2.1. Online EFL Learning Environments

The rapid shift to online education during the pandemic introduced substantial changes to teaching methodologies across academic disciplines, including EFL instruction. Studies have consistently highlighted the challenges and adaptations required in this digital transition, emphasizing the importance of flexibility and innovative pedagogical approaches, and well-designed online platforms (Alkhannani, 2021; Jaber, 2021; Kadir & Yunos, 2021; Lee, 2020; Mahyoob, 2020; Oraif & Elyas, 2021; Park & Park, 2018; Zapata-Cuervo, Montes-Guerra, Shin, Jeong, & Cho, 2022).

For instance, Mahyoob’s (2020) explored the technical and academic challenges that Saudi Arabian EFL student faced in online learning environments, noting barriers related to internet connectivity, limited digital resources, and communication constraints that hindered effective language learning. In contrast, Oraif and Elyas (2021) found that when online platforms were well-structured, students displayed higher engagement and participation, suggesting that the effectiveness of online EFL learning environments depends on the tools and interactivity they provide. Such findings highlight the variability in student experiences and underscore the importance of examining specific elements, such as immediacy and interaction, that differ between real-time and pre-recorded class formats.

Moreover, Yang and Xu (2023) emphasized the importance of e-readiness among both students and instructors, identifying teacher presence and student- teacher interaction as critical to overcoming online learning obstacles. Similarly, Reeves (2000) and Tinoca and Oliveira (2013) found that collaborative tools and clear instructional guidelines significantly improved student outcomes in online learning environments. These findings underscore the need to explore how specific online formats—such as real-time and pre-recorded classes—affect student engagement and emotional responses. Lee’s (2020) study also echoed these findings, revealing that while some students preferred a hybrid model post-pandemic, many favored a blended learning approach that allowed for both face-to-face and online interaction.

These findings indicate that instructional format and platform functionality can shape student engagement, underscoring the need to investigate the impacts of synchronous and asynchronous learning formats on student experiences.

2.2. Emotional Factors in EFL Learning: Anxiety, Confidence, and Willingness

EFL learning involves emotional factors such as anxiety, confidence, willingness, which significantly influence student engagement and outcomes. Anxiety, often characterized by tension and unease in language learning contexts, can hinder students’ participation in speaking, reading, and writing activities (Horwitz, Horwitz, & Cope, 1986; Jiang et al., 2023; Matsuda & Gobel, 2004). For example, Horwitz, Horwitz, and Cope (1986) identified foreign language anxiety as a major barrier to language acquisition, noting its negative impact on student performance. Similarly, Matsuda and Gobel (2004) observed that learners with higher anxiety levels avoided reading in English, leading to decreased comprehension and lower engagement. Studies suggest that synchronous interaction, where immediate feedback is available, can help reduce anxiety by creating a supportive and interactive environment (Jiang et al., 2023).

Confidence, closely linked to interaction and feedback quality, also plays a critical role. Studies suggest that online platforms with features like breakout room, collaborative tools, and real-time feedback mechanisms can bolster confidence by fostering interaction and providing timely support (Kadir & Yunos, 2021). Conversely, less interactive formats may contribute to feelings of isolation, exacerbating anxiety and hindering student engagement.

Willingness is another key factor influencing language acquisition. Maclntyre et al. (1998) emphasized willingness to communicate as a key predictor of learners’ language use and learning outcomes, highlighting its relevance in language acquisition. Peng and Woodrow (2010) further demonstrated that learners with higher willingness to use English were more likely to actively engage in using the language both inside and outside the classroom.

These studies suggest that real-time instructional formats, which offer immediate interaction and feedback, may encourage students to participate more actively in English. Conversely, pre-recorded formats, which lack immediacy, may pose challenges in sustaining learners’ motivation and willingness to use English.

2.3. Initial Attitudes and Format Preferences in Online EFL learning

Understanding students’ initial attitudes and preferences is critical in exploring how different online EFL learning formats impact their engagement and performance. Existing studies (Maclntyre et al., 1998; Peng and Woodrow, 2010) often focus on how students’ emotional responses evolve over time; however, this study emphasizes the role of pre-existing emotional factors―such as anxiety, confidence, and willingness―that influence students’ format preferences and early engagement patterns.

