Research introduction

The integration of the internet, big data, artificial intelligence, blockchain, and 5G has given rise to diverse and creative approaches in higher education. Many countries have recognized the pressing need to reform their educational practices within these new paradigms, prompting national strategies that intertwine digital technologies with pedagogical models. Lee and Means emphasize that technology-enhanced environments produce more meaningful academic experiences when instructional methods are carefully synchronized with technological tools1,2, while Singh argues that learner engagement is heightened by systems allowing flexible, customized interactions—perspectives often overlooked in purely traditional classrooms3.MOOCs began expanding worldwide around 2012, and China swiftly aligned with this trend by initiating the construction of online open courses, as documented in the “10-Year Development Plan for Education Informatization 2011–2020”4. In 2015, the Ministry of Education issued guidelines to strengthen the design, use, and oversight of online open courses in higher education institutions, followed in 2017 by the “13th Five-Year Plan for National Education Development,” which highlighted “Internet + Education” as a crucial measure for integrating information technology into teaching. The subsequent “Education Informatization 2.0 Action Plan”5,6 introduced in 2018 sustained this momentum, underscoring the importance of leveraging cutting-edge tools and methodologies to support instructional innovation7. Government-led initiatives in the United States also propelled online education forward, as reflected in the third report on the panoramic view of online teaching development in the country—published by Quality Matters (QM) and Eduventures Research in 2019—which emphasized the importance of comprehensive data, authenticity, and reliability in reflecting emerging trends and informing educational practice8,9.

Despite these large-scale policy efforts, a considerable body of academic research highlights the continued need for empirical evaluations to clarify how technology-based instruction affects learner success. For example, Nkomo found that digital platforms, even when backed by robust policy support, can remain underutilized if students’ actual modes of engagement are not carefully examined10. In this study, student learning engagement, within the context of learning engagement, refers to student participation, measuring the time and effort learners invest in their learning and teaching, and its alignment with school or societal expectations11. This form of engagement can be understood through multiple dimensions, encompassing behavioral, cognitive, emotional, and interactive aspects. The “47th Statistical Report on China’s Internet Development” further indicates that by December 2020, online teaching and mobile online teaching reached 34.60% of Chinese netizens, with a significant contingent enrolled in university-level computer science programs12,13. Such data underscore the unstoppable expansion of online modalities in higher education, yet they also illustrate the importance of blending policy frameworks with scholarly evidence to ensure that digital technologies genuinely enhance the learning experience. Recent analyses in global reports continue to underscore the importance of data-based, interactive approaches in online education. Kuh, a scholar in the field of student engagement, contends that the commitment and time students dedicate to academic tasks deeply influence their learning achievements14. Autonomy and convenience are among the attributes prompting many undergraduates to embrace online learning, although the extent of their motivation and involvement often varies. Researchers such as Fredricks, and Wang, who proposed a multidimensional conception of engagement, emphasize behavioral, emotional, and cognitive aspects, each shaped by learners’ perceptions of institutional or platform-based support15,16. Harkin notes that learners may struggle to sustain engagement if they lack a sense of ownership or face persistent ambiguity in monitoring progress17. Those who exhibit a high level of interaction and self-regulation in online settings tend to achieve stronger outcomes, while those with diminished engagement risk difficulties in maintaining goals.

Although online learning engagement has attracted increasing scholarly interest, there remains a notable scarcity of focused research on undergraduate computer science students13. These learners often work within specialized digital environments and rely on advanced technological competencies distinct from other populations, yet their particular challenges and learning processes have not been sufficiently explored. Amerstorfer warns that isolating any one of these components fails to capture the broader tapestry of factors that collectively shape learner engagement18. Examining undergraduate computer science students is especially vital given their deep-rooted familiarity with digital tools and the field’s emphasis on acquiring advanced, technology-driven competencies. By focusing on a population whose educational experiences are closely tied to interactive, collaborative, and hands-on methods, The present study addresses this gap by incorporating multiple dimensions, including fundamental student characteristics (gender, family background, major), personal traits, teacher-related factors, peer dynamics, and learning-environment conditions. This comprehensive framework is pivotal for capturing how these elements intersect to shape students’ online engagement and for informing pedagogical choices that acknowledge a wide spectrum of influences. Through examining these convergent factors, The findings are intended to provide invaluable theoretical insights and practical guidance for university online instruction.

