Narrative point of view as an instructional variable in English reading classrooms: Differential patterns of multidimensional reading performance

1 Introduction

Stories play a central role in English language teaching (ELT), particularly at the primary level, where narrative texts support reading comprehension, vocabulary development, inferential reasoning, and interpretive engagement (Grabe, 2009; Koda, 2005). Although ELT research has devoted substantial attention to vocabulary, grammar, and text difficulty, less attention has been given to how narrative perspective may shape learners' engagement with texts during reading. One such underexplored feature is narrative point of view (POV). For learners moving beyond literal comprehension toward inferential and interpretive reading, the way a story is narrated may influence both comprehension outcomes and meaning construction.

During the late primary years, reading increasingly involves inferential reasoning, perspective-taking, and the integration of information across multiple levels of meaning (Cain and Oakhill, 2007; Kintsch, 1998). Comprehension therefore depends not only on vocabulary knowledge and decoding ability but also on the ability to interpret implicit meanings, establish coherence, and construct broader representations of narrative understanding. Consequently, the organization and presentation of narrative information may influence engagement with different dimensions of comprehension.

First-person (FP) narration presents events through a character's subjective perspective and provides access to a limited range of narrative information. Readers may therefore rely more heavily on local textual cues and inferential interpretation when constructing meaning. Third-person omniscient (TP) narration, by contrast, provides broader access to events, relationships, and multiple perspectives, potentially supporting coherence building and information integration. These differences suggest that narrative POV may shape engagement during reading.

Reading comprehension is widely regarded as a multidimensional construct involving inferential reasoning, coherence building, local interpretation, and global integration of textual information (Bernhardt, 2005; Grabe, 2009; Kintsch, 1998). In the present study, Micro-Level Comprehension refers to understanding localized narrative information, whereas Macro-Level Comprehension refers to broader thematic, causal, and organizational understanding across the narrative. From this perspective, narrative POV may be associated with differing patterns of performance across distinct dimensions of comprehension rather than influencing comprehension uniformly. Examining narrative perspective within a multidimensional framework may therefore provide a more nuanced understanding of how textual design features relate to learner engagement and reading performance in ELT classrooms.

Despite its theoretical relevance, little classroom-based research has examined narrative POV as a pedagogically meaningful textual feature in ELT contexts. Although narrative perspective has received considerable attention in narratology and reading-comprehension research, less is known about how narration types relate to multidimensional comprehension performance among young English language learners. In practice, reading materials are often selected according to thematic relevance, vocabulary level, or linguistic difficulty, whereas narration type receives comparatively limited consideration.

Viewing narrative POV as a textual design feature rather than merely a literary characteristic may offer a more nuanced perspective on reading pedagogy, as different narrative perspectives may be associated with distinct patterns of reading performance across comprehension dimensions.

The present study investigates how FP and TP narration are associated with multidimensional comprehension performance among Grade 4 and Grade 5 learners in English reading classrooms. Specifically, it examines Inferential Processing, Structural Coherence Processing, Micro-Level Comprehension, and Macro-Level Comprehension. By situating narrative POV within a multidimensional framework, the study seeks to contribute to reading comprehension research and ELT pedagogy by examining how narrative design features may relate to comprehension engagement and performance in classroom contexts.

2 Literature review

Understanding how narrative POV relates to reading comprehension requires integrating insights from narratology, reading-comprehension research, and ELT. Although POV has been extensively examined as a literary construct, its relationship to multidimensional comprehension in classroom contexts remains comparatively underexplored.

This review synthesizes theoretical and empirical scholarship on narrative perspective and reading comprehension, focusing on its potential associations with inferential reasoning, structural coherence, and multidimensional comprehension outcomes in primary-level ELT settings.

Section 2.1 outlines theoretical perspectives on narrative POV and reading comprehension. Sections 2.2 and 2.3 examine first-person (FP) and third-person omniscient (TP) narration in relation to inferential engagement and structural coherence. Section 2.4 reviews multidimensional approaches to reading comprehension, and Section 2.5 identifies the research gap and study rationale.

2.1 Narrative point of view and reading comprehension

Narrative point of view (POV) is a central feature of storytelling that shapes readers' access to, interpretation of, and organization of narrative information. Narratological scholarship conceptualizes perspective through focalisation, narrative voice, and discourse orientation, highlighting how narration influences access to information and interpretive positioning within a text (Chatman, 1978; Genette, 1980; Fludernik, 2002). Cognitive narratology further views narrative as a cognitive structure through which readers construct and orient themselves within narrative worlds (Herman, 2009). Despite these developments, discussion of POV in English language teaching (ELT) has remained largely confined to literary analysis rather than classroom reading comprehension.

Contemporary reading-comprehension research increasingly views comprehension as a multidimensional process involving inferential reasoning, coherence construction, perspective-taking, and the integration of textual information across multiple levels of meaning (Grabe, 2009; Koda, 2005). Meaning is constructed through interactions between textual information and prior knowledge (Kintsch, 1998; Van den Broek et al., 1999), suggesting that narrative structure may influence how readers engage with information during comprehension.

Narrative perspective may be particularly relevant during the late primary years, when learners increasingly move beyond literal comprehension toward inferential and interpretive engagement with texts. Research on reading development suggests that successful comprehension during this stage depends not only on vocabulary and linguistic knowledge but also on learners' ability to generate inferences, track perspectives, and build coherent representations of text (Cain and Oakhill, 2007; Grabe, 2009). Research on children's engagement with narrative fiction further indicates that narrative understanding involves the integration of characters, events, and social information during the construction of meaning (Gasser et al., 2022).