Building on this foundation, this study uses initial survey data collected at the beginning of the semester to explore students’ attitudes toward real-time and pre-recorded online EFL courses. By prioritizing these initial emotional responses, this study offers a unique perspective on self-selection tendencies in online EFL learning environments, highlighting how emotional factors shape students’ experiences and engagement patterns.

2.4. Synchronous vs. Asynchronous Learning Formats

Synchronous (real-time) and asynchronous (pre- recorded) learning formats each offer unique benefits and challenges. Synchronous classes provide immediacy and interaction, enabling students to actively participate and receive immediate feedback. This aligns with findings by Jiang et al. (2023), which highlight the positive impact of live interaction on motivation and engagement. Similarly, Oraif and Elyas (2021) observed high engagement levels among Saudi EFL learners using designated online platforms, suggesting that well-structured online environments can foster active participation. Furthermore, the real-time format fosters a sense of community, reducing feelings of isolation and enhancing students’ confidence to engage with the target language. Altınay (2017) emphasized the value of peer learning and assessment in online collaborative environments, emphasizing the importance of structured interaction and feedback for fostering student engagement. This aligns with the benefits observed in real-time formats, where immediacy and interaction contribute to a more dynamic learning experience.

Conversely, asynchronous formats offer flexibility and self-paced learning opportunities, which can benefit students with varying schedules or those who prefer independent study. However, research also indicates potential drawbacks, such as reduced engagement and increased isolation. MacIntyre et al. (1998) and Peng and Woodrow (2010) noted that lower interaction levels in asynchronous settings might lead to decreased motivation and higher anxiety, as students lack immediate support from instructors and peers.

Tinoca and Oliveira (2013) demonstrated how formative assessment in online environments can enhance teacher feedback and improve student outcomes, reinforcing the importance of integrating interactive and formative elements into online courses. These findings underscore the potential for structured feedback mechanisms to address motivational challenges in asynchronous settings. While prior studies have extensively explored the benefits and limitations of synchronous and asynchronous formats, there is limited research focusing on students’ initial emotional responses and preferences to synchronous and asynchronous learning. Previous studies (Chong, 2019, 2021, 2022, 2023a, 2023b) by the author examined how students’ emotional responses evolve throughout a semester but did not address how these initial emotions influence students’ preferences or engagement. This study fills that gap by comparing how real-time and pre-recorded formats impact students’ early emotional responses and academic performance, offering insights into the interplay between emotions, format preferences, and outcomes.

By situating this study within the broader discourse on EFL learning environments and emotional factors, this review highlights the importance of tailoring online instructional formats to meet diverse learner needs. It underscores the need for research that not only compares synchronous and asynchronous formats but also examines the underlying emotional and motivational drivers shaping students’ early engagement and success in online EFL courses.

3.1. Research Participants

The participants in this study were 127 college students enrolled in general education English courses, offered as part of their elective curriculum. The study included two groups corresponding to different online instructional format: 53 students in real-time online English classes conducted during the COVID-19 pandemic (2020) and 74 students in pre-recorded video online classes in the post-pandemic period (2023). Despite the time difference between the two groups, baseline equivalence was ensured by maintaining consistent course content, the same instructor, and standardized grading criteria across both formats. Both groups consisted of students who had completed two mandatory general English courses (College English 1 and College English Conversation 1) as part of their college curriculum. This selection criteria aimed to ensure a comparable level of English language background across both groups, minimizing the influence of language proficiency on emotional responses and academic performance.

In 2020, real-time online classes were implemented to address the urgent need for synchronous interaction, providing opportunities for live engagement between students and instructors. In 2023, the transition to pre-recorded video classes reflected the evolving preferences for flexibility and self-paced learning while maintaining the same pedagogical objectives.