Theoretical foundations and research framework

Theoretical foundations

TPACK model

The TPACK model, standing for Technological Pedagogical Content Knowledge, profoundly elucidates the integrative knowledge system required by teachers in the digital age. This theory emphasizes that technology is not merely an auxiliary tool but a necessary component constituting effective teaching practice. Shulman’s early work laid the foundation for understanding teachers’ pedagogical content knowledge (PCK)19, upon which Mishra and Koehler further developed the TPACK model, emphasizing the dynamic interaction between technology, content, and pedagogy. Specifically, the TPACK model comprises seven elements: content knowledge (CK), pedagogical knowledge (PK), technological knowledge (TK), and their intersections, namely technological content knowledge (TCK), technological pedagogical knowledge (TPK), pedagogical content knowledge (PCK), ultimately forming TPACK20,21. The core concept of this model lies in the understanding that effective technology integration is not simply applying technology to teaching but deeply comprehending how technology shapes and facilitates the teaching and learning processes of specific subjects. In educational practice, teachers need to carefully select and utilize technology to support specific teaching activities and ultimately promote student learning22. In this study, TPACK offers a foundational framework for the online teaching model. It relates directly to the research objective of examining factors influencing students’ online learning engagement by spotlighting how technological selections shape the efficacy of content delivery. The technology component in TPACK guides decisions regarding learning platforms, while the pedagogical and content dimensions interact to create opportunities for meaningful online participation.

Behaviorist learning theory

Behaviorist learning theory posits that human thought and behavior result from “stimulus–response” connections formed through interactions with the external environment. Represented by Watson and Skinner, behaviorists emphasize the decisive role of the external environment in shaping behavior, believing that almost all behaviors can be acquired through learning23,24. In educational practices, the principles of behaviorism are reflected in teachers shaping and adjusting student behavior through specific instructional interventions, aiming to create a positive learning environment, reinforce appropriate behaviors, and eliminate inappropriate ones25. According to LaBrot, behaviorist theory emphasizes that stimuli trigger responses, and the environment determines learning; programmed instruction follows the principle of small steps, establishing teacher and peer assessment systems to provide learners with timely feedback on learning outcomes, focusing on the learner26. Thorndike’s law of effect and law of exercise also provide important support for this theory. The law of effect states that behaviors leading to satisfactory outcomes are more likely to be repeated, while the law of exercise emphasizes the importance of repeated practice in consolidating learning outcomes27. Consequently, factors such as the learner’s environment, the structure of the learning plan, relevant learning objectives, the evaluation of learning outcomes, and the learners themselves significantly influence learner engagement. In the present investigation, behaviorist concepts inform the assessment of how stimuli—ranging from online course design to peer interactions—elicit responses such as consistent class attendance, assignment completion, and interactive communication. The research focus on behavioral, cognitive, emotional, and interactive engagement relates to these behaviorist mechanisms, as learners’ online activities and responses are inextricably linked to the environmental and instructional cues provided by their teachers, peers, and course structure. Structured reinforcement through prompt feedback may heighten students’ willingness to participate, particularly when such stimuli are perceived as supportive and clear.

Learning participation theory

Learning participation theory argues that to comprehensively understand human learning activities, not only should the natural attributes and tendencies of learning and the application of learning technologies and networks be considered, but also the social attributes of learners and the continuously shortening half-life of knowledge. Lave and Wenger’s situated learning theory laid an important foundation for this, emphasizing that learning is a process of participating in communities of social practice28. Koob further points out that this theory emphasizes relying on learning technologies to build personal knowledge networks, learning networks, and social networks, enhancing learning outcomes through interactions among learning members29. Wenger’s concept of Communities of Practice is also closely related, emphasizing that learning occurs through joint participation, interaction, and knowledge sharing. Therefore, the use of online learning systems, the understanding of knowledge learning, and peer communication are significant manifestations of learning participation30. This study draws upon Learning Participation Theory to illuminate how students develop and sustain engagement within virtual environments. The emphasis on personal knowledge networks resonates with the research objective of mapping how self-factors, teacher influences, and peer interactions collectively generate momentum in online learning. This perspective clarifies why various participants’ roles, digital resources, and social networks become instruments for deepening student engagement, particularly when learners rely on technology to bridge temporal and geographical distances.

Situated cognition learning theory

Situated cognition learning theory posits that individuals possess the ability to actively construct knowledge; learning is essentially a collaborative socialization process, an interaction between learners and situations, a progression from the periphery to the core, and an active exploration in real activities31. This theory emphasizes that knowledge is situated, that learning is closely linked to situational activities, and that learning should focus on the learner’s rational understanding of knowledge participation in various situations. Lambert’s research also indicates that action and cognition are situated and that the acquisition of knowledge and skills is inseparable from specific social and material environments32. Contrary to traditional cognitivist views, situated cognition theory argues that knowledge does not exist independently of context but is constructed and used within specific situations33. Due to the specificity of online teaching locations and organized activities in this study, the learning situation is particularly crucial. The online learning environment, interaction methods, and tools used all constitute specific learning situations, which profoundly affect students’ learning experiences and levels of engagement. The learning situation greatly enriches the specific manifestations of learning participation, hence it can be combined with learning situations to delineate dimensions, thereby more comprehensively examining students’ participation in online learning.

Research framework

Building upon the aforementioned theoretical foundations, this study primarily investigates students’ participation in online teaching modalities (behavioral Engagement, cognitive Engagement, emotional Engagement, and interactive Engagement). It examines the relationships between antecedent variables (such as gender, home location, major, and Daily online learning duration) and independent variables (like individual factors, teacher factors, Peer factors, and environmental factors). An influential factors research framework for online learning has been established, as depicted in Fig. 1.

