Research on Factors Affecting Online Learning Stickiness among College Students from the Perspective of NAM-TPB Integration

doi:10.21203/rs.3.rs-3215017/v1

Download PDF

Article

Research on Factors Affecting Online Learning Stickiness among College Students from the Perspective of NAM-TPB Integration

https://doi.org/10.21203/rs.3.rs-3215017/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

With the rise of digitalization in the post-pandemic era, online learning among college students has become increasingly common. However, stu-dents' participation in online learning has been hindered by low efficiency and poor sustainability, which have negatively impacted their learning out-comes. This study examines the factors influencing the stickiness of online learning among college students from the perspectives of moral responsibil-ity, psychology, and society, using the integrated framework of the Norm Activation Model (NAM) and the Theory of Planned Behavior (TPB). A sur-vey questionnaire with 25 measurement items was constructed based on the NAM and TPB integrated model as the theoretical framework to collect data on students' online learning, and a structural equation model was built to explore the factors influencing the stickiness of online learning among col-lege students in Henan province. The results show that: (1) personal norms have a significant direct positive effect on the stickiness of online learning, as revealed by the hypothesized path analysis; (2) subjective norms have a significant indirect effect on personal norms through responsibility attribu-tion, as revealed by the mediation analysis; and (3) responsibility attribution is the most important influencing factor, as revealed by the analysis of di-rect, indirect, and total effects. Based on the research findings, it is recom-mended to cultivate students' belief and moral responsibility in online learn-ing, and to guide them to actively take learning responsibility and prevent learning fatigue.

Biological sciences/Psychology/Human behaviour

Earth and environmental sciences/Environmental social sciences

Online Learning

Theory of Planned Behavior

Norm Activation Model

Psychological Factors

Stickiness of Learning

Since the global outbreak of COVID-19 in 2020, universities both domestically and internationally have widely adopted online teaching methods, achieving the goal of "suspending classes without suspending learning, and suspending learning without suspending classes". Although the COVID-19 epidemic has temporarily come to an end, the country is increasingly valuing online learning. The "National Informationization Development Strategy Outline" proposes to increase the promotion of educational informationization, establish an open learning mode under the network environment, and promote more schools to use online courses. The "2017 Work Points for Educational Informationization" issued by the Ministry of Education's Office explicitly states that it is necessary to organize and promote the universal action of online education and implement the "Opinions on Strengthening the Construction, Application, and Management of Open Courses in Universities". Therefore, online learning has become a major trend in the development of education in today's era.

The question then becomes, how can online learning "stick" to learners, that is, improve learning stickiness, which refers to the stimulation of learning motivation, the intelligent provision of learning resources, and the efficient improvement of learning effectiveness. In the face of the normalization trend of online learning, how to improve the stickiness of online learning for university students, narrow the gap between online and offline teaching, and promote the sustainability of online learning has become a key focus of online teaching in universities. As the backbone of the new era, the quality of college students' classes and the sustainability of online learning are related to students' learning outcomes, employment prospects, and future contributions to society. Therefore, in the era of digital intelligence and the popularization of online learning, it is of great significance to explore the factors affecting the stickiness of online learning for university students.

Currently, scholars have already established a certain foundation for researching students' online learning. Scholars have applied community exploration theory, group dynamic theory, constructivist learning theory, knowledge contextualization theory, technology acceptance model, and expectancy-value theory to study the factors affecting online learning stickiness from the perspectives of responsibility, psychology, sociology, and other fields. This study will combine the regulation activation theory and the planned behavior theory as the basis, collect data through questionnaire surveys, use structural equation models to test and discuss the impact of model variables on the stickiness of online learning for college students, and explore the factors influencing the stickiness of online learning for university students.

1.1 Studying the Factors Affecting Student Online Learning Stickiness from a Moral Responsibility Perspective

Gu et al. (2015)¹, Donatella et al. (2016)², Haugland et al. (2022)³, and others have analyzed the research on student online collaborative learning and believe that students who take on more roles in the online learning process can stimulate stronger group cognitive and responsibility senses, pay more attention to relationship maintenance in collaborative learning, and improve the effectiveness of online learning.

Preget et al. (2023)⁴, Geletu et al. (2022)⁵, Li et al. (2022)⁶, and others have proposed that students should place greater emphasis on values such as morality and responsibility in online learning from the perspective of responsibility management for students and teachers. Morality and responsibility serve as intersection points in the learning process to enhance students' sense of learning responsibility and learning outcomes.

Ahmad et al. (2021)⁷, Kalichman et al. (2021)⁸, Schaeper et al. (2020)⁹, and others have taken the perspective of learning responsibility and conducted periodic scientific skill and cognitive ability tests on students' responsibility characteristics. They believe that consciously cultivating and training students' sense of responsibility skills can help to quickly stimulate their learning potential and improve their academic performance.

1.2 Factors Affecting Students' Online Learning Stickiness: A Psychological Perspective

Guo et al. (2022)¹⁰, Erika et al. (2019)¹¹, Müller et al. (2021)¹², and others have applied the "Technology Acceptance Model" to online learning, using subjective attitudes, intrinsic beliefs, and behavioral intentions to explain people's acceptance of information technology. Their analysis shows that perceived usefulness is an important factor affecting students' active participation in online learning.

Schneider et al. (2022)¹³, Zambrano et al. (2022)¹⁴, Shekinah et al. (2022)¹⁵, Arbaugh et al. (2014) ¹⁶using metacognition theory, emphasize that students can improve their online learning effectiveness through self-monitoring, self-awareness, and self-regulation in cognitive activities. They believe that students' self-efficacy and learning motivation clarity are important factors affecting their online learning.

Emos et al. (2010)¹⁷, Chen et al. (2012)¹⁸, Alenezi et al. (2023)¹⁹, and others have analyzed hundreds of potential factors that directly or indirectly influence students' online learning, and argue that students' satisfaction with online learning is a key factor in improving their classroom participation.