Empirical studies indicate that narrative perspective is associated with differences in how readers construct and engage with narrative meaning. Readers may adopt perspective-dependent viewpoints during narrative processing (Brunyé et al., 2009, 2011, 2016), while changes in narrative perspective have also been associated with differences in experiential engagement and emotional involvement during reading (Hartung et al., 2016, 2017). Such experiences may also vary across readers, with individual differences in reading experience, mental simulation, perspective taking, and narrative appreciation shaping how readers engage with literary texts (Mak et al., 2022).

From a comprehension perspective, FP narration may encourage attention to locally situated meanings, whereas TP narration may provide broader contextual information that supports coherence building and information integration. These differences suggest that narrative POV may be associated with differing patterns of reading engagement and comprehension performance.

Despite its theoretical and empirical relevance, relatively little ELT research has examined narrative POV as a pedagogically meaningful feature of classroom reading materials. Reading texts are often selected according to thematic relevance, vocabulary level, or linguistic complexity, whereas narration type receives comparatively limited attention. Addressing this gap may clarify whether narrative POV should be considered a component of instructional text design.

2.2 First-person narration and inferential processing

First-person (FP) narration presents events through a character-bound perspective, providing access to experiences, thoughts, and interpretations from a limited viewpoint. Because information is filtered through a single narrator, readers may rely more heavily on local textual cues, perspective-taking, and inferential interpretation to construct narrative meaning. Inferential Processing involves generating connections among ideas, interpreting implicit meanings, and integrating textual information with prior knowledge (Kintsch, 1998; Van den Broek et al., 1999). These demands may become more salient in FP narration, where access to broader contextual information is comparatively restricted. Recent work with young children further suggests that story understanding involves updating mental representations as readers integrate verbal information across the developing narrative (Lee, 2025).

Research suggests that readers often adopt perspective-dependent viewpoints during narrative engagement (Brunyé et al., 2009, 2011, 2016), indicating that FP narration may encourage closer attention to character-centered meaning construction, particularly among learners whose inferential reasoning abilities are still developing (Cain and Oakhill, 2007; Oakhill and Cain, 2012).

From an instructional perspective, FP narration may support activities emphasizing interpretation, perspective-taking, character motivation, and inferential reasoning (Bland, 2018; Lazar, 1993). At the same time, its interpretive demands may increase the need for scaffolding through guided questioning and explicit modeling of inferential reading strategies (Gibbons, 2015; Walqui, 2006).

Taken together, these considerations suggest that FP narration may encourage greater inferential engagement and attention to localized meaning construction during reading.

2.3 Third-person omniscient narration and structural coherence

Third-person omniscient (TP) narration differs from first-person narration in the scope of information available to readers. Rather than restricting access to a single perspective, it provides broader access to characters, events, contextual information, and narratorial insights, allowing relationships and motivations to be interpreted within a wider narrative framework.

Within reading-comprehension research, coherence construction is regarded as a fundamental component of understanding, requiring readers to integrate information across textual units and construct meaningful representations (Grabe, 2004; Kintsch, 1998; Van den Broek et al., 1999). Readers must establish connections among events, recognize causal relationships, and integrate information into coherent representations of narrative meaning. Textual features that make such relationships more explicit may facilitate information integration and broader coherence.

TP narration may provide this support by offering broader contextual information and multiple perspectives. Because readers are not restricted to a single viewpoint, relationships among events and motivations may be more readily accessible, while narratorial explanations may assist information integration across the text. Consequently, TP narration may support broader organization and integration of narrative meaning.

These considerations may be particularly relevant during the primary years, when learners continue to develop their ability to integrate information across larger textual units and construct coherent representations of narrative meaning (Cain and Oakhill, 2007; Oakhill and Cain, 2012; Lindgren, 2023).

Research suggests that readers' engagement with narratives is influenced by the availability and organization of information within texts (Graesser et al., 1994; Kintsch, 1998). Narratives that provide greater contextual guidance may facilitate coherent mental representations by reducing the need to infer relationships that are not explicitly represented.

From an instructional perspective, TP narratives may support activities emphasizing thematic analysis, narrative organization, causal relationships, and connections between local details and broader narrative structures. Recent intervention research also suggests that explicit attention to narrative structure can support young learners' development of narrative understanding (Lilja et al., 2025).

Taken together, these considerations suggest that TP narration may encourage greater attention to the organization and integration of narrative information across a text. The characteristics of TP narration nevertheless provide a plausible basis for examining its relationship with structurally oriented dimensions of reading comprehension in ELT contexts.

2.4 Multidimensional comprehension in ELT contexts

Reading comprehension is widely understood as a complex process involving the construction of meaning through interactions among the reader, the text, and the activity or context of reading (Snow, 2002). Contemporary research further views comprehension as involving multiple processes, including inferential reasoning, coherence construction, and the integration of information across levels of meaning (Bernhardt, 2005; Grabe, 2009; Kintsch, 1998; Venneker et al., 2025). Successful reading involves interpreting implicit meanings, organizing relationships among ideas, understanding local textual details, and constructing broader representations of narrative meaning. This perspective is particularly relevant in ELT contexts, where learners encounter varying linguistic, interpretive, and organizational demands (Grabe, 2009; Koda, 2005). Consequently, a multidimensional framework may provide a more nuanced understanding of learners' engagement with texts and textual features.

The present study distinguishes between Processing-Oriented and Comprehension-Oriented dimensions. Processing-Oriented dimensions concern how readers engage with and organize textual information during reading, whereas Comprehension-Oriented dimensions reflect the outcomes of that engagement.

Inferential Processing refers to the ability to interpret implicit meanings, establish connections among ideas, and construct relationships not explicitly stated in the text (Cain and Oakhill, 2007; Kintsch, 1998). Structural Coherence Processing refers to the ability to recognize causal relationships, organize information across events, and integrate textual elements into coherent representations of narrative meaning.