Fig. 1
figure 1

Research framework for factors influencing online learning.

Full size image

Antecedent variable

Gender: Gender refers to the biological characteristics of humans as male or female. In this study, males and female refer to the biological characteristics of the survey respondents. Home Location: This term denotes the geographical location where one typically works and resides34. In this context, it signifies the place where the student’s family resides, primarily categorized into urban, town, and rural areas. Major: It pertains to the academic categories set up in universities based on the needs of professional specialization in society35. In this study, the majors include Software Engineering, Computer Science and Technology, Network Engineering, and Information Security. Daily online learning duration: This measures the amount of time learners spend on online learning daily36. In this research, it is categorized as 0–2 h, 2–6 h, and more than 6 h daily.

Independent variable

Influencing Factors: These are reasons or conditions that determine the success or failure of something. In scientific experiments, factors or reasons that impact the experiment’s indicators are also termed as factors37. In this study, these denote the reasons or elements affecting students’ participation in online learning, understood through indicators such as individual attributes, teacher characteristics, peer influences, and environmental conditions. Self factors: This refers to the contribution of an individual’s efforts to the degree of participation in online teaching and learning modes38. In this research, it is the first metric of influencing factors, measured by elements like course interest, personal control, and individual information literacy. Teacher Factors: This pertains to the contribution of teachers to university students’ participation in online learning39. In this study, it is a metric of influencing factors, determined by the teacher’s attitude towards online teaching, their information literacy, and preparedness for online instruction. Peer factors: This denotes the contribution of peers to university students’ participation in online learning40. In this study, it is a metric of influencing factors, ascertained through peer interactions, collaborative learning, and peer assessments. Environmental Factors: This refers to the environment’s contribution to the participation level of university students in online learning41. In this research, it’s a metric of influencing factors, determined by aspects like the availability of robust internet equipment, whether the home environment supports student learning, and whether the school offers convenience for online studies.

Dependent variable

Student Learning Engagement: This refers to the “student engagement” in the context of “learning engagement”, and the measure of the time and effort learners invest in their own learning and effective teaching activities, and whether this time and effort align with the expectations of the school or society42. In this study, it is understood through behavioral, cognitive, emotional, and interactive engagement. Behavioral Engagement: Refers to the observable behaviors of students actively participating in classroom learning activities43. In this study, it is determined by classroom interaction, assignment submission, and active participation in online tests. Cognitive Engagement: Refers to a student’s commitment to learning, reflecting the cognitive and psychological efforts of the learner44. In this study, it is determined by the depth of student thinking, the quality of student thought, and reflection on learning materials. Emotional Engagement: Refers to the emotional reactions and attitudes of learners towards learning content and teaching activities45. In this study, it is determined by student satisfaction with the course, student satisfaction with teaching activities, and student satisfaction with their learning outcomes. Interactive Engagement: Refers to the interactions between students and teachers, students and peers, students and materials, and students and technology46. In this study, it is determined by communication interactions between students and teachers, cooperative learning among students, student feedback on materials, and student feedback on the use of technology.

Research questions

This study, based on questionnaire surveys and in-depth interviews, focuses on the current status of online learning and its influencing factors among undergraduate students majoring in computer science in China. Through thorough analysis, it aims to gain insights into the impact and significance of various factors on students’ online learning engagement. Combining the characteristics and background of higher education in China, it further understands the specific situation and needs of computer science undergraduates’ online learning. The main questions of this study are:

  1. (1)

    Are there significant differences and relationships in students’ online learning engagement (encompassing behavior, cognition, emotion, and interaction) across personal characteristics such as gender, Home location, major, and online learning duration?

  2. (2)

    Are there significant differences and relationships in students’ perceptions of Self factors, teachers, peers, and the learning environment across personal characteristics such as gender, Home location, major, and daily online learning duration?

  3. (3)

    Are there significant relationships between the multi-dimensional factors associated with student engagement (such as Self factors, teachers, peers, and the learning environment) and students’ online learning engagement (encompassing behavior, cognition, emotion, and interaction)?

Research design

Research subjects

This study aims to investigate the current status, influencing factors, and related optimization strategies of online learning among university students in China, primarily targeting undergraduate students from H University located in the central region of China. These students participated in the “Principles of Computer Organization” course through online learning mode in the 2022 academic year. A total of 800 students undertook the online instruction of this course that academic year, comprising 280 majoring in Software Engineering, 220 in Computer Science and Technology, 160 in Network Engineering, and 140 in Information Security. This data reflects the course’s popularity and acceptance among students from different majors, as the course is typically offered in the second and third years of university, with the participating students being predominantly sophomores and juniors. The online teaching model for this course was designed as an integrated synchronous-and-asynchronous learning environment. Teachers delivered real-time (synchronous) lectures to students, followed by assignments and preview materials (asynchronous) made available on the “SuperStarLearn”platform. This approach is similar to a MOOC format, wherein both synchronous live sessions and asynchronous tasks are offered through digital platforms47,48. To ensure the accuracy and representativeness of our research, we employed a simple stratified random sampling method and selected 267 students as the research sample. Using Slovin’s formula (n = \frac{N}{1 + Ne^2}), ensures that within the given error range (e = 0.05), the sample is sufficiently representative49. According to this method, the sample sizes for Software Engineering, Computer Science and Technology, Network Engineering, and Information Security are 98, 80, 44, and 45 students, respectively. In terms of gender, out of the 267 respondents, 142 (53%) were female, and 145 (47%) were male. The age of the participants primarily ranged from 18 to 25 years old. Through such filtering and sampling, we can ensure the reliability and validity of our research findings, providing empirical evidence for further discussions on the current status, influencing factors, and optimization strategies for online learning among computer science undergraduates.