1.3 Factors Affecting Student Online Learning Stickiness from a Social Perspective

Kayode et al. (2014)²⁰, Alismaiel et al. (2022)²¹, and Terblanche et al. (2023)²² explored the behavioral characteristics of individual learners in social groups from the perspective of online learning dialogue interaction and found that interactive activities in online learning significantly affect learning performance and have a positive regulatory effect on student learning outcomes.

Constantine Han et al. (2018)²³, Kartsoni et al. (2023)²⁴, and Miah et al. (2023)²⁵ analyzed the factors affecting individuals in online learning from multiple perspectives such as teachers, learners, and social environment and found that social influence and outcome cognition have a significant impact on student online learning.

Florence et al. (2019)²⁶, Simon et al. (2020)²⁷, and Junus et al. (2021)²⁸ argue that improving the online teaching and interaction ability of university teachers can enhance student participation in online learning and generate stronger learning interest, thus promoting more active participation in online learning.

Currently, there are numerous research results on the factors affecting student online learning stickiness. However, most of them only analyze the influencing factors from a single perspective of moral responsibility, psychology, or social aspects. There are some deficiencies in the research on the stickiness of online learning for college students: 1. There is limited research on the online learning behavior of college students as a group; 2. Most of the studies only analyze the influencing factors from a single perspective of moral responsibility, sociology, or psychology, and rarely combine the three perspectives to analyze the factors affecting the stickiness of college students' online learning; 3. Previous researchers have mostly followed the analysis mode of online learning behavior analysis-reason analysis-countermeasure suggestions, and classical social science theories have not yet matured in the sustainable theory of online learning, and empirical research is relatively scarce. Based on this, this article starts from the micro-individual level, combines the regulatory activation theory and planned behavior theory, and constructs a structural equation model from the moral responsibility, social, and psychological perspectives to explore the factors affecting the stickiness of online learning for college students.

2.1 The theory of norm activation and online learning stickiness

The norm activation theory was proposed by Schwartz (1977)²⁹ to predict human behavior patterns from the perspective of moral responsibility and psychology. According to Schwartz, 1) society has an impact on human behavior, which can either motivate or constrain individuals' actions, but not all individuals conform to social norms; 2) individual behavior can be analyzed from the perspective of moral responsibility and psychology, emphasizing that individuals can exhibit positive social behavior while considering their own interests. For example, individuals can help regulate their own behavior, develop a sense of moral obligation, enhance their sense of social responsibility, beliefs, and values. Those who do not follow social norms will face social punishment and self-discipline; 3) whether individuals can exhibit positive social behavior depends on whether their personal norms conform to social attributes, whether their sense of moral obligation is activated, and the influence of psychological activities such as cost-benefit analysis, which ultimately reflects on the stickiness of online learning. The norm activation theory suggests that individuals will perceive the actual consequences of their behavior during the implementation process and develop a sense of responsibility for positive social behavior, which is reflected in their awareness of consequences (AC), attribution of responsibility (AR), and personal norms (PN), which are ultimately reflected in their intention to participate in online learning stickiness(OLS).

In the proposed model, the factors affecting the stickiness and efficiency of online learning among college students can be considered from multiple perspectives. Assuming that students are aware that participating in online learning not only meets the requirements of contemporary learning methods but also facilitates their learning and improves their grades, they will be motivated to actively participate in online learning. Secondly, when students realize that their serious participation in online learning not only enhances their knowledge but also has a positive impact on others and society, they will consider it their responsibility to participate seriously. Finally, when students are aware of their major responsibility for online learning and society, they are more willing to participate actively. Based on the norm activation theory and the actual situation of online learning, the following hypotheses are proposed. The path assumption is shown in Figure 1:

H₁: Awareness of consequences has a positive impact on attribution of responsibility.

H₂: Awareness of consequences has a positive impact on personal norms.

H₃: Attribution of responsibility has a positive impact on personal norms.

H₄: Personal norms have a positive impact on online learning stickiness.

2.2 Theory of Planned Behavior and Online Learning Stickiness

The Theory of Planned Behavior (Ajzen, 1977)³⁰ posits that individuals typically engage in social and psychological behaviors in social activities, and their primary basic principle in social activities is to satisfy their own interests by taking practical actions. Its main tenets are: 1. An individual's subjective intention to act is the most important factor that influences actual behavior. An individual's intention to act is primarily influenced by their attitude towards the event. When an individual perceives a behavior as favorable, they are more likely to take positive actions when making decisions. The individual's attitude is a potential deciding factor for whether a behavior is implemented or not; 2. Individual behavior is influenced by social pressure. Social behavior to some extent impacts individual behavior. When an individual feels pressure from a social group, they are more likely to have the intention to act and take action; 3. Individuals can perceive the difficulties and obstacles that objective factors such as resources and abilities may cause for implementing a behavior, which has a direct or indirect impact on actual behavior. The Theory of Planned Behavior suggests that when an individual has the intention to act, their behavior is primarily influenced by their attitude towards the behavior (Attitude, ATT). When individuals perceive a behavior as favorable, they are more likely to be aware of the importance of the event to society when making decisions and form subjective norms (Subjective Norm, SN). At the same time, individuals are more likely to perceive the ease and difficulty of implementing the behavior, that is, Perceptual Behavior Control (PBC), which ultimately makes it easier for them to take actual action.

In the process of students participating in online learning, if they perceive positive consequences of their behavior, they are more likely to make positive judgments and actively participate in online learning. Therefore, in research, the students' attitude towards behavior is considered an important variable. Once an individual realizes that a specific behavior may bring positive effects, they are more likely to understand the impact of the behavior on themselves and the environment. Students' positive attitudes towards behavior play an important role in understanding the perception of online learning outcomes. Based on this, the following hypotheses are proposed. The path assumption is shown in Figure 1:

H₅: Attitude towards behavior has a positive impact on online learning stickiness.

H₆: Attitude towards behavior has a positive impact on personal norms.