Micro-Level Comprehension refers to understanding localized narrative information, including character actions, immediate contextual meanings, and perspective-specific details. Macro-Level Comprehension refers to broader thematic, causal, and organizational understanding derived from integrating information across the narrative as a whole.

Although related, Processing-Oriented and Comprehension-Oriented dimensions are conceptually distinct. Patterns of engagement with textual information may be associated with particular comprehension outcomes, but no single processing orientation necessarily determines a specific outcome. A multidimensional framework therefore provides a basis for examining how different narration conditions may be associated with differing patterns of processing and comprehension. Such an approach may contribute to more informed reading instruction and text selection in ELT contexts.

2.5 Research gap and study rationale

The literature reviewed above suggests that narrative POV is a theoretically significant feature of narrative texts, and that reading comprehension is best understood as a multidimensional construct. However, several gaps remain.

First, research on narrative perspective has been conducted largely within literary, narratological, or general reading contexts, with comparatively limited attention to how narration conditions relate to reading performance among young English language learners in authentic classroom-based ELT settings.

Second, reading comprehension has often been examined through overall performance measures, providing limited insight into how narrative perspective may relate to specific dimensions of reading. Comparatively few studies have investigated whether narration conditions are associated with differing patterns across Processing-Oriented and Comprehension-Oriented dimensions.

Third, although theoretical scholarship suggests that FP and TP narration provide differing forms of informational access and organizational support, their relationship with multidimensional reading performance remains insufficiently understood. Existing research offers valuable theoretical insights but limited classroom-based evidence regarding their relationship with learner engagement and comprehension in primary-level ELT contexts.

Addressing these gaps is important for both theory and practice. Examining narrative POV within a multidimensional framework may contribute to a more nuanced understanding of how textual features relate to reading performance, while insights into learners' responses to different narration conditions may inform text selection, instructional design, and reading-comprehension practices.

Accordingly, the present study investigates the relationship between FP and TP narration and multidimensional reading performance among Grades 4 and 5 learners. Specifically, it examines Inferential Processing, Structural Coherence Processing, Micro-Level Comprehension, and Macro-Level Comprehension to explore how narration conditions may be associated with differing patterns of reading performance. By integrating narrative perspective, multidimensional comprehension, and classroom-based ELT practice, the study seeks to extend understanding of how narrative design features relate to learner engagement and comprehension outcomes.

3 Conceptual framework

The present study is guided by a conceptual framework that integrates narrative point of view (POV), Processing-Oriented dimensions of reading, and Comprehension-Oriented outcomes within classroom-based ELT contexts. Drawing on narratology (Chatman, 1978; Genette, 1980), reading-comprehension theory (Kintsch, 1998; Van den Broek et al., 1999), and narrative-engagement research (Brunyé et al., 2009, 2011, 2016; Hartung et al., 2016, 2017), the framework conceptualizes POV as a pedagogically meaningful textual feature that may be associated with differing patterns of learner engagement and reading performance. Rather than assuming a uniform influence of narrative perspective on comprehension, the framework proposes that different narration conditions may encourage different forms of engagement with textual information during reading.

Within this framework, first-person (FP) narration is associated with an Inferential Processing orientation. Because information is presented through a character-bound perspective, readers may rely more heavily on interpreting implicit meanings, establishing unstated relationships, and drawing inferences from localized textual information. Third-person omniscient (TP) narration, by contrast, is associated with a Structural Coherence Processing orientation. Through broader access to events, relationships, and contextual information, it may provide greater support for organizing and integrating information across the narrative. These associations represent relative emphases rather than exclusive processing tendencies.

The framework further proposes that these Processing-Oriented dimensions may be associated with differing Comprehension-Oriented outcomes. Inferential Processing is conceptually linked to forms of understanding that depend on interpreting localized meanings, character perspectives, and implicit relationships, whereas Structural Coherence Processing is linked to the organization and integration of information across broader narrative structures. These expectations are theoretical rather than deterministic and serve as conceptual guides for the investigation.

Importantly, the framework does not assume fixed or linear relationships among POV, processing orientations, and comprehension outcomes. Reading performance may also be shaped by factors such as linguistic proficiency, prior knowledge, reading experience, and developmental differences. Accordingly, the framework proposes associations rather than direct causal pathways.

As illustrated in Figure 1, FP and TP narration are theorized to be associated with differing processing orientations that may be reflected in multidimensional patterns of comprehension performance. Grade Level is conceptualized as a moderating factor that may influence the strength and expression of these associations.

By distinguishing between Processing-Oriented dimensions and Comprehension-Oriented outcomes, the framework provides a basis for examining how narrative design features may be associated with multidimensional patterns of reading performance among young English language learners. The research questions guiding the investigation are presented below.

3.1 Research questions

The study was guided by the following research questions:

RQ1. What differences emerge in overall multidimensional reading performance between first-person and third-person omniscient narration among upper-primary English language learners?

RQ2. How do first-person and third-person omniscient narration differentially relate to inferential processing, structural coherence processing, and their associated micro-level and macro-level comprehension outcomes?

RQ3. To what extent does grade level moderate the relationship between narrative point of view, processing orientations, and associated comprehension outcomes?

4 Methodology

This study examined how narrative point of view (POV) may be associated with multidimensional reading performance among upper-primary English language learners. Guided by the conceptual framework presented in Section 3, a classroom-based quasi-experimental design was employed within an authentic educational setting.

The following sections describe the research design, participants and context, materials and narrative adaptation procedures, assessment framework, data collection procedures, and analytical methods.

4.1 Research design

The study employed a classroom-based quasi-experimental design comparing first-person (FP) and third-person omniscient (TP) narration. The design examined their associations with two Processing-Oriented dimensions (Inferential Processing and Structural Coherence Processing) and two Comprehension-Oriented dimensions (Micro-Level and Macro-Level Comprehension). Rather than treating reading comprehension as a single outcome, the study adopted a multidimensional approach to examine patterns of performance across these four dimensions.