Research tools

The core data collection instrument employed in this study is a questionnaire devised to capture the multifaceted characteristics and influencing factors of university students’ online learning. Its development relied on insights drawn from a previously validated survey50, which was recognized for its rigorous design and thorough consideration of both internal and external elements. The original scale identified learning readiness, learning motivation, online learning environment, and organizational management of the learning process as crucial dimensions, reflecting its emphasis on a broad range of factors that shape online engagement. In adapting this survey for the local context, particular attention was given to preserving the core constructs while refining language and item relevance to fit the specific characteristics of our research participants. The final version of the questionnaire in this study consists of three major components. The opening section addresses students’ demographic background by collecting information on gender, home location, academic major, and frequency of online engagement. Subsequent items examine engagement in greater detail through four dimensions—behavioral, cognitive, emotional, and interactive—which collectively capture the intensity and quality of online participation. Another portion explores the wide-ranging factors that influence students’ participation patterns, focusing on the learner as an individual, instructors, peers, and the learning environment. Both the engagement and influencing-factor components employ a five-point Likert scale, inviting students to rate their level of agreement from “Strongly Disagree” through “Strongly Agree.” Adaptation of the survey followed a careful process involving language revisions and minor structural changes, ensuring alignment with the local learning context. This approach was guided by the theoretical premise that students’ readiness, intrinsic motivation, contextual supports, and organizational features of the curriculum determine key aspects of online learning behaviors. Evidence of strong reliability and validity in prior studies further justified the decision to incorporate this scale. A pilot test conducted with a smaller group of respondents confirmed the questionnaire’s suitability, leading to final refinements that maintained conceptual consistency across dimensions. In order to ascertain robust psychometric properties, the study team performed reliability and validity checks following data collection. As shown in Table 1, Cronbach’s α for each dimension exceeded 0.7, reflecting high internal consistency across the adapted survey items. The KMO values were above 0.9, and the Bartlett’s test of sphericity indicated statistical significance at p < 0.001, suggesting that the data were appropriate for factor analysis. The cumulative variance contribution surpassed 70% for each scale dimension, revealing a solid factorial structure aligned with the theoretical framework that guided item construction.

Table 1 Reliability and validity analysis of online learning engagement and influencing factors scale.
Full size table

Research process

The study utilized SPSS 25.0 software to conduct a comprehensive analysis of all collected data. Initially, frequency statistics and percentages were employed to describe and summarize the basic characteristics of students, such as gender, Home location, major, and the duration of their online learning participation. Subsequently, t-tests and Analysis of Variance (ANOVA) were used to determine whether there are significant differences between students’ varying characteristics (like gender, Home location, major, and online learning engagement duration) and their online learning engagement and its influencing factors. Then, the Pearson r or Pearson correlation coefficient was employed to explore the linear relationship between students’ basic characteristics and online learning engagement and its influencing factors, as well as the linear relationship between online learning engagement and its influencing factors. Lastly, based on the research findings, suggestions for optimization strategies for student online learning were proposed.

Results analysis and discussion

Basic characteristics of students

As shown in Table 2, the participants are primarily female, with most residing in townships, and majoring in software engineering. Furthermore, most of them have a study participation time of 0–2 h in the online teaching mode. In terms of gender, out of 267 respondents, 142 (53%) are female, while 145 (47%) are male. In present-day China, the phenomenon of the number of female students exceeding that of males in general higher education institutions has become common. Research by Liao, Xiong, and Hu indicates that among students in Chinese normal schools, 65.3% are female, and only 34.7% are male51. Regarding home location, most (72%) reside in rural or township areas, while 28% live in urban areas. Professionally, results demonstrate that the majority of students major in software engineering (37%); 80 individuals (30%) are in computer science and technology; 44 (16%) in network engineering; and 45 (17%) in information security. Since China’s modernization, the demand for graduates in the computer domain has risen, as shown in Zhou’s study, the enrollment numbers in these four undergraduate majors: software engineering, computer science and technology, network engineering, and information security have all increased52. Additionally, concerning daily online study duration, most respondents (49%) have study times between 0–2 h; 116 students (44%) study between 2 and 6 h; while 19 students (7%) study for more than 6 h daily. Wang’s publication titled 'The Current Status and Influencing Factors of College Students’ Online Learning’ also confirms that online learning seems set to become a daily compulsory course for college students in the future53.