H₇: Attitude towards behavior has a positive impact on perceived behavior control.

H₈: Subjective norms have a positive impact on perceived behavior control.

2.3 Theoretical Framework of NAM-TPB Integration and Online Learning Stickiness

By combining the Norm Activation Model (NAM) and the Theory of Planned Behavior (TPB), it is assumed that students who are aware of the positive outcomes of participating in online learning and who recognize their obligation to fully engage in online learning will have a stronger inclination towards online learning. However, considering that students may be influenced by other social groups and may have uncertain thoughts about whether online learning is beneficial to society, they may not actively participate in online learning. Additionally, it is assumed that students who can assess the difficulties they may face when actively participating in online learning will have a greater sense of responsibility and will invest their sense of responsibility and belief in their actual work. Based on the significant impact of subjective norms and perceived behavioral control on students, the following hypotheses are proposed. The path assumption is shown in Figure 1:

H₉: Perceived behavioral control has a positive impact on responsibility attribution.

H₁₀: Subjective norms have a positive impact on responsibility attribution.

H₁₁: Subjective norms have a positive impact on personal norms.

Based on the research hypotheses, the theoretical model framework can be illustrated as shown below in Figure 1:

This study conducted a questionnaire survey on undergraduate students in Henan Province universities. In the era of digitization, many universities are continuously introducing and applying online learning to improve students' learning motivation and persistence. However, there is currently limited research on the factors that influence the stickiness of online learning for university students. Improving the sustainability and learning efficiency of students' online learning is an important issue that future higher education needs to face. Based on this, this study conducted a survey of undergraduate students in Henan Province universities, analyzed the factors that influence students' online learning stickiness, and chose a questionnaire survey as the most suitable way to obtain data, and formulated a questionnaire survey for this research purpose and conducted data collection and analysis.

3.1 Variable Measurement

This study constructed measurement variables based on the internal and external factors of students' participation in online learning, taking result consciousness, responsibility attribution, personal norms, perceived behavioral control, subjective norms, and behavioral attitudes as independent variables, and online learning stickiness as the dependent variable, to explore the key factors that affect students' online learning stickiness. The first part of the questionnaire explains the purpose of the study and provides instructions for the questionnaire survey, which includes a privacy statement and requires sampling personnel to provide certain personal information, including gender, grade, device usage, and the number of courses taken per semester. The second part describes the relevant items of result consciousness, responsibility attribution, personal norms, perceived behavioral control, subjective norms, behavioral attitudes, and online learning stickiness. Each latent variable contains three items, and all item statements use a 7-point Likert scale (1 point indicates strongly disagree, and 7 points indicate strongly agree). Before officially collecting data for this study, the reliability and validity of the questionnaire were checked through preliminary research on 50 college students from Henan Province. All participants in the pilot study had similar demographic characteristics to the main study sample, so that final data analysis could be conducted experimentally. Based on their feedback, some basic modifications were made to the questionnaire, and final data collection was conducted based on the revised questionnaire.

3.2 Variable Measurement

In this paper, the structural equation model and Amos 24.0 software were used to test the reliability and validity of the variables. The test results are shown in Table 1.

Awareness of Consequences

This item contains three items and uses a 7-point Likert scale. This latent variable contains the items "Participating seriously in online learning can effectively reduce student turnover and reduce the risk of epidemic spread", "Participating seriously in online learning is beneficial to the development and improvement of learning software", and "Participating seriously in online learning is beneficial to the rapid development of Internet platforms." The Cronbach's alpha value for result consciousness is 0.755. (Table 1)

Responsibility Attribution

This item contains three items and uses a 7-point Likert scale. This latent variable contains the items "I am responsible for improving my own academic performance", "I am responsible for improving my own academic performance", and "I am responsible for the formation of a good learning atmosphere in the class." The Cronbach's alpha value for responsibility attribution is 0.754. (Table 1)

Personal Norms

This item contains three items and uses a 7-point Likert scale. This latent variable contains the items "If I do not participate seriously in online learning, I will feel guilty and empty", "Participating seriously in online learning conforms to my principles and beliefs", and "As students, we all have a responsibility to participate seriously in online learning." The Cronbach's alpha value for personal norms is 0.753. (Table 1)

Attitudes

This item contains three items and uses a 7-point Likert scale. This latent variable contains the items "I think it is valuable for students to participate seriously in online learning", "I think it is a positive performance for students to participate seriously in online learning", and "I think it is beneficial for both myself and society for students to participate seriously in online learning." The Cronbach's alpha value for behavioral attitudes is 0.751. (Table 1)

Subjective Norms

This item contains three items and uses a 7-point Likert scale. This latent variable contains the items "My family and friends think that students have an obligation to participate seriously in online learning", "My family and friends often urge and educate me to participate seriously in online learning", and "My family and friends hope that I can participate seriously in online learning." The Cronbach's alpha value for subjective norms is 0.754. (Table 1)

Perceived Behavioral Control

This item contains three items and uses a 7-point Likert scale. This latent variable contains the items "When I make up my mind, I have the ability to participate seriously in online learning", "When I encounter difficulties, I can overcome them and participate seriously in online learning", and "Whether or not someone is monitoring, I can participate seriously in online learning." The Cronbach's alpha value for perceived behavioral control is 0.754. (Table 1)

Online Learning Stickiness

This item contains three items and uses a 7-point Likert scale. This latent variable contains the items "I advocate online learning", "I am willing to practice online learning", and "I am willing to persist in online learning and share my online learning experience with others." The Cronbach's alpha value for online learning stickiness is 0.751. (Table 1)