A quasi-experimental approach was adopted because the research was conducted within an authentic classroom environment where existing class groups were maintained for instructional and administrative reasons. This design enabled the investigation of narrative perspective while preserving ecological validity and reflecting typical English reading instruction.

Consistent with the conceptual framework, the study focused on associations among narrative perspective, Processing-Oriented dimensions, and comprehension outcomes rather than on direct observation of cognitive mechanisms.

To minimize the influence of extraneous variables, narrative versions were adapted from equivalent source texts and administered under comparable instructional conditions. Additional procedures were implemented to support consistency in text difficulty, assessment structure, administration, and scoring across narration conditions, as described in the following sections.

4.2 Participants and context

A total of 115 upper-primary learners participated in the study, comprising 55 boys and 60 girls enrolled in Grade 4 (n = 57) and Grade 5 (n = 58) at a CBSE-affiliated private English-medium school in Karnataka, India. Participants were drawn from four intact classroom groups: Grade 4A (n = 28), Grade 4B (n = 29), Grade 5A (n = 31), and Grade 5B (n = 27). Learners were aged approximately 9–11 years. These grade levels were selected because learners during the late primary years increasingly engage in inferential and interpretive reading while continuing to develop broader comprehension and coherence-building skills.

Participants were selected through intact classroom sampling to preserve the natural instructional context. English served as the primary language of instruction, although it was not the primary home language for most learners. The school primarily served lower-middle and middle-income semi-urban communities, with a small proportion of learners transitioning from nearby rural settings.

The linguistic profile reflected the multilingual diversity typical of Karnataka's English-medium schools, with Kannada reported as the primary home language for most learners and smaller groups reporting Tamil, Telugu, Malayalam, Marathi, or Hindi. This context provided a relevant setting for examining reading performance in multilingual ELT classrooms.

All participants had prior experience with narrative texts through classroom reading activities and graded English reading materials appropriate to their instructional level, reducing the likelihood of task-novelty effects.

Prior to the main study, a baseline classroom reading assessment was administered to obtain a general indication of English reading proficiency. The assessment was used primarily to ensure that participants possessed sufficient reading ability to engage meaningfully with the narrative materials and multidimensional reading tasks employed in the study. Although Grade 5 learners demonstrated somewhat stronger baseline performance than Grade 4 learners, all groups showed sufficient proficiency to participate effectively in the reading activities and assessments.

4.3 Materials and narrative adaptation

Two narrative texts appropriate for upper-primary English language learners were selected from graded reading materials aligned with learners' proficiency and developmental level. The two narratives represented realistic and non-realistic story forms, respectively, a distinction relevant to narrative materials used with young English language learners (Panchal, 2026). Selection was based on linguistic accessibility, narrative clarity, thematic appropriateness, and suitability for Grades 4 and 5. Both texts reflected story-based materials commonly encountered in English language classrooms.

Each narrative was adapted into two parallel versions representing first-person (FP) and third-person omniscient (TP) narration. The adaptation process aimed to preserve narrative equivalence while varying POV as the primary textual condition under investigation.

To maintain comparability across conditions, storyline, sequence of events, characters, settings, and thematic content were retained across versions. Narrative perspective was modified primarily through changes to pronouns, perspective markers, and narrative voice, with minimal linguistic adjustments to preserve readability and coherence. The adaptations were intended to ensure that any differences in reading performance could be more plausibly associated with narration condition rather than broader textual variation. The adapted texts were reviewed for approximate equivalence in length, vocabulary level, and sentence complexity. Each version contained approximately 220–250 words, with vocabulary and grammatical structures aligned to graded-reader specifications appropriate for upper-primary learners. Key characteristics of the adapted materials are summarized in Supplementary Appendix A.

Readability equivalence was established through graded-reader selection, controlled adaptation procedures, and expert review, with emphasis placed on consistency in instructional difficulty, narrative organization, and classroom readability rather than reliance solely on formula-based readability indices.

To support content validity and classroom appropriateness, the adapted materials were reviewed by three specialists in ELT and reading instruction. The review focused on linguistic suitability, developmental appropriateness, narrative clarity, and comparability across narration conditions. Feedback informed minor revisions to wording, narrative flow, and presentation prior to classroom implementation.

4.4 Instrumentation and assessment framework

A structured reading-comprehension assessment was developed to examine learners' performance across multiple dimensions of narrative understanding. Consistent with the conceptual framework presented in Section 3, the assessment was designed to capture both Processing-Oriented and Comprehension-Oriented aspects of reading performance rather than relying solely on an overall comprehension score.

The framework comprised four dimensions organized into two Processing-Oriented dimensions (Inferential Processing and Structural Coherence Processing) and two Comprehension-Oriented dimensions (Micro-Level Comprehension and Macro-Level Comprehension). Operational definitions are presented in Table 1.

Table 1. Assessment dimensions and operational definitions.

Together, the four dimensions were intended to capture complementary aspects of narrative reading performance among upper-primary learners. The Processing-Oriented dimensions focused on how learners engaged with and organized textual information during reading, whereas the Comprehension-Oriented dimensions focused on the outcomes of that engagement as reflected in learners' understanding of narrative meaning. This distinction provided a multidimensional basis for examining reading performance across narration conditions.

Each narrative text was followed by a structured assessment designed to examine all four dimensions separately. The assessment comprised eight items per dimension, including four multiple-choice items, three short constructed-response items, and one extended-response item. This mixed-format structure was intended to capture both recognition-based and interpretive aspects of reading performance.

