Sage Journals: Discover world-class research

Abstract

During the COVID-19 pandemic, Online Tutoring Platforms (OTPs) have been applied extensively in education in China. The aim of this study is to investigate the determinants that influence college students’ behavioural intention of using OTPs to acquire knowledge. This study also explores how students’ attitude change in using OTPs five months after the outbreak of the COVID-19 pandemic based on a hierarchical component model which is combined by the Social Impact Theory (SIT) and Unified Theory of Acceptance and Use of Technology (UTAUT). Partial least squares (PLS) analysis is employed to analyse the data collected from 1133 students in Mainland China. The results of the analysis indicate that social impact consists of three dimensions (compliance, identification, internalization) significantly influences college students’ attitude toward OTPs and further affects college students’ behavioural intention toward OTPs usage. Furthermore, performance expectancy and effort expectancy also positively affect students’ behavioural intention toward using OTPs to acquire knowledge. This study makes several suggestions for universities to encourage students using OTPs to cope with the situation of Covid-19 pandemic and for educators to promote online tutoring for reforming universities in the future.

Keywords

Online Tutoring Platform Social Impact Theory COVID-19 UTAUT

Introduction

In educational technologies, Online Tutoring Platform (OTP) refers to using Artificial Intelligence Technology (AIT) to help students acquire knowledge without face to face teaching. At present, with the rapid growth of educational information technology, online education has arisen as a substitute and/or supplement to conventional offline teaching. After the outbreak of COVID-19 pandemic, online education which brings convenience to students for learning courses anytime and anywhere has become mainstream. Online education is enabled by means of OTPs that integrate pre-recorded online courses and live teaching services into simple solutions. “Suspension of Classes and Non-stop Learning” is a fresh scheme proposed by the Ministry of Education of China to deal with the situation of postponement of spring semester due to the prevention and control of the pandemic. As of May in 2019, 1454 colleges across China have applied OTPs to cope with the situation of COVID-19 pandemic. 1.03 million teachers offered 1.07 million theoretical and experimental courses online. The number of college students adopting OTPs has skyrocketed during the last 5 months. A total of 17.75 million college students used OTPs to participate in online learning.

In previous studies, scholars have employed different technology acceptance models, such as the Unified Theory of Acceptance and Use of Technology (UTAUT), to identify the factors influencing students in accepting OTPs.^{1–5, 19} However, they mainly focused on technological determinants. During COVID-19, the government implemented an online learning policy through an executive order, which has formed a strong social influence which may inspire college students’ willingness to adopt OTPs. On considering OTPs is an emerging means of educational information technology, researchers may not realize the importance of social influence in the decision-making process of accepting new technology. However, the decision process is complicated and always preconditioned by societal and cultural assumptions.⁶ The mechanism of social influence on college students’ attitude toward OTPs usage under certain circumstance still needs to be deeply investigated.

This study attempts to identify the determinants for adopting OTPS as well as to gain a clear understanding of the mechanism of social impact on college students’ attitude and behavioural intention toward using the OTPs. This study devotes efforts to the following theoretical contributions. Firstly, this study plays a vanguard role that integrates the Social Impact Theory (SIT) with the UTAUT model to devote a synthetic second layer model to explain how the social impact would influence college students’ behavioural intention toward using the OTPs. The combination of the SIT and UTAUT diversifies researchers’ focus on implementing the systems that are social-related. Secondly, as China was the first hit by the COVID-19 pandemic, and its countermeasures have achieved great effects. Thus, practices of colleges in China provide sufficient evidence of online learning strategies that are significant contributing elements for educational sustainability during the COVID-19 pandemic. Moreover, research of experience in China could be of reference for other countries to cope with the pandemic. Thirdly, even after COVID-19 pandemic, online teaching will be an indispensable teaching method in the future. Therefore, the results of this study can provide recommendations for OTPs developers to improve the design and content, which can enable students to acquire knowledge better from the OTPs. At the same time, the use of OTPs can help the teachers and the students to learn while teaching.

The structure of this paper is organized as follows. In section experimental details, which offers a short overview of SIT and UTAUT and proposes a research model to evaluate the important determinants that influence college students’ behavioural intention of using OTPs. Six hypotheses are set up for testing, and an empirical survey is conducted. Then, the research results of the model and discussion are presented, at the end of this study, a conclusion is provided.

Experimental details

Literature review

Social impact theory

In 1958, Kelman⁷ brought up a critical issue that there are three processes of attitude change: compliance, identification, and internalization. Compliance is acceptance because of rewarding gaining or punishment avoiding. Identification embodies acceptance of influence and goal attachment to maintain satisfying relationships with peers. Internalization portrays acceptance of individuals’ value system being congruent with induced behaviour. The framework can be applied directly to various forms of social impact on attitudes and subsequent actions. Attitude toward behaviour refers to the feelings that an individual likes or dislikes a target behaviour.⁸ According to the results of the literature research, there is no research using SIT to interpret college students’ intention to adopt the OTPs in the context of the COVID-19 pandemic at present. In such particular circumstances, a crucial issue in OTPs research relates to three critical questions that are brought about by social impact. Firstly, what determinants are important for college students’ behavioural intention toward using OTPs. Although there is more research have been done on the influencing factors of the college students’ intention, the research in COVID-19 pandemic context is rare. Secondly, whether college students’ attitude toward OTPs would be changed; Thirdly, if there was a change, what kind of change it is. Is it a change that happens because a college student accepts the influence under pressure due to expecting rewards or avoiding punishment? or is he/she eager to build desired relationships just to fit in as an act of conforming? or the teaching belief of OTPs is intrinsically rewarding? To put it in other terms, is it a more lasting change (if any) of college students’ attitude and conviction toward OTPs even after COVID-19 pandemic? Only if we understand the essence and depth of changes (if any), we can make meaningful predictions about subsequent use behaviour and willingness to use continuously even without surveillance. Although attitude toward behaviour and behavioural intention of college students to adopt the OTPs to learn courses are crucial to ensure the tutoring effects and college students’ willingness to use continuously, most studies focus on “what” determinants have influence on college students’ behavioural intention toward using OTPs.^1,2,5 College students’ attitude change toward using OTPs and how the changes (if any) happen have received little scrutiny in research. Hence, SIT is integrated into this study to initiate an alternative way to understand “how” the determinants impact college students’ behavioural intention to adopt OTPs and what college students’ cognitive state is right now toward online education through the OTPs.