Table 1. Test table of variable reliability and validity

Latent Variable	Dimensional Quantity	Topic Number	Standard Factor Load	Cronbach's alpha	AVE Square Root	AVE	CR
AC	3	AC₁	0.869	0.755	0.874	0.764	0.907
		AC₂	0.884
		AC₃	0.869
AR	3	AR₁	0.973	0.754	0.91	0.828	0.935
		AR₂	0.886
		AR₃	0.868
PN	3	PN₁	0.879	0.753	0.867	0.751	0.9
		PN₂	0.851
		PN₃	0.869
PBC	3	PBC₁	0.88	0.754	0.881	0.776	0.912
		PBC₂	0.875
		PBC₃	0.887
ATT	3	ATT₁	0.894	0.751	0.889	0.79	0.919
		ATT₂	0.894
		ATT₃	0.878
SN	3	SN₁	0.886	0.754	0.881	0.776	0.912
		SN₂	0.896
		SN₃	0.861
OLS	3	OLS₁	0.871	0.751	0.874	0.764	0.907
		OLS₂	0.884
		OLS₃	0.867

3.3 Sampling and Data Collection

The questionnaire survey targeted undergraduate students in Henan Province universities, aiming to gain in-depth understanding of their participation in online learning and analyze the main factors considered for their learning effectiveness. Empirical data was collected through the questionnaire survey. Relevant survey instructions were attached to the questionnaire survey to ensure that each research subject who participated in the survey understood the purpose of the survey and that the collected data was only used for research purposes. More than 500 questionnaires were distributed through online and face-to-face methods in Henan Province universities, and after eliminating and organizing the questionnaires that were not rigorous, 481 valid questionnaires were collected, with a valid recovery rate of over 96%. Further data analysis was conducted on the main information and details of the survey subjects in this study, as shown in Table 2.

3.4 Summary of Respondents

Out of the 481 survey responses collected, basic information of college students who participated in the survey included four characteristics: gender, grade level, devices used for online learning, and number of courses taken per semester. In terms of gender, female college students represented 61.75% of the respondents, while male students represented 38.25%. Regarding grade level, second and third year students had the highest representation, accounting for 36.80% and 44.91% respectively. In terms of devices used for online learning, 74.22% and 76.72% of the respondents reported using computers and mobile phones respectively. In terms of the number of online courses taken per semester, the majority of respondents (56.55%) reported taking 3-6 courses per semester.

Table 2. Basic characteristics of respondents

Personal attributes	Classification	Frequency	Percentage	Class participation	Classification	Frequency	Percentage
Gender	Male	184	38.25%	Do you think that every student has the responsibility to study online seriously?	Not much responsibility	143	77.72%
					More responsible	30	16.30%
					Very responsible	11	5.98%
	Female	297	61.75%		Not much responsibility	269	90.57%
					More responsible	18	6.06%
					Very responsible	10	3.37%
Grade	Freshman	12	2.49%	Are you willing to practice online learning?	Be reluctant to	11	91.67%
					Normal	1	8.33%
					Be willing to	0	0.00%
	Sophomore	177	36.80%		Be reluctant to	108	61.02%
					Normal	48	27.12%
					Be willing to	21	11.86%
	Junior	216	44.91%		Be reluctant to	137	63.43%
					Normal	61	28.24%
					Be willing to	18	8.33%
	Senior	76	15.80%		Be reluctant to	45	59.21%
					Normal	23	30.26%
					Be willing to	8	10.53%
Use Equipment	Mobile phone	357	74.22%
	Computer	369	76.72%
	Flat Plate	95	19.75%
Number of classes per semester	1—3	176	36.59%	Do you advocate online learning?	Less advocate	86	48.86%
					Comparative advocacy	70	39.77%
					Very advocate	20	11.36%
	3—6	272	56.55%		Less advocate	120	44.12%
					Comparative advocacy	100	36.76%
					Very advocate	52	19.12%
	6—9	33	6.86%		Less advocate	15	45.45%
					Comparative advocacy	9	27.27%
					Very advocate	9	27.27%

In the analysis, the variables of gender, grade level, and number of courses were compared with attitudes and willingness towards online learning. Regarding gender, there was a significant difference in attitudes towards online learning, with 21% of male students feeling responsible for taking online learning seriously compared to only 10% of female students. Grade level also had an impact on attitudes towards online learning, with the percentage of students who were less willing to take online courses decreasing from 91.67% in the first year to 59.21% in the fourth year. It is evident that grade level has a significant influence on attitudes towards online learning. The number of courses taken per semester did not have a significant impact on attitudes towards online learning, with approximately 55% of respondents reporting no difference.

4.1 The Measurement Model

To assess the chance relationships in the model, a structural equation model was used for measurement in this study, with SPSS 24.0 and AMOS 24.0 used for quantitative analysis of the survey data. The results showed that the overall Cronbach's alpha coefficient of the model was 0.841, exceeding the threshold condition of 0.7, indicating good internal consistency of the overall scale. The Cronbach's alpha coefficients of each latent variable ranged from 0.751 to 0.755, all exceeding the evaluation standard value of 0.7, with the accompanying probability of the Bartlett sphere test being less than 0.001. The standardized factor loading values of each corresponding item in the latent variables ranged from 0.851 to 0.973, all exceeding the standard value of 0.6, indicating good representativeness of the items in the latent variables. In addition, the combination reliability of the latent variables all exceeded the evaluation standard of 0.7, indicating good composite reliability and convergent validity of the model; and the average variance extracted all exceeded the standard value of 0.5, indicating good convergent validity of the model. The specific values are shown in Table 1.

The model was analyzed for fit using AMOS 24.0, and the results are shown in Table 3. According to the evaluation standards of fit indices, the overall fit of the model was considered good. The actual value of CMIN/DF in the absolute fit index was 1.187, which was less than the evaluation standard of 3.000, indicating ideal fit. The actual value of RMSEA was 0.023, which was less than the evaluation standard of 0.08, indicating ideal fit. The actual values of NFI, CFI, IFI, TLI, and RFI were 0.961, 0.994, 0.994, 0.992, and 0.953, respectively, all exceeding the evaluation standard of 0.900, indicating ideal fit.