To minimize potential confounding effects due to the time gap, this study focused on format-based differences (real-time vs. pre-recorded) rather than external factors. The primary emphasis is on format-based differences and how the unique characteristics of each (e.g., immediacy in real-time classes vs. flexibility in pre-recorded classes) influenced student responses. The findings are intended to shed light on each format’s advantages and limitations in fostering student engagement and learning outcomes, independent of the pandemic or post-pandemic context. Although the two classes occurred in different years, this setup enables a reliable comparison of synchronous and asynchronous instructional formats under a consistent curriculum, offering insights that remain relevant beyond specific timeframes. Both groups were drawn from the same university, participated in similar general English courses, and followed a standardized syllabus and evaluation criteria. These measures ensured a reliable basis for comparison, emphasizing the distinct characteristics of each format.

To prioritize initial emotional responses, a voluntary online survey was conducted during the second to third weeks of the semester, before significant engagement with course activities. This timing allowed the study to capture students’ pre-existing emotional factors, such as anxiety about language learning, confidence in English skills, and willingness to participate, which often shape students’ early engagement behaviors. These responses were treated as baseline indicators of student engagement potential. The survey was optional and conducted outside of regular class activities, leading to differing participation rates between the two groups: 84.9% of students in the real-time group (45 out of 53) and 55.4% in the pre-recorded group (41 out of 74). These response rates are summarized in Table 1.

<Table 1>

Research Participants by Online Format and Survey Participation

The differing response rates between the two groups highlight a limitation of this study. Although voluntary participation allowed students to provide honest feedback, the discrepancy in response rates may affect the representativeness of the data. This limitation is acknowledged in the discussion section, and future studies will aim to use mandatory surveys or more balanced sampling methods to mitigate this issue. By ensuring consistency in course content and grading, this study provides a comparison of emotional engagement and academic performance between real-time and pre-recorded formats, despite the inherent differences in response rates and timeframes.

3.2. Research Materials and Processes

The study employed two main data sources: surveys to capture students’ emotional responses and academic performance metrics to assess engagement and outcomes.

3.2.1. Survey Instrument

The survey instrument was designed based on the Foreign Language Classroom Anxiety Scale (FLCAS) developed by Horwitz, Horwitz, and Cope (1986). It consisted of 11 Likert-scale questions aimed at assessing students’ emotional and motivational responses to English learning, with specific focus on anxiety, confidence, positivity, and willingness. Table 2 outlines the Likert- scale questions, addressing anxiety (items 1-2), confidence (items 3-7), positivity (item 8), and willingness to engage with English (items 9-10), as well as familiarity with learning strategies (item 11).

<Table 2>

Survey Questions on Emotional and Motivational Responses to English Learning

Minor modifications were made to align the FLCAS items with the study’s specific focus on online instructional formats. The survey was administered online to ensure accessibility and anonymity, allowing students to provide honest feedback.

3.2.2. Academic Performance Metrics

To assess the impact of each instructional format on student learning, this study analyzed academic performance through two key evaluation metrics: attendance and exam scores. These metrics were chosen for their objective nature, providing insights into both student engagement and content mastery.

Attendance was assigned a 30% weight in the overall grade and tracked throughout the semester. It served as a proxy for student participation, particularly in the pre-recorded classes where active engagement was more challenging to observe.

Midterm and final exams were weighted at 50% of the overall grade, collectively measuring students’ mastery of the course material. For both the midterm and final exams, students were required to practice speaking on specific topics multiple times and record themselves on video. These video recordings were then submitted for evaluation. The assessments focused on aspects such as fluency, coherence, the story flow, and the use of vocabulary and expressions learned throughout the course, reflecting the students’ speaking progress over the semester. Reviewing the video recordings allowed the instructor to assess students’ speaking development and provided valuable insight into the effort and self-directed practice students had invested in their learning. Table 3 summarized the components of academic performance evaluation.

<Table 3>

Academic Performance Evaluation Components

Assignments, weighted at 20%, were excluded from the analysis as they were primarily used to encourage participation in the online courses rather than to measure content mastery. The assignments were designed to incentivize students to engage with the course material, but their focus was on timely participation rather than assessing students’ English proficiency or content mastery. Consequently, the points awarded for completing assignments reflected engagement rather than the development of English language skills. Therefore, it was deemed inappropriate to include assignments in the analysis of content mastery. By focusing on attendance and exams, the study provided objective measures of both engagement and learning outcomes. This grading scheme allowed for a clear comparison of student engagement and learning outcomes across the two formats, focusing on attendance as an indicator of participation and exams as a measure of content mastery.