Multiple-choice items were scored dichotomously (1 = correct, 0 = incorrect), whereas constructed-response items were evaluated using an analytic rubric. Scores were aggregated separately for each dimension and used to examine patterns of performance across narration conditions. A summary of the assessment framework and item distribution is provided in Supplementary Appendix B.

4.5 Pilot testing, validation, and reliability

Prior to the main study, the assessment was reviewed by three specialists in ELT, reading instruction, and language assessment. The review focused on item clarity, developmental appropriateness, representativeness, and alignment with the four assessment dimensions. Particular attention was given to the alignment of assessment tasks with the intended Processing-Oriented and Comprehension-Oriented dimensions. Feedback informed minor revisions to item wording, task instructions, and response requirements, thereby strengthening content validity and classroom suitability.

Following expert review, the assessment was piloted with a comparable group of learners from a similar educational context to evaluate item clarity, response feasibility, and overall suitability for upper-primary English language learners. Pilot administration helped identify ambiguities in item interpretation and task completion. Pilot responses indicated that the assessment was appropriate for the target population. Minor revisions were introduced to improve wording clarity and task presentation, while the overall assessment framework remained unchanged.

To support scoring consistency, multiple-choice items were scored dichotomously (1 = correct, 0 = incorrect), whereas constructed-response items were evaluated using an analytic rubric. Approximately 20% of learner responses were re-evaluated by the same scorer after an interval period, yielding a scoring agreement rate of approximately 88%, indicating acceptable scoring consistency.

Internal consistency reliability was evaluated using Cronbach's alpha coefficients prior to the primary statistical analyses. Reliability estimates were calculated for individual dimensions and composite measures, with results reported in the Results section.

4.6 Procedure

The study was conducted during regular classroom hours across multiple sessions. Data collection followed a structured sequence consisting of a baseline assessment followed by two reading sessions corresponding to the narration conditions examined in the study.

4.6.1 Baseline assessment

Prior to the reading sessions, all participants completed a researcher-developed baseline assessment designed to provide a general indication of English reading proficiency. The assessment served as a classroom-based indicator of readiness for the narrative reading tasks rather than as a standardized proficiency measure.

The assessment consisted of a grade-appropriate passage (approximately 130–150 words) followed by seven comprehension items, including four multiple-choice and three short-answer questions. Multiple-choice items were scored dichotomously (1 = correct, 0 = incorrect), whereas short-answer responses were evaluated using a rubric-based scoring scheme ranging from 0 to 2, yielding a maximum score of 10. Although both the baseline assessment and the multidimensional assessment evaluated reading comprehension, the baseline assessment focused on general classroom reading proficiency, whereas the multidimensional assessment examined performance across the four study dimensions. A sample version of the assessment is provided in Supplementary Appendix A.

As shown in Table 2, Grade 5 learners demonstrated somewhat higher baseline reading performance than Grade 4 learners; however, both groups showed sufficient proficiency to engage meaningfully with the narrative materials and assessments.

Table 2. Baseline reading assessment performance by grade level.

4.6.2 Reading sessions

Participants subsequently completed two reading sessions corresponding to the first-person (FP) and third-person omniscient (TP) narration conditions. During each session, learners read one narrative text followed immediately by the associated multidimensional assessment.

To minimize order effects, narration conditions were counterbalanced across groups. Groups 4A and 5A initially completed the FP condition, whereas Groups 4B and 5B initially completed the TP condition. Existing classroom groupings were maintained throughout the study.

The two sessions were separated by an interval of approximately 2 to 3 days to reduce immediate recall effects while maintaining continuity of classroom engagement. Learners read the assigned narrative independently within a 30–40-min classroom period under supervised conditions. Standardized instructions were provided across groups, and no additional explanation or instructional scaffolding was introduced during reading. Immediately following each session, participants completed the multidimensional assessment individually without peer discussion or collaborative interaction.

All sessions were conducted by the same instructor to maintain consistency across groups and narration conditions. Timing, instructions, and assessment conditions were standardized throughout the study. Participants were not informed about the specific focus on narrative POV in order to minimize response bias associated with narration awareness.

4.7 Data analysis procedures

Data were analyzed using Jamovi (Version 2.6.45). Prior to inferential analyses, descriptive statistics were calculated to examine score distributions and variability across narration conditions and assessment dimensions. Internal consistency reliability was evaluated using Cronbach's alpha coefficients, as described in Section 4.5.

For RQ1, a repeated-measures ANOVA (RM ANOVA) examined differences in overall multidimensional reading performance between the first-person (FP) and third-person omniscient (TP) conditions. Composite scores were calculated by aggregating performance across the four assessment dimensions.

For RQ2, a 2 × 4 repeated-measures ANOVA examined whether narration conditions were associated with differing patterns of performance across the four assessment dimensions.

For RQ3, a mixed-design ANOVA examined whether grade level moderated the relationship between narration condition and reading performance, with narration condition as the within-subject factor and grade level as the between-subject factor.

Assumptions of normality and sphericity were examined prior to analysis. Where violations of sphericity occurred, Greenhouse–Geisser corrections were applied. Statistical significance was evaluated at p < 0.05, and effect sizes were reported using partial eta squared (ηp²).

4.8 Ethical considerations

The study was conducted in accordance with established ethical principles for classroom-based educational research involving young learners. As the investigation was embedded within regular English reading activities, procedures were designed to maintain normal instructional conditions and minimize educational risk.

Approval to conduct the study was obtained from the school principal prior to data collection. The purpose of the study, data collection procedures, and intended use of the findings were communicated to the school administration, and parents or guardians were informed through established school communication procedures.

Participation was voluntary, and verbal assent was obtained from all learners. Participants were informed that they could withdraw from the study at any stage without academic consequence.