Table 3. Results of model fit test

Fit Test
Fitting index	CMIN/DF	RMSEA	NFI	CFI	IFI	TLI	RFI
Evaluation criteria	<3	<0.08	>0.9	>0.9	>0.9	>0.9	>0.9
Actual value	1.187	0.023	0.961	0.994	0.994	0.992	0.953

4.2 Structural Model

The structural paths reflect the degree of influence between the latent variables in the model, and all of the 11 hypotheses designed in this paper were verified. Table 4 shows the results of the hypothesis path tests in the model. At the significance level of 0.001, result consciousness (AC) positively influenced responsibility attribution (AR), with a standardized path coefficient of 0.251, and hypothesis H1 was established. At the significance level of 0.01, result consciousness (AC) positively influenced personal norms (PN), with a standardized path coefficient of 0.101, and hypothesis H2 was established. At the significance level of 0.001, responsibility attribution (AR) positively influenced personal norms (PN), with a standardized path coefficient of 0.305, and hypothesis H3 was established. At the significance level of 0.001, personal norms (PN) positively influenced online learning stickiness (OLS), with a standardized path coefficient of 0.296, and hypothesis H4 was established. At the significance level of 0.001, attitude (ATT) positively influenced online learning stickiness (OLS), with a standardized path coefficient of 0.152, and hypothesis H5 was established. At the significance level of 0.001, attitude (ATT) positively influenced personal norms (PN), with a standardized path coefficient of 0.156, and hypothesis H6 was established. At the significance level of 0.001, attitude (ATT) positively influenced perceived behavioral control (PBC), with a standardized path coefficient of 0.280, and hypothesis H7 was established. At the significance level of 0.005, subjective norms (SN) positively influenced perceived behavioral control (PBC), with a standardized path coefficient of 0.126, and hypothesis H8 was established. At the significance level of 0.001, perceived behavioral control (PBC) positively influenced responsibility attribution (AR), with a standardized path coefficient of 0.180, and hypothesis H9 was established. At the significance level of 0.001, subjective norms (SN) positively influenced responsibility attribution (AR), with a standardized path coefficient of 0.200, and hypothesis H10 was established. At the significance level of 0.01, subjective norms (SN) positively influenced personal norms (PN), with a standardized path coefficient of 0.033, and hypothesis H11 was established.

Table 4. Model path hypothesis testing

Hypothesis	Hypothesis path test	Standardized regression coefficient	Standard error	Critical value	Conclusion
H₁	AC→AR	0.271	0.056	4.585^***	Support
H₂	AC→PN	0.108	0.059	1.720^*	Support
H₃	AR→PN	0.304	0.062	4.923^***	Support
H₄	PN→OLS	0.292	0.06	4.927^***	Support
H₅	ATT→OLS	0.152	0.058	2.619^***	Support
H₆	ATT→PN	0.158	0.059	2.664^***	Support
H₇	ATT→PBC	0.28	0.061	4.588^***	Support
H₈	SN→PBC	0.127	0.06	2.110^**	Support
H₉	PBC→AR	0.182	0.054	3.311^***	Support
H₁₀	SN→AR	0.205	0.058	3.465^***	Support
H₁₁	SN→PN	0.034	0.061	0.537^*	Support
Note: *, and * represent significant at 1%, 5% and 10% levels respectively.

4.3 Mediating Effect

The structural paths reflect the degree of influence between the latent variables in the model, and all of the 11 hypotheses designed in this paper were verified. Using AMOS 24.0 software, the mediating effects included in the model were analyzed using the deviation-corrected non-parametric percentile bootstrap method. The sampling frequency was set to 2000, and the bootstrap test results of the direct, indirect, and total effects between the variables in the model and the mediating paths are shown in Tables 5 and 6, respectively. The standardized regression coefficients of the model are shown in Figure 2.

Table 5 presents the results of the bootstrap test for each mediating path in the model. The point estimate for the mediating path from subjective norm to perceived behavioral control to responsibility attribution was 0.023 with a confidence interval of [0.002, 0.055], which does not contain zero, indicating a significant partial mediating effect and a valid mediating path. The point estimate for the mediating path from subjective norm to responsibility attribution to personal norm was 0.068 with a confidence interval of [0.032, 0.120], which does not contain zero, indicating a significant partial mediating effect and a valid mediating path. The point estimate for the mediating path from result awareness to responsibility attribution to personal norm was 0.078 with a confidence interval of [0.040, 0.133], which does not contain zero, indicating a significant partial mediating effect and a valid mediating path. The point estimate for the mediating path from behavioral attitude to personal norm to online learning stickiness was 0.051 with a confidence interval of [0.018, 0.100], which does not contain zero, indicating a significant partial mediating effect and a valid mediating path.

Table 5. Bootstrap Test for Each Mediating Path in the Model

Intermediate path	SE	Standardized point estimates	95%CI after correction for deviation		P value
			Lower Limit	Upper Limit
SN→PBC→AR	0.013	0.023	0.002	0.055	0
SN→AR→PN	0.022	0.068	0.032	0.12	0
AC→AR→PN	0.021	0.078	0.04	0.133	0
ATT→PN→OLS	0.02	0.051	0.018	0.1	0

Table 6 shows the direct, indirect, and total effects of each path on students' willingness to participate in online learning. Among all the paths, responsibility attribution has the largest total effect on personal norm, with a value of 0.304, while perceived behavioral control has the smallest total effect on online learning stickiness, with a value of 0.016. In terms of direct effects, responsibility attribution has the largest direct effect on personal norm, with a value of 0.304, while subjective norm has the smallest direct effect on personal norm, with a value of 0.034. In terms of indirect effects, responsibility attribution has the largest indirect effect on online learning stickiness, with a value of 0.089, while behavioral attitude has the smallest indirect effect on online learning stickiness, with a value of 0.015.