3.2.3. Instructional Formats and Learning Environments

The study included two distinct online learning environments that evolved as remote education shifted during and after the COVID-19 pandemic. The real-time and pre-recorded formats each provided unique structures for student engagement and interaction, allowing this study to explore how synchronous and asynchronous learning settings impacted students’ emotional responses and academic outcomes.

In contrast, as the educational landscape shifted back toward traditional face-to-face learning in the post- COVID era, the researcher transitioned to 100% pre- recorded video classes using the university’s Cyber Class platform. This asynchronous format allowed students to access lecture content at their convenience. While pre-recorded online formats offer flexibility, they lack the direct, interactive components found in real-time online classes. Students’ engagement in this format was primarily asynchronous, with interactions generally limited to designated Q&A sessions and emails, potentially impacting their connection to the material and peers.

① Real-time Online Classes

Real-time online classes, widely adopted during the pandemic, facilitated synchronous interaction among professors, students, and peers through platforms like Microsoft Teams and Zoom. These classes were conducted in 2020 and prioritized active student engagement. The real-time format provided immediate feedback, fostering a sense of immediacy that allowed students to address questions and concerns promptly. Students could interact directly with the instructor and peers through video sessions, live chats, and shared document spaces for collaborative work. This setup was designed to create a sense of community and encourage active participation, enabling dynamic and engaging learning experiences.

② Pre-recorded Video Classes

In 2023, the study transitioned to 100% pre-recorded video classes using the university’s Cyber Class platform. This asynchronous format allowed students to access recorded lectures and course materials at their convenience, offering flexibility in their learning schedules. However, unlike real-time online classes, pre-recorded formats lack direct, real-time interaction. Interaction was primarily asynchronous, limited to designated Q&A sessions, discussion forums, and email communication. While this setup provided resources for self-paced learning, the lower immediacy of interaction required students to take more initiative in their learning experiences. This format potentially impacted students’ connection to the material and peers, as engagement depended largely on individual effort.

③ Course Topics and Activities

To maintain a controlled comparison, both formats followed a standardized 15-week curriculum, as shown in Table 4, which details the weekly topics covered across both learning environments. The content ranged from introductory activities (e.g., self-introductions, describing people and places) to more advanced topics on personal goals and plans, culminating in midterm and final exams. By adhering to a consistent curriculum and grading structure, the study ensured that any differences in student experiences and outcomes could be attributed to the instructional format rather than content variations.

<Table 4>

Weekly Course Topics

Table 4 outlines the weekly course topics, which were designed to progressively build students’ English language skills in comprehension, writing, and speaking. Each week, students engaged in structured activities aimed at enhancing their understanding of the topic while encouraging self-directed learning. These activities were tailored to the unique features of each instructional format (real-time vs. pre-recorded) while maintaining consistency in content delivery and learning objectives.

In the first week, students were introduced to the course structure, learning objectives, and assessment methods. In the second week, to establish a foundation, students wrote short paragraphs introducing themselves, focusing on basic vocabulary and sentence structures, and practiced reading their introductions aloud. Subsequent weeks focused on specific themes, such as describing people and places, leisure activities, and goals.

The weekly activities included 1) reading example texts, where students engage with topic-relevant materials to familiarize themselves with vocabulary, expressions, and sentence patterns. These texts served as models to guide their writing and speaking exercises; 2) writing practice, where students wrote short descriptive paragraphs based on the weekly topic. These assignments were designed to encourage active participation without being included in the performance evaluation, allowing students to focus on skill-building without performance pressure; 3) oral practice and delivery, where students practiced reading their written work aloud to improve pronunciation, fluency, and confidence in speaking; and 4) recording and submitting short videos based on their written work, providing opportunities for self-assessment while allowing the instructor to provide targeted feedback on pronunciation, intonation, and clarity. Reviewing these submissions also enabled the instructor to evaluate students’ efforts and engagement with class activities.