To ensure confidentiality, learner responses were anonymised through numerical codes, and no personally identifiable information was included in the reporting of findings. All data were used exclusively for research purposes and stored securely throughout the study.

Procedures were designed to minimize disruption to classroom instruction while ensuring fairness and confidentiality throughout the research process.

5 Results

This section presents the study findings. Reliability evidence for the multidimensional reading assessment is reported first, followed by descriptive statistics and inferential analyses addressing the three research questions. The analyses examine overall differences in reading performance across narration conditions (RQ1), differential patterns across the four assessment dimensions (RQ2), and the moderating role of grade level (RQ3).

5.1 Reliability of composite and dimension-level measures

Internal consistency reliability was evaluated using Cronbach's alpha coefficients for composite scores and individual assessment dimensions across narration conditions. The resulting reliability estimates are presented in Table 3.

Table 3. Cronbach's alpha coefficients for assessment dimensions by narration condition.

FP, first-person narration; TP, third-person omniscient narration. Bold values indicate the composite score, whereas non-bold values represent the respective assessment dimensions.

As shown in Table 3, reliability coefficients ranged from acceptable to strong across the assessment dimensions. Composite scores demonstrated particularly high internal consistency under both narration conditions (FP α =0.90; TP α =0.92), supporting the use of aggregated scores in subsequent analyses. Dimension-level coefficients also remained acceptable for a classroom-based multidimensional assessment involving multiple item formats and varied comprehension demands. Reliability estimates were generally comparable across narration conditions, with slightly higher coefficients observed for several dimensions under the TP condition.

Overall, the findings indicate satisfactory internal consistency at both the composite and dimension levels, supporting the reliability of the assessment framework.

5.2 Descriptive statistics across narration conditions and assessment dimensions

Descriptive statistics were computed for the four assessment dimensions under the first-person (FP) and third-person omniscient (TP) narration conditions. Means and standard deviations are presented in Table 4.

Table 4. Descriptive statistics for assessment dimensions by narration condition (N = 115).

M, mean; SD, standard deviation. Composite scores represent the sum of performance across the four assessment dimensions within each narration condition.

As shown in Table 4, FP narration was associated with higher mean scores in Inferential Processing and Micro-Level Comprehension, whereas TP narration was associated with higher mean scores in Structural Coherence Processing and Macro-Level Comprehension. Composite scores were also descriptively higher under the TP condition than under the FP condition.

The raw score difference between conditions was approximately 2.6 points on the composite scale, indicating that the statistically significant effect was associated with a moderate difference in observed performance across narration conditions. Further analyses examined whether this overall difference varied across the individual assessment dimensions.

These patterns are broadly consistent with the conceptual framework proposed in Section 3, which suggests differing associations between narration conditions, Processing-Oriented dimensions, and Comprehension-Oriented outcomes. Inferential analyses examining the statistical significance of these differences are presented in the following sections.

5.3 Overall multidimensional reading performance (RQ1)

A repeated-measures ANOVA was conducted to examine whether overall multidimensional reading performance differed across narration conditions. The analysis revealed a significant main effect of narration condition, F(1, 113) = 83.70, p < 0.001, ηp² = 0.425, indicating that overall reading performance varied significantly between first-person (FP) and third-person omniscient (TP) narration. As shown in Table 5, composite scores were higher under the TP condition than under the FP condition.

Table 5. Repeated-measures ANOVA for overall multidimensional reading performance (RQ1).

FP, first-person narration; TP, third-person omniscient narration; ηp², partial eta squared.

The raw score difference between conditions was approximately 2.6 points on the composite scale, indicating a moderate difference in observed performance across narration conditions. These findings suggest an overall advantage for TP narration in multidimensional reading performance. Further analyses examined whether this overall difference varied across the individual assessment dimensions.

5.4 Differential effects of narrative point of view across assessment dimensions (RQ2)

A 2 × 4 repeated-measures ANOVA was conducted to examine whether performance patterns differed across narration conditions and assessment dimensions. Table 6 presents the results.

Table 6. Repeated-measures ANOVA for assessment dimensions (RQ2).

Type III sums of squares reported.

The non-significant main effect of assessment dimension indicates that average scores did not differ substantially across dimensions when narration conditions were considered collectively. However, the significant Narration × Dimension interaction indicates that the pattern of performance across dimensions differed according to narration condition. Figure 2 illustrate the crossover pattern.

As illustrated in Figure 2, FP narration was associated with higher performance in Inferential Processing and Micro-Level Comprehension, whereas TP narration was associated with higher performance in Structural Coherence Processing and Macro-Level Comprehension. This crossover pattern reflects differing performance profiles across narration conditions and assessment dimensions and is broadly consistent with the conceptual framework proposed in Section 3, which suggests differing associations between narration conditions, Processing-Oriented dimensions, and Comprehension-Oriented outcomes.

5.5 Moderating role of grade level (RQ3)

A mixed-design ANOVA examined whether grade level moderated the relationship between narration condition and multidimensional reading performance. The analysis revealed a non-significant main effect of grade level, F(1, 113) = 0.84, p = 0.363, ηp² = 0.007, but a significant Narration × Grade interaction, F(1, 113) = 13.50, p < 0.001, ηp² = 0.107 (Table 7).

Table 7. Mixed-design ANOVA examining the moderating role of grade level (RQ3).

ηp², partial eta squared. FP, first-person narration; TP, third-person omniscient narration.

The non-significant main effect indicates that overall multidimensional reading performance did not differ substantially across grades. However, the significant interaction indicates that performance patterns across narration conditions varied by grade level. The interaction is illustrated in Figure 3.

As illustrated in Figure 3, grade-level differences were more pronounced under FP narration than under TP narration. Grade 5 learners demonstrated moderately higher performance than Grade 4 learners under the FP condition, whereas differences between grades were comparatively small under the TP condition. These findings suggest that grade-level variation was associated with narration condition rather than reflecting a uniform difference in performance across the study.