Table 6. Analysis of Direct, Indirect, and Total Effects Between Variables in the Theoretical Model

Path	Direct effect	Indirect effect	Total effect
AC→PN	0.101	0.082	0.183
AC→AR	0.271		0.271
AC→OLS		0.056	0.056
AR→PN	0.304		0.304
AR→OLS		0.089	0.089
PN→OLS	0.292		0.292
SN→PN	0.034	0.069	0.103
SN→OLS		0.03	0.03
SN→AR	0.205	0.023	0.228
SN→PBC	0.127		0.127
PBC→AR	0.182		0.182
PBC→PN		0.055	0.055
PBC→OLS		0.016	0.016
ATT→PBC	0.28		0.28
ATT→AR		0.051	0.051
ATT→PN	0.158	0.015	0.173
ATT→OLS	0.152	0.051	0.203

This study aims to explore the factors influencing the stickiness of online learning among college students through the combination of the norm activation theory and the theory of planned behavior. Most previous studies have focused on analyzing college students' willingness to participate in online learning from the perspectives of task-technology match and interaction imbalance. This study is the first to investigate the factors influencing the stickiness of online learning among college students in the post-pandemic era within the framework of the norm activation theory and the theory of planned behavior, combining moral responsibility, psychological, and social factors to analyze the sustained impact of online learning.

The results show that among all the assumed paths, behavioral attitude and personal norms have a significant direct impact on the stickiness of online learning, with personal norms having a greater positive impact (coefficient: 0.292). This can be explained from the perspective of moral responsibility. Personal norms can evoke a sense of pride or guilt, thereby influencing actual online learning behavior to some extent. This is somewhat different from the study of Pedro et al. (2021)³¹, which may be due to differences in learning courses and the influence of personal norms on online learning. It is also somewhat different from the study results of Singh et al. (2023)³², which may be due to the traditional teaching concept that emphasizes learning as a process of social interaction. The separation of time and space in online learning has hindered the social interaction between teachers and students, so it is believed that online learning can only have better learning results if it matches students' ability.

In the mediation analysis, subjective norms have the greatest indirect effect on personal norms through responsibility attribution, with an impact degree of 0.022. This may be because students are more likely to perceive the importance of actively participating in online learning when influenced by external factors, and feel a strong sense of moral responsibility and belief that they should be responsible for their own and others' learning environment, thereby regulating their own behavior and increasing their willingness to participate in online learning. This is consistent with the research results of Ader et al. (2023)³³, because online learning is the cooperation and mutual assistance generated by the community of learning individuals, teachers, and learning peers under certain rules and division of labor, which restores the sociality of teaching, stimulates students' sense of moral responsibility, and is conducive to promoting the sustainability of students' online learning.

Furthermore, in direct effects and total effects analysis, responsibility attribution has the largest effect on personal norms (coefficient: 0.304), indicating that when students have a strong sense of moral responsibility towards online learning, they believe that their behavior will affect their own and others' learning outcomes, resulting in a certain belief and regulating their own behavior. This is inconsistent with the study of Glassman et al. (2021)³⁴, which may be due to the fact that individual learning effectiveness is not only a process of its own dynamic evolution, but also influenced by collective feedback and interaction, which leads to certain deviations in research results.

In the indirect effect, responsibility attribution also has the largest indirect effect on the stickiness of online learning among students (impact degree: 0.089), as students who perceive their responsibility for online learning are more likely to actively participate and take responsibility for their learning, even in the face of setbacks and difficulties. This is consistent with the research results of Vannini et al. (2022)³⁵, which indicates that cultivating students' moral responsibility for learning is one of the important ways to stimulate students' willingness to learn and enhance their learning effectiveness, and it is also an important influencing factor to improve students' learning effectiveness in the process of online learning.

In conclusion, cultivating students' sense of moral responsibility is an important way to enhance their learning willingness and effectiveness, and is an important factor in enhancing the stickiness of online learning.

The argument of this study supports that the stickiness of online learning among college students is influenced by various factors including responsibility, morality, social, and psychological aspects. The research findings suggest that: 1. through the analysis of hypothetical paths, personal norms have a significant direct positive impact on the stickiness of online learning, and it is recommended that universities should cultivate students' beliefs in online learning and improve their self-management abilities. 2. in the analysis of mediating effects, subjective norms significantly influence personal norms through responsibility attribution, and it is suggested that universities cultivate learning models and establish a learning image to motivate students to learn from exemplary representatives and develop a sense of learning responsibility to regulate personal behavior. 3. From the analysis of the effects, including direct, indirect, and total effects, responsibility attribution is the most important influencing factor, and it is recommended that universities cultivate students' sense of responsibility and guide them to actively assume learning responsibility to prevent learning fatigue.

However, this study also has certain limitations: 1. The research sample mainly consists of college students in formal learning settings, and future research can compare formal online learning with non-formal online learning. 2. The study does not consider technological devices and equipment as factors that affect learning stickiness, which are essential in online learning, and future research can incorporate them into the study.

Author Contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Xiaoyan Li, Yaxin Tan, Yuanqing Wei and Lei Zhou. The first draft of the manuscript was written by Yaxin Tan and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Conceptualization: Xiaoyan Li; Methodology: Xiaoyan Li, Yaxin Tan; Data curation: Yaxin Tan; Resources: Xiaoyan Li; Writing-original draft preparation: Xiaoyan Li, Yaxin Tan; Writing-review and editing: Lei Zhou, Yuanqing Wei; Visualization: Xiaoyan Li, Yaxin Tan; Supervision: Xiaoyan Li, Yaxin Tan; Funding acqui-sition: Xiaoyan Li. All authors have read and agreed to the published version of the manuscript.

Funding

Undergraduate Research and Teaching Series Project on Industry-Specialized Universities in North China: "Research on the Implementation of Research-oriented Teaching and Value-added Development of Students in Management Majors" (No: 2022-9); Research on Digital Deep Integration and Innovative Practice of Materials in Higher Education in the School of Management and Economics at North China University of Water Resources and Electric Power (No: 86); General Project at North China University of Water Resources and Electric Power: "Research and Practice on the Reform of Training Mode in Applied Economics Under the Perspective of the 'Three News'" (2021SJGLX71Y); Philosophy and social Science innovation team, North China University of Water Resources and Electric Power (No: 03)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon request.