In both real-time and pre-recorded online English courses, students were encouraged to adopt self-directed learning strategies. These included regular attendance in online classes, consistent independent practice, and timely submission of writing and speaking assignments. These strategies were intended to help students reinforce their language skills and maintain consistent engagement across both online course formats.

3.3. Data Analysis

To assess how students’ initial emotional reactions towards the online class format related to subsequent engagement and performance, the study employed a multi-step data analysis using Excel 2019 and SPSS Statistics 25.

① Data Cleaning and Preparation

Survey responses and academic performance data were collected and cleaned to ensure accuracy. Missing values were identified and excluded from the analysis to maintain data integrity. Outliers were reviewed to determine their impact and were excluded if deemed non-representative of the participant groups.

② Descriptive Statistics

Measures of central tendency (mean and median) and variability (standard deviation) were calculated for each survey item and academic performance metric. This step allowed for a comprehensive understanding of the data’s distribution and provided insights into general patterns within the emotional responses and academic performance metrics.

③ Reliability Analysis

To ensure the internal consistency of the survey and performance metrics, Cronbach’s alpha was calculated. The reliability of the survey instrument and academic performance metrics was assessed, with values of 0.74 for the survey (11 items) and 0.69 for the academic performance metrics (2 items). Cronbach’s Alpha values above 0.7 are generally considered acceptable for exploratory studies, ensuring the reliability of the measures used in this research. These reliability values indicate that the survey items and performance metrics consistently capture the intended constructs, supporting the validity of the subsequent analyses.

④ Comparative Analysis

To evaluate differences between the two instructional formats, independent samples t-tests were conducted. For emotional responses, each survey item was analyzed individually to identify statistically significant differences in anxiety, confidence, and willingness to engage in EFL learning between the two groups. Additionally, for academic performance, attendance rates and exam scores were compared across the two formats to examine distinctions in engagement and learning outcomes. This analysis helped reveal any significant distinctions in academic engagement and achievement related to the instructional approach, further contributing to understanding how each format influenced student learning outcomes.

⑤ Correlation Analysis

To explore the relationship between emotional responses and academic outcomes, a correlation matrix analysis was conducted. This analysis examined the strength and direction of the relationships between three key emotional factors (anxiety, confidence, and willingness) and two academic performance metrics (attendance and exam scores). Pearson’s correlation coefficient (r) was used as the statistical measure, which ranges from -1 to +1, indicating the strength and direction of the relationship. The correlation matrix provided valuable insights into the interplay between students’ initial emotional responses and their subsequent engagement and performance in the course. By identifying these relationships, the correlation analysis complemented the comparative findings from descriptive statistics and t-tests, offering a nuanced understanding of how emotional factors influence engagement and outcomes across both online formats.

Emotional responses, including anxiety, confidence, and willingness, were analyzed to investigate variations by online format, while correlations between survey responses and academic metrics (attendance and exam scores) were assessed to understand whether preliminary emotional states could predict engagement or performance trends. This approach offers a nuanced understanding of how synchronous and asynchronous learning environments impact student experiences in EFL education.

4.1. Survey Responses in Real-time and Pre-recorded Online Courses

This section examines students’ initial emotional responses to real-time and pre-recorded online courses, focusing on differences in anxiety, confidence, and willingness to engage in English learning. By capturing students’ emotional states early in the semester, the survey provides insights into how these factors may influence their choice of instructional format and subsequent engagement.

Table 5 presents the findings from the initial survey responses gathered from students in both real-time and pre-recorded online course formats. These data highlight student responses regarding anxiety, confidence, and willingness to engage in English learning, offering insights into the emotional and motivational factors that may influence students’ choice of online learning format prior to the start of the semester. The overall scores for anxiety, confidence, and willingness presented in Table 5, were calculated as the mean of specific survey items. These aggregated scores provide a comprehensive view of students’ initial emotional and motivational states across the two instructional formats.

