA. (2022). The effectiveness of teachers’ written and verbal corrective feedback (CF) during formative assessment (FA) on male language learners’ academic anxiety (AA), academic performance (AP), and attitude toward learning (ATL). Lang. Testing in Asia12:19. doi: 10.1186/s40468-022-00169-2
33
PawlakM. (2014). “The role of written corrective feedback in promoting language development: an overview” in Language learning, discourse and communication. eds. Szubko-SitarekW.SalskiŁ.StalmaszczykP., vol. 19 (Springer), 3–21.
34
PengC. X.StorchN.KnochU. (2023). Greater coverage vs. deeper processing? Comparing individual and collaborative processing of teacher feedback. Lang. Teach. Res. doi: 10.1177/13621688231214910
35
RasoolU.MahmoodR.Zammad AslamM.BarzaniS. H. H.QianJ. (2023). Perceptions and preferences of senior high school students about written corrective feedback in Pakistan. SAGE Open13:21582440231187612. doi: 10.1177/21582440231187612
36
RasoolU.QianJ.SaqlainM.AbbasiB. N. (2022). Written corrective feedback strategies: a systematic review. Voyage J. Educ. Stud.2, 67–83. doi: 10.58622/vjes.v2i2.19
37
SheenY.WrightD.MoldawaA. (2009). Differential effects of focused and unfocused written correction on the accurate use of grammatical for ms by adult ESL learners. System37, 556–569. doi: 10.1016/j.system.2009.09.002
38
ShintaniN.EllisR.SuzukiW. (2014). Effects of written feedback and revision on learners’ accuracy in using two English grammatical structures. Lang. Learn.64, 103–132. doi: 10.1111/lang.12029
39
StorchN.WigglesworthG. (2010). Learner’s processing, uptake, and retention of corrective feedback on writing: case studies. Stud. Second. Lang. Acquis.32, 303–334. doi: 10.1017/S0272263109990532
40
SwainM.LapkinS. (2002). Talking it through: two French immersion learners’ response to reformulation. Int. J. Educ. Res.37:285. doi: 10.1016/S0883-0355(03)00006-5
41
SwellerJ. (1988). Cognitive load during problem solving: effects on learning. Cogn. Sci.12, 257–285. doi: 10.1207/s15516709cog1202
42
SwellerJ. (2010). Element interactivity and intrinsic, extraneous, and germane cognitive load. Educ. Psychol. Rev.22, 123–138. doi: 10.1007/s10648-010-9128-5
43
SwellerJ.Van MerrienboerJ. J. G.PaasF. (1998). Cognitive architecture and instructional design. Educ. Psychol. Rev.10, 251–296. doi: 10.1023/a:1022193728205
44
TruscottJ. (2004). Evidence and conjecture on the effects of correction: a response to Chandler. J. Second. Lang. Writ.13, 337–343. doi: 10.1016/j.jslw.2004.05.002
45
WondimB. M.BishawK. S.ZelekeY. T. (2024). Effectiveness of teachers’ direct and indirect written corrective feedback provision strategies on enhancing students’ writing achievement: Ethiopian university entrants in focus. Heliyon10:e24279. doi: 10.1016/j.heliyon.2024.e24279
46
WuY. (2025). Effects of written corrective feedback on Chinese EFL learners’ cognitive processes. J. Linguist. Commun. Stud.4, 1–8. doi: 10.56397/jlcs.2025.02.01
Summary
Keywords
Cognitive Load Theory, past tense acquisition, second language acquisition, senior high school learners, written corrective feedback
Citation
Rasool U, Qian J and Chen H (2026) Educational psychology perspectives on cognitive load and written corrective feedback: enhancing past tense acquisition in L2 learning. Front. Educ. 11:1695913. doi: 10.3389/feduc.2026.1695913
Received
05 September 2025
Revised
04 February 2026
Accepted
05 February 2026
Published
20 February 2026
Volume
11 - 2026
Edited by
Ana Remesal, University of Barcelona, Spain
Reviewed by
Soumita Sarkar, Gujarat University, India
Abdulaziz Alshahrani, Najran University, Saudi Arabia
Updates
Copyright
© 2026 Rasool, Qian and Chen.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Ushba Rasool, ushba.rasool@gmail.com
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.