6 Discussion

This study examined whether narrative point of view (POV) is associated with multidimensional reading performance among upper-primary English language learners. The findings suggest that narrative perspective may function not only as a stylistic feature of narrative texts but also as a pedagogically meaningful textual feature associated with differing patterns of reading performance. Consistent with the proposed conceptual framework, first-person (FP) and third-person omniscient (TP) narration were associated with distinct performance profiles across assessment dimensions rather than a uniform advantage in comprehension.

In addition, grade-level differences varied across narration conditions, suggesting that developmental variation may be expressed differently across narrative contexts rather than as a general advantage in reading performance. The following sections discuss these findings in relation to the study's research questions, conceptual framework, and implications for English language teaching (ELT).

6.1 Narrative point of view and multidimensional reading performance

The most distinctive finding of the study was the significant interaction between narration condition and assessment dimension. Rather than producing a uniform advantage, FP and TP narration were associated with distinct performance profiles across Inferential Processing, Structural Coherence Processing, Micro-Level Comprehension, and Macro-Level Comprehension. These findings suggest that narrative POV may be understood not only as a literary feature but also as a pedagogically meaningful textual variable associated with multidimensional reading performance. This pattern is broadly consistent with research indicating that narrative perspective can shape readers' engagement with textual information and influence how narrative meaning is constructed during reading (Brunyé et al., 2011; Hartung et al., 2016).

The significant interaction suggests that narrative POV may be associated not only with overall reading performance but also with the dimensions through which learners construct meaning from narrative texts.

FP narration was associated with stronger performance in Inferential Processing and Micro-Level Comprehension. One possible explanation is that its character-bound perspective encourages closer engagement with perspective-specific information, internal states, and local textual cues. Because information is presented through a single narrator, readers may need to infer relationships, motivations, and meanings that are not explicitly stated (Brunyé et al., 2011; Hartung et al., 2017). Such conditions may promote greater reliance on inferential interpretation and attention to immediate contextual information, consistent with models of reading comprehension that emphasize the integration of explicit information and reader-generated inferences during meaning construction (Kintsch, 1998; Van den Broek et al., 1999).

By contrast, TP narration was associated with stronger performance in Structural Coherence Processing and Macro-Level Comprehension. The broader narratorial perspective may provide greater access to relationships among events, characters, and narrative developments, thereby supporting coherence building and broader narrative interpretation. This explanation is consistent with perspectives that emphasize the role of structural guidance and organized meaning-making during reading (Herman, 2009; Kintsch, 1998).

Importantly, the findings do not indicate the superiority of one narration condition over the other. Rather, they suggest that FP and TP narration may support different dimensions of reading performance depending on the nature of the comprehension demands involved. Taken together, the findings reinforce the view that reading comprehension is multidimensional and that narrative design features may be associated with different aspects of meaning construction (Van den Broek et al., 1999).

6.2 Developmental sensitivity across narration conditions

The findings further suggest that grade-level variation in reading performance was associated with narration condition rather than expressed uniformly across reading contexts. Although no significant overall grade-level effect was observed, the significant Narration × Grade interaction indicates that differences between Grades 4 and 5 varied across narration conditions.

Grade-level differences were more pronounced under FP narration than under TP narration. One possible explanation is that FP narration places greater emphasis on dimensions such as Inferential Processing and Micro-Level Comprehension, where learners may differ in their ability to interpret perspective-specific information and construct meaning from local textual cues. Under such conditions, developmental differences may become more readily observable (Cain and Oakhill, 2007).

By contrast, differences between grades were less pronounced under TP narration. The broader narratorial perspective may provide greater structural support for organizing and interpreting narrative information, thereby reducing the extent to which developmental differences are reflected in overall performance. This interpretation is consistent with perspectives that emphasize the supportive role of textual organization and scaffolding during comprehension (Vygotsky, 1978; Hammond and Gibbons, 2005).

The absence of a significant overall grade-level effect suggests that developmental variation was associated with the interaction between learner characteristics and narration conditions rather than with grade level alone.

Taken together, the findings suggest that developmental variation may be shaped not only by learner characteristics but also by narrative context. Narrative POV may therefore influence the conditions under which developmental differences become more or less visible. From an instructional perspective, this highlights the importance of aligning narrative materials with learners' developmental readiness and instructional goals (Walqui, 2006; Gibbons, 2015).

6.3 Pedagogical implications for ELT

The findings suggest that narrative POV may be considered a pedagogically meaningful feature of instructional text design in ELT. Although reading materials are often selected on the basis of topic, vocabulary level, readability, or thematic relevance, the results indicate that narrative perspective may also be associated with differing patterns of multidimensional reading performance. Consequently, narration condition may be considered alongside other instructional factors when selecting or adapting classroom reading materials.

From an instructional perspective, FP narration was associated with stronger performance in Inferential Processing and Micro-Level Comprehension, suggesting potential value for activities emphasizing inferential interpretation, perspective-taking, character understanding, and close attention to local textual information. TP narration, by contrast, was associated with stronger performance in Structural Coherence Processing and Macro-Level Comprehension, suggesting potential value for activities emphasizing thematic understanding, narrative organization, summarisation, and broader interpretation of relationships among events (Walqui, 2006; Gibbons, 2015; Hammond and Gibbons, 2005).

These associations should not be interpreted as fixed pedagogical prescriptions. Both narration conditions engage multiple dimensions of reading performance, and their effectiveness is likely to vary according to learner characteristics, instructional support, task design, and broader classroom objectives. Rather, the findings suggest that narrative perspective may be strategically aligned with particular instructional goals depending on the dimensions of reading performance teachers wish to emphasize.