<Table 5>

Survey data Analysis for Real-time and Pre-recorded Online Courses

The survey was conducted during the second to third weeks of the semester, aiming to capture students’ initial emotional states before they were significantly exposed to their respective instructional formats. This timing allowed for an exploration of potential pre-existing emotional factors, such as anxiety, confidence, and willingness, that might influence students’ choice of either real-time or pre-recorded courses. By distinguishing between emotional responses at this early stage, the study highlights the importance of addressing these factors when designing instructional strategies to optimize engagement in both real-time and pre-recorded settings.

① Anxiety

As shown in Table 5, there was a significant difference (p < .00) in overall anxiety between the two groups, with students in the pre-recorded format reporting higher levels of worry about making mistakes (p = .00) and feeling anxious when using English in class (p = .07). This pattern suggests that students with higher anxiety may be more inclined to choose the pre-recorded format, potentially preferring limited peer interaction and a more self-paced environment where they can avoid immediate social pressure or live feedback. These findings align with Horwitz et al. (1986) and Matsuda & Gobel (2004), who identified anxiety as a significant barrier to active participation in language learning. The data suggest that students with elevated anxiety levels may prefer pre-recorded formats, likely due to reduced social pressure and the flexibility to learn at their own pace without the immediacy of live feedback.

② Willingness

There was a significant difference in students’ overall willingness to continue learning English, with those in the real-time classes demonstrating a greater desire to engage in English learning (p = 0.01). Students in the real-time courses also showed a higher willingness to continue using English outside of class (p = 0.05). This trend, though modestly significant, reflects increased motivation among students in the real-time format, which may be attributed to the interactive, community-oriented nature of the live classes. The preference for real-time classes may suggest that these students value a learning environment where they can use English actively, benefit from immediate feedback, and develop language skills in a collaborative setting. This supports findings by MacIntyre et al. (1998), who identified willingness as a key predictor of sustained language use and active participation in language learning contexts.

③ Confidence

Confidence showed no significant difference between the real-time and pre-recorded learning formats (p = 0.41). This indicates that both instructional formats had similar effects on students’ confidence levels, suggesting that the format did not significantly impact their overall confidence in learning English. However, students in real-time classes consistently reported higher mean values for confidence across skills like writing, speaking, and pronunciation. This trend suggests that students choosing real-time online classes may prefer environments where they receive direct feedback and interaction, which can contribute to building confidence over time. In contrast, students in pre-recorded classes may prioritize flexibility to learn at their own pace, even if it means sacrificing immediate feedback.

The survey data suggest that students in real-time online courses generally had a more positive outlook toward learning English, as reflected in lower anxiety levels and greater willingness to use English beyond the classroom. These findings highlight the role of emotional factors in shaping students’ preferences for specific online learning formats. Students who prefer a supportive, interactive environment tend to select real-time classes, which provide opportunities for direct engagement and feedback. The reduced anxiety and increased willingness observed in real-time classes may reflect the collaborative benefits of synchronous learning environments. Altınay (2017) emphasized the importance of structured peer learning and assessment in online environments for fostering student engagement. This finding aligns with this study’s observation that real-time formats, by enabling immediate interaction and feedback, create a more collaborative and engaging atmosphere for learners. Meanwhile, students who favor more flexibility and less immediate interaction may choose pre-recorded courses, potentially due to higher initial anxiety or a preference for self-paced learning.

Overall, these results suggest that students’ initial emotional states influence their selection of online learning formats. Future research could further explore how additional factors, such as self-efficacy, prior online learning experiences, and individual learning styles, interact with emotional responses to shape students’ preferences and engagement in online EFL contexts.

4.2. Comparison of Academic Performances in Real-time and Pre-recorded Online Courses

This section examines differences in academic performance between real-time and pre-recorded online English courses, focusing on two key metrics: attendance and exam scores. Assignment scores are excluded from the analysis as they primarily assessed participation rather than content mastery, which could confound the results. To maintain consistency, Table 6 presents only attendance and exam performance. The academic performance metrics have been standardized to a 100-point scale to facilitate a clear and consistent comparison across both real-time and pre-recorded course formats. The exam scores are based on the average of the exams, weighted at 50% of the total grade, and the attendance score is weighted at 30% of the total grade.