Compliance with Ethical Standards

The authors have no competing interests to declare that are relevant to the content of this article. All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.

The survey objects involved in this study are college students in Henan Province, China.

All authors are aware of and agree to the contents of the manuscript.

Acknowledgments

In this section you can acknowledge any support given which is not covered by the author contribution or funding sections. This may include administrative and technical support, or donations in kind (e.g., materials used for experiments).

Experiments Statement

A statement to confirm that all methods were carried out in accordance with relevant guidelines and regulations.

A statement to confirm that all experimental protocols were approved by North China University of Water Resources and Electric Power.

Confirming that informed consent was obtained from all subjects and their legal guardians.

Before collecting the data of this study, the reliability and validity of the questionnaire were tested through a preliminary survey of 50 college students in Henan Province. All participants in the pilot study had similar demographic characteristics to the main study sample, so final data analysis could be performed experimentally. Based on their feedback, some basic changes were made to the questionnaire and final data collection was carried out based on the revised questionnaire.

Finally, the survey objects are college students in Henan Province, and the informed consent of all respondents has been obtained before the survey. The questionnaire is accompanied by a relevant survey description to ensure that each participant understands the purpose of the survey and that the data collected is used for research purposes only. More than 500 questionnaires were distributed online and face-to-face in colleges and universities in Henan Province. After sorting out the inaccurate questionnaires, 481 valid questionnaires were recovered, with an effective recovery rate of more than 96%.