6.4 Theoretical contributions

Beyond its pedagogical implications, the study contributes to the theoretical understanding of narrative POV in reading comprehension. Whereas narrative perspective has traditionally been discussed primarily as a literary feature associated with characterization, voice, and narrative presentation (Herman, 2009), the findings suggest that it may also function as a pedagogically meaningful textual variable associated with multidimensional reading performance.

A second contribution lies in its support for a multidimensional view of reading comprehension. Rather than exerting a uniform influence on comprehension, narrative POV was associated with differing patterns across assessment dimensions. These findings suggest that narrative perspective may be associated with particular dimensions of reading performance rather than with comprehension as a single undifferentiated outcome (Kintsch, 1998; Van den Broek et al., 1999).

More broadly, the findings suggest that different narration conditions may provide differential support for particular dimensions of reading performance rather than producing a uniform enhancement of comprehension. Together, these findings extend current discussions of narrative perspective beyond overall comprehension performance and highlight the importance of examining how textual features relate to multiple dimensions of reading.

The findings also contribute to discussions concerning the relationship between textual design and learner performance. The significant interaction between narration condition and assessment dimension suggests that narrative features may be associated with how different aspects of reading performance are supported during comprehension tasks. From a broader text-comprehension perspective, this finding is consistent with accounts that view comprehension as involving the construction and integration of representations at different levels of text processing (Van Dijk and Kintsch, 1983; Kintsch, 1998). Thus, narrative perspective may be considered not only a stylistic characteristic but also a textual feature that may be associated with different dimensions of comprehension performance.

A further contribution concerns developmental variation. Although no significant overall grade-level effect was observed, developmental differences were expressed differently across narration conditions. This finding suggests that learner development and textual design may be more closely interconnected than is often assumed, with developmental variation becoming more or less visible depending on the narrative context in which reading occurs.

Taken together, these contributions support a view of narrative POV as a meaningful component of instructional text design associated with differing patterns of multidimensional reading performance. By bringing together perspectives from narrative theory, reading-comprehension research, and classroom-based inquiry, the study contributes to a more nuanced understanding of how narrative design features relate to reading performance among young English language learners.

6.5 Limitations and directions for future research

The findings should be interpreted within several methodological and contextual limitations, which also suggest directions for future research.

First, although a counterbalanced within-subject design was employed, the study relied on a specific set of adapted narrative materials. Despite efforts to maintain equivalence across narration conditions, subtle linguistic and discourse-level differences may have influenced performance. Future research employing additional texts, genres, and discourse structures would help determine the robustness of the observed patterns.

Second, the study was conducted within a single English-medium school involving learners from a limited grade range. Consequently, the findings should not be assumed to generalize across broader age groups, proficiency levels, or educational settings. Studies involving multiple schools, instructional contexts, and learner populations would strengthen understanding of how narrative perspective relates to multidimensional reading performance.

Third, the investigation focused exclusively on FP and TP narration. Future research may examine additional narrative forms, including third-person limited narration, second-person narration, shifting focalisation, and multimodal narrative formats. Future work may therefore extend the framework by examining additional dimensions and outcomes.

The study also focused primarily on immediate classroom-based reading performance. Alternative approaches, including think-aloud protocols, eye-tracking, response-time analyses, discourse-based investigations, and longitudinal interventions, may provide deeper insight into how narrative perspective relates to reading processes and comprehension development over time. Such approaches could complement the performance-based measures used in the present study and contribute to a more comprehensive understanding of narrative perspective in educational contexts.

Finally, future research may explore how narrative POV functions within broader instructional and technological environments, including digital reading platforms, collaborative reading tasks, multimodal narratives, and AI-supported learning environments.

Taken together, these directions suggest that narrative POV remains a promising area for continued investigation within reading-comprehension research and ELT. Further work integrating diverse narrative forms, learner populations, and methodological approaches may contribute to a more comprehensive understanding of how narrative design features relate to multidimensional reading performance and instructional practice.

7 Conclusion

This study examined whether narrative point of view (POV) is associated with multidimensional reading performance among upper-primary English language learners. The findings indicate that narrative perspective was associated with both overall and dimension-specific patterns of performance. Although third-person omniscient (TP) narration demonstrated a modest overall advantage in composite performance, FP and TP narration were associated with different strengths across assessment dimensions. FP narration was associated with stronger performance in Inferential Processing and Micro-Level Comprehension, whereas TP narration was associated with stronger performance in Structural Coherence Processing and Macro-Level Comprehension. In addition, grade-level differences varied across narration conditions, suggesting that developmental variation may be expressed differently across narrative contexts.

Taken together, the findings suggest that narrative POV may function as a meaningful component of instructional text design rather than merely a literary feature of narrative texts. By demonstrating differing associations across Inferential Processing, Structural Coherence Processing, Micro-Level Comprehension, and Macro-Level Comprehension, the study highlights the value of considering narrative design alongside linguistic and instructional factors in ELT reading pedagogy.

More broadly, the findings reinforce the importance of multidimensional perspectives on reading comprehension. Rather than viewing comprehension as a single undifferentiated outcome, the results suggest that narrative design features may be associated with differing patterns of reading performance across dimensions. Continued investigation of narrative design variables may contribute to a more nuanced understanding of how textual features, learner characteristics, and instructional contexts interact to shape reading performance in English language classrooms.

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/flang.2026.1813257/full#supplementary-material

Supplementary material Supplementary material at the source

Supplementary material Supplementary material at the source

Footnotes

Edited by: Antonio Bova, Catholic University of the Sacred Heart, Italy

Reviewed by: Mirjam Anugerahwati, State University of Malang, Indonesia

Hartono Hartono, Sultan Agung Islamic University, Indonesia