Table 2 Basic characteristics of students in online teaching mode.
Full size table

Analysis of online learning engagement

Through the use of Analysis of Variance (ANOVA), a significant analysis was conducted on students’ varying characteristics such as gender, Home location, major, and daily duration of online learning in relation to their online learning Engagement. As illustrated in Table 3 and Table 4, initially, when analyzing from the perspective of gender, the Behavioral Engagement results revealed no significant difference between males and females in online Behavioral Engagement (t = − 0.885, df = 265, Sig. = 0.377). This suggests that regardless of gender, students exhibit similar frequencies and durations in online activities. In terms of Cognitive, Emotional, and Interactive Engagement, gender likewise appeared as an insignificant influencing factor. Further analysis on other characteristics revealed that Home location (F = 0.197, Sig. = 0.821) and major (F = 0.575, Sig. = 0.632) had no significant relationship with Behavioral Engagement. This may indicate that these factors do not directly impact students’ frequency and duration of activities in an online Environmental context. However, there was a significant difference in terms of daily online learning duration (F = 10.056, Sig. = 0.000), suggesting that the length of online learning could be a crucial factor influencing Behavioral Engagement. Extended periods of online learning might imply greater student Interactive behavior, task submissions, and discussion participation. Moreover, data on Cognitive Engagement showed that Home location (F = 1.974, Sig. = 0.141) and major (F = 0.259, Sig. = 0.855) had no clear association with Cognitive Engagement. Yet, a prolonged duration of online learning Engagement had a significant impact on Cognitive Engagement (F = 7.620, Sig. = 0.001), implying that extended online learning durations might offer students more opportunities to delve deeper into understanding, reflecting upon, and mastering study materials. Furthermore, the Emotional Engagement data indicated a near significant relationship between Home location (F = 2.731, Sig. = 0.067) and Emotional Engagement, potentially suggesting that the family Environmental background might, to some extent, influence students’ Emotional Engagement. There was no evident relation between major (F = 1.522, Sig. = 0.209) and Emotional Engagement. However, the duration of online learning had a pronounced impact on Emotional Engagement (F = 5.676, Sig. = 0.004), denoting that the more time students spend in online learning, the more intense their Emotional response might be towards the content and the learning process. Lastly, data on Interactive Engagement showed no significant relationship between Home location (F = 1.329, Sig. = 0.267) or major (F = 1.843, Sig. = 0.140) and Interactive Engagement. A clear relation existed between online learning duration and Interactive Engagement (F = 7.039, Sig. = 0.001), suggesting that the longer students learn online, the more they might interact with Peers and Teachers, thereby enhancing their learning experience. Overall, it is evident that for undergraduate computer science students in China, the primary factor associated with their online learning Engagement is the duration of online learning, while factors like gender, Home location, and major appear to have negligible effects on online Engagement.

Table 3 Analysis of student gender characteristics in online learning engagement.
Full size table
Table 4 Analysis of basic student characteristics in online learning engagement.
Full size table

Using Pearson correlation analysis, as shown in Table 5, firstly, from a gender perspective, the data indicates that the correlation between gender and all types of learning engagement (such as Behavioral, Cognitive, Emotional, and Interactive) is relatively low, with correlation coefficient (r) values all below 0.1. Specifically, for Behavioral Engagement, the correlation coefficient for gender is 0.054 (p = 0.377). This aligns with the aforementioned Analysis of Variance (ANOVA) results, both suggesting that gender does not seem to be a significant influencing factor. Secondly, the analysis on Home location indicates that its correlation with various types of engagement is also relatively weak. Moreover, all \(p\) values are above 0.05, indicating its lack of statistical significance. Furthermore, when discussing the influence of academic majors on online learning engagement, the correlation also appears to be relatively weak. Although the primary target group of this research consists of computer science students, the data indicates that the academic background might not be the critical determinant of online learning engagement. This conclusion is consistent with the aforementioned analysis. However, the most significant finding is that the daily duration of online learning shows a positive and significant correlation with all types of learning engagement. For instance, the correlation coefficient for Behavioral Engagement is 0.264 (p < 0.000), for Cognitive Engagement it’s 0.231 (p < 0.000), for Emotional Engagement it’s 0.200 (p = 0.001), and for Interactive Engagement it’s 0.225 (p < 0.000). This suggests that the more time students spend in online learning, the higher their engagement in various aspects. This is consistent with previous analysis. Extended online learning durations possibly imply that students have more opportunities to delve deeper, participate in discussions, interact, and reflect upon the learning content.