<Table 6>

Comparison of Academic Performance: Real-time vs. Pre-recorded Online English Courses

The analysis in the Table 6 reveals no statistically significant difference in exam scores between students in the real-time format (M = 80.53, SD = 19.19) and those in the pre-recorded format (M = 81.65, SD = 15.62), with p = .72. This indicates that both instructional formats were similarly effective in supporting students’ acquisition of course material. Additionally, a small percentage of students in each group scored below 60% (7.55% in real-time and 6.76% in pre-recorded), suggesting that neither format posed significant challenges to academic success. These findings imply that, in terms of knowledge acquisition and exam performance, both real-time and pre-recorded formats provide adequate support for student learning.

Conversely, attendance rates showed a significant difference (p = .01), with students in real-time classes achieving higher average attendance (M = 95.41%, SD = 8.70) compared to those in pre-recorded classes (M = 90.44%, SD = 12.20). These findings align with participation rates from the initial survey, where response rates were higher for real-time students (84.9%) compared to pre- recorded students (55.4%). The higher attendance in real-time classes may reflect the motivational benefits of synchronous formats, which foster a sense of accountability and engagement similar to traditional face-to-face classes. These findings are consistent with the survey data discussed in Section 4.1, where students in real-time classes reported greater willingness to engage with English learning.

To further explore the relationship between emotional factors and academic outcomes, Table 7 presents the correlation matrix of students’ anxiety, willingness, and confidence with their attendance and exam scores. The correlation matrix, showing the relationships between students’ emotional factors (anxiety, willingness, and confidence) and their academic performance (attendance and exam scores), is as follows.

<Table 7>

Correlation between Emotional Factors and Academic Performance

Table 7 highlights how emotional factors are linked to academic outcomes in the study. However, none of correlations are statistically significant, as all p-values exceed .05. In education research, correlations below 0.20 are generally considered weak. Given that these correlations are weak and statistically insignificant, they should not be interpreted as evidence of causal relationships. This weak correlation suggests that other influencing factors, which were not measured in this study, could be influencing. As mentioned in 4.1, students in pre-recorded classes reported higher anxiety levels, which correlates weakly with attendance (r = 0.12) and negatively correlates with exam scores (r = -0.08). This suggests that students with higher anxiety may reduce both participation and exam performance. Conversely, real-time class students exhibited higher willingness. The correlation matrix supports this, showing a positive correlation with both attendance (r = 0.15) and exam scores (r = 0.09), indicating that greater willingness leads to better engagement and academic outcomes. While confidence levels were similar across both groups in 4.1, the correlation matrix shows that confidence positively correlates with attendance (r = 0.22) and exam scores (r = 0.11). This suggests that higher confidence contributes to better engagement and performance, even though the initial differences were not statistically significant. Confidence, therefore, plays an important role in fostering engagement and academic success, regardless of the online instructional format. This finding aligns with prior studies (Horwitz et al., 1986; MacIntyre et al., 1998), highlighting the role of confidence in language learning and engagement. However, the weak correlations imply that while emotional factors may have some impact on academic outcomes, other unmeasured variables are likely influencing the results, and these relationships should not be interpreted as direct causal links. Future study should consider additional variables or use longitudinal data to better understand these dynamics.

Additionally, the significant difference in attendance, despite comparable exam scores, suggests that while both formats effectively support knowledge acquisition, real- time classes may foster stronger engagement. This may be attributed to the immediate interaction and structured participation opportunities in synchronous formats. These findings align with previous studies (e.g., Jiang et al., 2023), emphasizing the role of interaction in promoting motivation and sustained engagement.

By focusing on students’ initial emotional reactions and their correlations with engagement patterns, this study extends the author’s previous works, providing a nuanced understanding of how instructional formats influence academic performance and emotional engagement.

4.3. Overall Discussion

The findings provide insights into how real-time and pre-recorded online English courses influence student engagement and performance, focusing on the interplay between emotional responses and academic outcomes. One of the key findings is that attendance rates were significantly higher in real-time classes compared to pre-recorded classes. This highlights the motivational and structural advantages of synchronous learning environments. The accountability and immediacy of interaction in real-time classes create a stronger sense of community and engagement, motivating students to participate consistently.