Gu, X.Q., Shao, Y.J., Guo, X.F., Cher, P.L. Designing a role structure to engage students in computer-supported collaborative learning. The Internet and Higher Education. 1(24), 13-20, DOI: https://doi.org/10.1016/j.iheduc.2014.09.002 (2015)
Donatella.C, Stefano, C., Nobuko, F. Role taking and knowledge building in a blended university course. International Journal of Computer-Supported Collaborative Learning. 11(1), 9-39, DOI: https://doi.org/10.1007/s11412-015-9224-0 (2015)
Haugland, M.J., Rosenberg, I., Aasekjær, K. Collaborative learning in small groups in an online course – a case study. BMC Medical Education. 22(1), 165-165, DOI: https://doi.org/10.1186/s12909-022-03232-x (2022)
Preget, L. A conceptual framework for understanding the learning processes integral to the development of responsible management practice. Journal of Management Development. 2(42), 125-140, DOI: https://doi.org/10.1108/JMD-07-2022-0181 (2023)
Geletu, G.M., Mihiretie, D.M. Professional accountability and responsibility of learning communities of practice in professional development versus curriculum practice in classrooms: Possibilities and pathways. International Journal of Educational Research Open. 4, 100223, DOI: https://doi.org/10.1016/J.IJEDRO.2022.100223 (2023)
Li, C.L., Xing, W.L., Leite, W. Building socially responsible conversational agents using big data to support online learning: A case with Algebra Nation. British Journal of Educational Technology. 53(4), 776-803, DOI: https://doi.org/10.1111/BJET.13227 (2022)
Ahmad, F., Mastuang, M., Misbah, M.(2021). The implementation of guided discovery learning model to improve students’ characters of responsibility and academic skills. Journal of Physics: Conference Series. 1(2104), 012020, DOI: https://doi.org/ 10.1088/1742-6596/2104/1/012020 (2021)
Kalichman, M.W., Devereaux, M.L., Plemmons. D.K. A Course for Teaching and Learning About the Responsible Conduct of Research. Journal of empirical research on human research ethics : JERHRE. 3(17), 15562646211047, DOI: https://doi.org/ 10.1177/15562646211047163 (2021)
Schaeper, H. The first year in higher education: the role of individual factors and the learning environment for academic integration. Higher Education. 79, 95-110, DOI: https://doi.org/10.1007/s10734-019-00398-0 (2020)
Guo, J.P., Yang, L.Y., Zhang, J., & Gan, Y.J. Academic self-concept, perceptions of the learning environment, engagement, and learning outcomes of university students: relationships and causal ordering. Higher Education. 83, 809-828, DOI: https://doi.org/10.1007/s10734-021-00705-8 (2022)
Erika, A.P., Jeanette, Z. Facilitating Student Outcomes by Supporting Autonomy: Implications for Practice and Policy. Policy Insights from the Behavioral and Brain Sciences. 2(6), 1115-122, DOI: https://doi.org/ 10.1177/2372732219862572 (2019)
Müller, F.H., Thomas, A.E., Carmignola, M., Dittrich, A.K., Eckes, A., Großmann, N., Martinek, D., Wilde, M., Bieg, S. University Students’ Basic Psychological Needs, Motivation, and Vitality Before and During COVID-19: A Self-Determination Theory Approach. Frontiers in Psychology. 12, 775804, DOI: https://doi.org/10.3389/FPSYG.2021.775804 (2021)
Schneider, S., Nebel, S., Meyer, S., Rey, G.D. The interdependency of perceived task difficulty and the choice effect when learning with multimedia materials. Journal of Educational Psychology. 114(3), 443-461, DOI: https://doi.org/10.1037/EDU0000686 (2022)
Zambrano, J., Kennedy, A.A.U., Aguilera, C., Yates, N., Patall, E.A. Students’ beliefs about agentic engagement: A phenomenological study in urban high school physical science and engineering classes. Journal of Educational Psychology. 114(5), 1028-1047, DOI: https://doi.org/10.1037/EDU0000690 (2022)
Shekinah, S.I., Chinnasamy, P., Deepsheka, K., Singaram, V. Impact of online education due to the pandemic among college students: Knowledge, Attitude and Practices analysis with structural equation modeling. Journal of education and health promotion. 11, 189-189, DOI: https://doi.org/10.4103/JEHP.JEHP_995_21 (2022)
Arbaugh, J.B., Eom, S.B. What Might Online Delivery Teach Us About Blended Management Education? Prior Perspectives and Future Directions. Journal of Management Education. 38(6), 1052562914534244, DOI: https://doi.org/10.1177/1052562914534244 (2014)
Eom, S.B., Arbaugh, J.B. Student Satisfaction and Learning Outcomes in e_Learning: An Introduction to Empirical Research. Hershey, PA: IGI Global. 37-56, DOI: https://doi.org/10.4018/978-1-60960-615-2 (2010)
Chen, H.C., Stotlar, D. An Examination of the Motivation and Satisfaction of College Students Enrolled in Physical Education Courses. Sport Science Review. XXI(1-2), 43-63, DOI: https://doi.org/10.2478/v10237-012-0003-9 (2012)
Alenezi S., Bahathig A., Soliman M., Alhassoun H., Alkadi N., Albarrak M., Albadrani W., Alghoraiby R., Alhaddab A., AlEyadhy A. Performance and satisfaction during the E-learning transition in the COVID-19 pandemic among psychiatry course medical students. Heliyon. 9(6), e16844, DOI: https://doi.org/10.1016/j.heliyon.2023.e16844 (2023).
Kayode C.V.E.O., Tian L.T. The impact of transactional distance dialogic interactions on student learning outcomes in online and blended environment. Computers & Education. 7(8), 414-427, DOI: https://doi.org/10.1016/j.compedu.2014.06.011 (2014).
Alismaiel O.A., Cifuentes F.J., AlRahmi W.M. Online Learning, Mobile Learning, and Social Media Technologies: An Empirical Study on Constructivism Theory during the COVID-19 Pandemic. Sustainability. 14(18), 1-15, DOI: http://doi.org/10.3390/su141811134 (2022).
Terblanche W., Lubbe I., Papageorgiou E.,Van D. M. N. Acceptance of e-learning applications by accounting students in an online learning environment at residential universities. South African Journal of Accounting Research. 37(1), 35-61, DOI: https://doi.org/10.1080/10291954.2022.2101328 (2022).
Han, J., Yin, H., & Wang, J. A case study of faculty perceptions of teaching support and teaching efficacy in China: characteristics and relationships. Higher Education. 76, 519-536, DOI: https://doi.org/10.1007/s10734-017-0223-0 (2018).
Kartsoni E., Bakalis N., Markakis G., Zografakis S. M., Patelarou E., Patelarou A. Distance Learning in Nursing Education during the COVID-19 Pandemic: Psychosocial Impact for the Greek Nursing Students-A Qualitative Approach. Healthcare (Basel, Switzerland). 8(1178), 1-12, DOI: https://doi.org/10.3390/healthcare11081178 (2013).
Miah M. S., Singh J. S. K., Rahman M. A. Factors Influencing Technology Adoption in Online Learning among Private University Students in Bangladesh Post COVID-19 Pandemic. Sustainability. 15(4), 1-12, DOI: https://doi.org/10.3390/su15043543 (2023).
Florence M., Chuang W., Annika J., Birgit M., Sonja G. Examining Faculty Readiness to Teach Online: A Comparison of US and German Educators. European Journal of Open, Distance and E-Learning. 22(1), 53-69, DOI: https://doi.org/10.2478/eurodl-2019-0004 (2019)
Mumford, S., Dikilitas, K. Pre-service language teachers reflection development through online interaction in a hybrid learning course. Computers & education. 144(1), 103706, DOI: https://doi.org/10.1016/j.compedu.2019.103706 (2020).
Junus, K., Santoso, H.B., Putra, P.O.H., Gandhi, A., Siswantining, T. Lecturer readiness for online classes during the pandemic: A survey research. Education Sciences. 11(3), 139, DOI: https://doi.org/10.3390/EDUCSCI11030139 (2021)
Schwartz, S.H.Normative influences on altruism. Advances in Experimental Social Psychology. 10, 221-279, DOI: https://doi.org/10.1016/S0065-2601(08)60358-5 (1977)
Ajzen, I. The theory of planned behavior. Organizational Behavior and Human Decision Processes. 50(2), 179-211, DOI: https://doi.org/10.1016/0749-5978(91)90020-T (1991)
Pedro, E.M., Alves, H., & Leitão, J. In search of intangible connections: intellectual capital, performance and quality of life in higher education institutions. Higher Education. 83, 243-260, DOI: https://doi.org/10.1007/s10734-020-00653-9 (2022)
Singh, P., Alhassan, I., Binsaif, N., Alhussain, T. Standard measuring of E-Learning to assess the quality level of E-Learning outcomes: Saudi electronic university case study. Sustainability. 15(1), 844, DOI: https://doi.org/10.3390/SU15010844 (2023).
Ader, M., Hassane, S., Van, B.J., Vermeulen, M. Comparing metacognitive regulation and socially shared metacognitive regulation in face-to-face and online learning settings in ill-structured problem solving. Learning, Culture and Social Interaction. 39, 100684, DOI: https://doi.org/10.1016/J.LCSI.2022.100684 (2023).
Glassman, M., Kuznetcova, I., Peri, J., Kim, Y.H. Cohesion, collaboration and the struggle of creating online learning communities: Development and validation of an online collective efficacy scale. Computers and Education Open. 2, 100031, DOI: https://doi.org/10.1016/j.caeo.2021.100031 (2021)
Vannini, V., Alberti, S., Epifani, C., Valentini, O., Ferri, P. The effects of online Team-Based Learning on undergraduate nursing students' performance, attitudes and accountability during COVID-19 pandemic. Acta bio-medica : Atenei Parmensis. 93(6), e2022346, DOI: https://doi.org/10.23750/ABM.V93I6.13769 (2023)

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Research on Factors Affecting Online Learning Stickiness among College Students from the Perspective of NAM-TPB Integration

Status:

Version 1

Abstract

Figures

1 Introduction

2 Hypothesis Development and Theoretical Framework

3 Research Methods

4 Results

5 Discussion

6 Conclusions

Declarations

References

Additional Declarations

Status:

Version 1