Table 5 Correlation analysis of basic student characteristics in online learning engagement.
Full size table

Analysis of factors influencing online learning

As shown in Tables 6 and 7, firstly, from the perspective of gender, the data results of the Self factors indicate that the difference between males and females is not significant (t = 0.992, df = 256, Sig. = 0.322). This suggests that in terms of intrinsic motivation and drive, both males and females exhibit similar characteristics. Similarly, for Teacher, Peer, and Environmental factors, there are no apparent differences between genders, indicating that gender differences are not the key determinant affecting engagement and outcomes in online learning. This is in line with the contemporary notion of gender equality in education, emphasizing that both males and females have equal learning capabilities and opportunities. Furthermore, the data for the Home location feature shows that there are no significant differences in the influence of Self factors (F = 1.013, Sig. = 0.364), Teacher factors (F = 0.302, Sig. = 0.740), Peer factors (F = 0.787, Sig. = 0.456), and Environmental factors (F = 1.346, Sig. = 0.262) based on home location. This could be attributed to the widespread accessibility and advancement of technology, allowing students, regardless of their location, to access relatively uniform learning resources and support. In terms of professional features, the data indicates a significant relationship between Teacher factors and the student’s major (F = 4.259, Sig. = 0.006). This might be because different majors have varying teaching modes or different needs for teacher support. For instance, some disciplines might require more hands-on guidance, thereby increasing students’ dependence on teachers. However, there are no significant differences in the influences of Self factors (F = 1.462, Sig. = 0.225), Peer factors (F = 2.208, Sig. = 0.088), and Environmental factors (F = 2.712, Sig. = 0.045) based on professional features. Lastly, in terms of daily online learning duration, the data clearly delineates its relationship with online learning influencing factors. Significant differences are observed in the Self factors (F = 9.901, Sig. = 0.000), Peer factors (F = 5.452, Sig. = 0.005), and Environmental factors (F = 6.294, Sig. = 0.002), but not in Teacher factors (F = 1.117, Sig. = 0.329). This implies that the longer students engage in online learning, the stronger their intrinsic motivation, interaction with peers, and adaptability to the learning environment may become. This aligns with traditional learning theories suggesting a relationship between input and output. However, of note, in the current rapidly advancing technological landscape, especially within computer science disciplines, students have cultivated a culture of autonomous learning. They are more reliant on online resources, such as search engines, forum discussions, and open-source materials, potentially reducing direct dependence on teachers. This not only reflects an enhancement in students’ independent learning abilities but also implies that teachers in online education modes need to seek new teaching strategies and redefine their roles.

Table 6 Analysis of students’ gender characteristics in online learning influencing factors.
Full size table
Table 7 Analysis of students’ basic characteristics in online learning influencing factors.
Full size table

As illustrated in Table 8, using the Pearson correlation coefficient analysis, several observations were made: Firstly, the data regarding the relationship between gender and various influencing factors indicated that the correlation between gender and Self, Teacher, Peer, and Environmental factors is not significant. Furthermore, the correlation coefficient ‘r’ values for these factors did not exceed 0.1. Specifically, for the Self factor, the correlation coefficient regarding gender is − 0.061 (p = 0.322). This is consistent with previous analysis results, suggesting that the influence of gender on online learning factors is not pronounced. Whether male or female, their performance in this regard seems virtually indistinguishable. Secondly, the data pertaining to Home location also revealed that its correlation with Self, Teacher, Peer, and Environmental factors is relatively weak. For instance, the correlation coefficients are as follows: − 0.017 (p = 0.788) for Self factor, − 0.021 (p = 0.530) for Teacher factor, − 0.050 (p = 0.411) for Peer factor, and − 0.059 (p = 0.334) for Environmental factor. This further validates that Home location is not a significant influencing factor in this study, implying that regardless of where students come from, their online learning performance and engagement levels remain relatively consistent. While analyzing the relationship between academic majors and online learning influencing factors, especially focusing on students majoring in computer science, the data suggested that the correlation with the Teacher factor was the most significant at − 0.160 (p < 0.05). The correlations with Self, Peer, and Environmental factors showed weaker associations. Lastly, data on daily online learning duration indicated a strong positive correlation with Self, Peer, and Environmental factors, but no correlation with the Teacher factor. Specifically, the correlation coefficients are 0.260 (p < 0.001) for the Self factor, 0.091 (p < 0.136) for the Teacher factor, 0.199 (p < 0.001) for the Peer factor, and 0.213 (p < 0.000) for the Environmental factor. This reiterates that students who spend more time online tend to perform better in online learning. They are more likely to engage deeply, interact actively, and achieve a better learning experience within the online environment.

Table 8 Correlation analysis of online learning influencing factors with basic student characteristics.
Full size table

Correlation analysis between online learning influencing factors and learning engagement