In contrast, pre-recorded courses, despite offering greater flexibility, demonstrated lower attendance rates. This result underscores the difficulties of self-directed learning: students may lack the intrinsic motivation or self-regulation required to engage consistently without the structure provided by scheduled interactions. This aligns with prior studies (e.g., MacIntyre et al., 1998; Peng & Woodrow, 2010), which emphasized the importance of emotional and motivational factors, particularly for students in less interactive environments.

The finding that attendance rates were higher in real-time classes than in pre-recorded classes highlights the difficulties some students face with self-directed learning. While pre-recorded classes are designed to support learners, who benefit from flexible scheduling and self-paced study, the results suggest that such flexibility may inadvertently lead to lower motivation and engagement for some students. This finding contrasts with the initial assumption that pre-recorded courses would naturally support higher attendance due to their adaptable format. Moreover, the absence of direct interaction in pre-recorded classes might contribute to feelings of isolation, reducing the motivation to engage regularly.

Emotional factors such as anxiety, confidence, and willingness played a crucial role in shaping students’ preferences for online learning formats. Specifically, students with higher anxiety levels may prefer pre- recorded courses, as these formats offer a more self-paced environment with less immediate interaction, potentially reducing feelings of pressure. In contrast, students with greater confidence and a higher willingness to engage may thrive in real-time settings, where structured interaction and immediate feedback help build confidence and foster engagement. These findings suggest that emotional responses significantly influence students’ decision-making when choosing between real-time and pre-recorded formats.

However, traditional in-person classes may offer even greater opportunities for immediate interaction and real- time support, particularly in language learning contexts where live, dynamic exchanges play a critical role in language acquisition. Despite the advantages of real-time online learning, pre-recorded video courses continue to grow in popularity, both in educational institutions and through independent online platforms that provide EFL instruction to a wide audience.

The study acknowledges several limitations that should be considered when interpreting the findings. First, the differences in survey response rates between the two groups (real-time: 84.8%; pre-recorded: 55.4%) may affect the representativeness of the data. The voluntary nature of the survey introduces potential self-selection bias, as students who chose to participate may not fully represent the broader student population. Additionally, the data collection spanned two distinct timeframes (2020 and 2023), which could introduce external factors unrelated to the online instructional format, such as changes in student attitudes or external circumstances between these years.

Despite these limitations, the methodology employed is crucial for exploring the interplay between instructional formats, emotional responses, and academic outcomes. By addressing a gap in existing research, the study provides valuable insights into how different online formats shape student engagement and performance, offering practical implications for optimizing online EFL instruction in diverse educational settings.

Given this trend, future research should explore strategies to enhance student engagement and improve attendance and participation rates in pre-recorded video format courses, particularly in college general education curricula. These courses offer flexible scheduling but often lack structured, interactive components. The results suggest a need to address motivational challenges in pre-recorded formats by incorporating strategies that foster engagement. Reeves (2000) emphasized the importance of alternative assessment methods in online learning to meet diverse student needs. Applying such approaches in pre-recorded courses—such as integrating formative assessments and interactive tasks—may mitigate feelings of isolation and support sustained motivation.

For example, adjusting the grading structure to increase the weight of attendance and participation might effectively motivate students in pre-recorded courses. Establishing a clear attendance policy, where students must meet a minimum attendance threshold, could also help prevent disengagement. Setting a requirement for students to attend a certain number of weekly sessions or complete a designated number of interactive tasks each month may help maintain regular engagement. Lastly, a stricter policy―such as assigning an automatic failing grade (F) if absences exceed a specific limit―could also be tested to evaluate its effectiveness in promoting attendance and commitment in pre-recorded settings.

These suggestions and strategies highlight the importance of balancing flexibility with interaction and accountability in pre-recorded online learning. Such approaches not only cater to diverse learner needs but also ensure that pre-recorded formats remain an effective component of online education in higher education settings.