As shown in Table 9, the Self factors demonstrate a significant correlation with students online learning Engagement. Specifically, the correlation coefficient for Behavioral Engagement is 0.807 (p < 0.001), for Cognitive Engagement is 0.825 (p < 0.001), for Emotional Engagement is 0.797 (p < 0.001), and for Interactive Engagement is 0.857 (p < 0.001). This suggests that students intrinsic motivation, learning attitudes, and self-motivation play pivotal roles in online learning. Students self-driven initiative dictates their proactivity and efficiency in online learning. The high correlation for cognitive, emotional, and interactive engagements further corroborates the significance of individual attributes such as interest, motivation, and learning strategies in influencing their online learning experience. Next, the Teacher factors also present a noteworthy correlation with online learning Engagement. The correlation coefficient for Behavioral Engagement is 0.638 (p < 0.001), for Cognitive Engagement is 0.679 (p < 0.001), for Emotional Engagement is 0.717 (p < 0.001), and for Interactive Engagement is 0.707 (p < 0.001). This indicates that a teacher’s pedagogical methods, interactive strategies, and feedback mechanisms play a crucial role in stimulating students emotional engagement and establishing affective bonds among learners. How teachers present content online, and the frequency and quality of their communication with students, profoundly influence students learning experiences. Subsequently, the Peer factors also exhibit significant correlations. The coefficient for Behavioral Engagement stands at 0.753 (p < 0.001), for Cognitive Engagement at 0.757 (p < 0.001), for Emotional Engagement at 0.782 (p < 0.001), and for Interactive Engagement at 0.853 (p < 0.001). This underlines the importance of interaction and collaboration among students in online learning. Support, discussions, and collaborations with peers not only enhance learners comprehension but also elevate their level of interaction and engagement. Lastly, Environmental factors reveal correlation coefficients of 0.711 (p < 0.001) for Behavioral Engagement, 0.789 (p < 0.001) for Cognitive Engagement, 0.801 (p < 0.001) for Emotional Engagement, and 0.761 (p < 0.001) for Interactive Engagement. This highlights the positive role of the learning environment, such as technical support, resource availability, and platform user-friendliness, in influencing students emotional engagement. An efficient and user-friendly learning environment can enhance students satisfaction and level of commitment in learning. In summary, we discern that various influencing factors bear certain correlations with online learning Engagement, with the Self and Peer factors being particularly significant.

Table 9 Correlation analysis between online learning influencing factors and learning engagement levels.
Full size table

Optimization strategies and recommendations

Establishing a module for cultivating student self-drive

Self-drive is a pivotal factor for students to successfully complete online learning tasks. In traditional classroom settings, students might derive motivation and support from Teacher guidance and Peer interaction. However, in an online learning environment, students often require a heightened sense of autonomy and self-drive to maintain the continuity and depth of learning54. In this research, it’s evident that Self factors have a significant correlation with students’ online learning Engagement. Specifically, the correlation coefficients of Self factors with Behavioral Engagement, Cognitive Engagement, Emotional Engagement, and Interactive Engagement all indicate the central role of students’ intrinsic intentions, attitudes towards learning, and self-motivation in online learning. This further corroborates the significance of students’ personal attributes like interests, motivations, and learning strategies in influencing their online learning experiences. Consequently, it’s advised that educational institutions and educators incorporate a specialized module for nurturing student self-drive during online course design. This module should encompass:

  1. (1)

    Identification and establishment of motivation Assist students in recognizing their motivations for learning, be it out of interest, career advancement, or other reasons, and set specific learning objectives based on these motivations55.

  2. (2)

    Training in self-management skills Offer techniques and methods for time management, planning learning, and self-evaluation to aid students in organizing their studies more efficiently.

  3. (3)

    Resource and tool recommendations Introduce and suggest online tools and resources that can aid students in establishing and maintaining their self-drive, such as learning management software and goal-tracking tools.

  4. (4)

    Through the self-drive cultivation module, students can not only enhance their outcomes in online learning but also manage themselves better in their future work and life, realizing sustained growth driven by themselves.

Enhancing online interaction among teachers and students to foster the formation of an online learning community

In the context of online learning, Interactive Engagement stands as a crucial component of student Engagement, encompassing interactions between students and teachers, students and their peers, students and learning materials, as well as students and technology56. This research reveals that students’ self-drive, attitudes toward learning, and interactions and collaborations with peers play a key role in online learning and are essential for boosting students’ Interactive Engagement. To further amplify students’ online learning Engagement, educational institutions, and teachers should prioritize online interaction sessions and promote communication and collaboration among students. For instance, by utilizing online discussion forums, group collaborative tasks, real-time chat tools, etc., they can facilitate profound exchanges and knowledge sharing among students. Moreover, teachers should periodically organize online seminars and interactive activities, urging students to pose questions, share insights, and participate in discussions, thus fostering the formation of a learning community and elevating students’ Interactive Engagement.

Enhancing technical support and platform experience for online learning resources

The online learning environment plays a pivotal role in students’ learning outcomes and Engagement. The data from this study distinctly unveils the close relationship between the learning Environmental factors such as technical support, resource availability, and platform user-friendliness, with students’ Emotional Engagement, learning satisfaction, and commitment. To further elevate students’ online learning experiences and outcomes: The significance of technical support is self-evident. Throughout the online learning process, any technical glitch can potentially disrupt learning, consequently affecting students’ experiences and outcomes. Hence, ensuring the stability of the online learning platform is of utmost importance57. When students encounter technical issues, they should be able to swiftly receive assistance and support from the technical team. In addition, conducting regular technical checks and updates on the online platform to ensure its compatibility with the latest educational technologies and tools is key to enhancing students’ online learning experience. Secondly, the availability of resources directly correlates with students’ learning outcomes. All learning materials, whether they are videos, documents, quizzes, or discussion boards, should ensure 24/7 accessibility to cater to students’ anytime-anywhere learning needs. Simultaneously, it’s imperative to routinely update and optimize resources based on student feedback and learning data, ensuring the timeliness and quality of the materials, thus aiding students in achieving superior learning results. Lastly, the platform design should be intuitive and straightforward, allowing students to effortlessly locate and utilize the resources they require. Moreover, offering personalized learning paths and recommendations—suggesting suitable resources and courses tailored to each student’s learning history and interests—can further enhance students’ Engagement58.

Strengthening assessment and feedback of learning outcomes

Assessment and feedback constitute indispensable components of the teaching process. They not only provide students with information about their progress but also offer educators invaluable feedback on teaching efficacy59. Particularly in an online learning environment, where face-to-face Interactive Engagement is not present, efficient and timely assessment and feedback mechanisms become especially crucial. Based on the data presented in this paper, there is a pronounced relationship between students’ online Engagement and factors such as Self-motivation, Peer Interaction, and the quality of learning Environmental factors, all of which serve as pivotal determinants of online learning outcomes. Therefore, assessment should not be confined merely to learning results but should also encompass these key factors, providing students with holistic feedback on their learning experience. The data from this study indicates a significant association between the duration of online learning and multiple Engagement metrics, including Behavioral Engagement, Cognitive Engagement, Emotional Engagement, and Interactive Engagement. This suggests that during evaluations, it is necessary to consider the duration of a student’s online learning and acknowledge the potential impact it may entail. For instance, longer online learning sessions might imply that students have had ample opportunities to deepen comprehension, reflect on content, and master learning materials. Hence, it is essential during assessments to pay particular attention to these students, determining whether they have genuinely grasped the knowledge. In addition, incorporating diversified assessment methods can serve as viable alternatives to traditional pen-and-paper exams and homework evaluations, which may not fully capture the multifaceted dimensions of online learning. Online quizzes, automated grading systems, and Peer evaluations not only provide students with prompt feedback but also enable educators to obtain a more comprehensive view of students’ learning status. Lastly, the research data underscores that factors highly correlated with students’ online Engagement are their Self-motivation and Peer Interaction. Thus, during assessments, we should also investigate students’ level of engagement and Interaction, examining whether they are genuinely immersed in learning and whether they can effectively interact with their peers and Teachers. By drawing insights from data analysis, it is critical to refine and adapt evaluation techniques and content continuously, ensuring the overall effectiveness of online learning.

Conclusion and limitations

This study delves deeply into the online learning status and its influencing factors among undergraduate students majoring in Computer Science at universities. The research indicates that students’ online Engagement is shaped by a myriad of influencing variables, which underscores the multifaceted nature of digital learning contexts. A clear positive correlation emerges between the duration of online learning and the level of student Engagement. In contrast, factors such as gender, Home location, and major appear to exert a relatively limited influence on online Engagement. Moreover, individual elements, including subjective willingness, attitude, and motivation, play a pivotal role in determining students’ online Engagement. The Interactive aspects, involving both the Teacher and Peer, teaching methodologies, and timely feedback, also display a strong connection with students’ Emotional Engagement. Additionally, this study highlights the significance of technical support and the user-friendliness of online platforms in influencing Emotional Engagement. In the contemporary digital era, the role of technology in education has become increasingly critical, serving as an indispensable factor in elevating students’ satisfaction and efficacy in online learning. Overall, this research offers valuable suggestions for refining online learning strategies, including the reinforcement of students’ Self-motivation, the cultivation of robust online interactions between teachers and students, and the enhancement of technical support and platform experiences. It is anticipated that these recommendations could serve as a reference for further developments in the area of online education at higher education institutions.

Nevertheless, there are certain limitations in this study that should be considered. The sample was drawn from undergraduate students majoring in Computer Science at “H University” in central China, which may introduce a regional and academic bias, thus influencing the generalizability of the findings. The cross-sectional design employed herein may not fully capture long-term changes and trends in students’ online Engagement. Although the analysis encompasses individual differences such as gender, Home location, and the duration of online learning with respect to online Engagement, it may omit other potential variables, including learning motivation and prior online learning experiences. Further investigations could examine such elements and their impact on online Engagement. This work includes an exploration of various influences—covering students, Teacher, Peer, and Environmental aspects—yet certain crucial variables remain unaddressed, such as course content design and the specific technical platforms employed, which can substantially affect online learning outcomes and Engagement. The current study primarily relies on quantitative data, which may not fully convey the nuanced perspectives and experiences of learners, as interviews and observations were not integrated. Future research endeavors could adopt a mixed-methods approach to offer deeper insights into learners’ perceptions. Moreover, given the rapid evolution of online education technologies and the dynamism of Environmental factors, continuous monitoring is recommended to foster a more comprehensive and profound understanding of online Engagement.