Unified theory of acceptance and use of technology

In 2003, Venkatesh et al.⁶ studied and compared advantages and disadvantages among eight technology acceptance models to bring up the UTAUT model. There are many previous research studies about students’ behavioural intention toward OTPs based on technology acceptance theories. In the wake of educational information technology, more and more researchers adopt UTATU to explore users’ behavioural intention toward online learning, including OTPs. There is a consensus that social impact is a critical direct determinant of behavioural intention toward online learning. Many researchers have used the models to explore the adoption of online learning systems. Lai and Lai² extended UTAUT to explore influencing factors that affect students’ willingness to use the electronic schoolbag systems in Mainland China. Radovan and Kristl⁴ integrated UTAUT and community of inquiry (CoI) framework into a combined model to study teachers’ behavioural intention and use behaviour in a virtual classroom in the Republic of Slovenia. Lakhal and Khechine⁹ enriched UTAUT to understand the adoption of desktop videoconferencing (DVC) in higher education. Yang et al.⁵ employed UTAUT to investigate the determinants of “Flash Class” – Mobile Mini-lessons in Macau. Aini et al.¹ employed UTAUT to explore the motivation of students’ adoption of iLearning, and all the research above concluded that social influence (similar with social impact) positively affects users’ behavioural intention toward online learning. However, on the one hand, most of these studies only focused on the direct effect of social impact on students’ behavioural intention toward using OTPs. There are no studies that pay attention to apply this theory with the consideration of changes in college students’ psychological cognitive state influenced by social impact. On the other hand, there are no studies that have discussed whether college students’ attitude toward using OTPs will be affected or not under the circumstance of COVID-19 pandemic.

Previous studies of OTPs mainly discussed what determinants and to what extent the determinants would affect students’ intention toward using the OTPs. For example, the two most commonly used models are the Technology Acceptance Model (TAM) and the UTAUT model.⁶ TAM includes two key determinants, perceived ease of use (equal to effort expectancy) and perceived usefulness (equal to performance expectancy), and the UTAUT includes four elements, effort expectancy (EE), performance expectancy (PE), facilitating conditions (FC) and social influence, respectively. There is no relevant study found in a literature review to discuss the mechanism of social influence on college students’ attitude toward behaviour to adopt the OTPs under such special circumstances.

Research model

Based on the theories of the SIT and the UTAUT, social impact (SI) may have a direct positive effect on college students’ behavioural intention (BI) toward using OTPs and indirect effect on college students’ BI toward using OTPs through attitude. The SI consists of three dimensions: compliance, identification, and internalization. PE, EE, and FC may also have positive effects on college students’ BI toward using OTPs. The following hypotheses were put forward. Figure 1 shows the research model.

Figure 1.

Research model.

H1: Social impact positively affects college students’ attitude toward the adoption of OTPs

H2: Attitude positively affects college students’ behavioural intention to adopt OTPs.

H3: Social impact positively affects college students’ behavioural intention to adopt OTPs.

H4: Facilitating conditions positively affect college students’ behavioural intention to adopt OTPs.

H5: Effort expectancy positively affects college students’ behavioural intention to adopt OTPs.

H6: Performance expectancy positively affects college students’ behavioural intention to adopt OTPs.

Data collection

The questionnaire investigation was carried out in Mainland China from February to May in 2020 when all colleges were preparing for the long-distance tutoring program to organize online tutoring activities, and all students received notice of adopting online tutoring platforms at home in order to fight the coronavirus pandemic. Firstly, the questionnaire was issued to 30 responsible college student leaders and they gave both positive and negative feedback. Then, the questionnaire was adjusted and redistributed to another 50 college students to conduct a pilot test. After the pilot test, the official questionnaire was delivered through a professional platform what is named “Wenjuanxing”. The survey promotion campaigns were launched via a lot of WeChat and Mobile QQ groups from colleges around Mainland China which were two main principal channels for educators and students to communicate with each other there. Finally, a total of 1202 questionnaires were gathered. However, 69 respondents were found to give wrong or contradictory answers, leaving 1133 questionnaires as effective for analysis. The measurable items for FC, EE, PE and BI were adopted from Venkatesh et al.⁶ The measurable items for ATT are adopted from Davis.¹⁰ The measurable items for compliance, identification, and internalization were adopted from Sutton and Harrison.¹¹ Table 1 shows the personal profiles of the respondents.

Table 1.

Summary of respondent background (N = 1133).

	Frequency	Percentage
Gender
Male	544	48.01
Female	589	51.99
Age
18 and under	104	9.18
19–21	902	79.61
22 and above	127	11.21
Grade
Grade 1	579	51.10
Grade 2	347	30.63
Grade 3	207	18.27

Results

The SmartPLS version 3.2.8 was used to analyse the data.¹² PLS-SEM has been applied in many research areas, for example, information and communication technology (ICT)³ and behaviour sciences.¹³ One of the reasons for choosing PLS-SEM is that the data doesn’t have to be normally distributed.¹⁴ Another reason is that this study adopts a complicated second-layer Hierarchical Component Model (HCM) which consists of reflective-reflective components and PLS-SEM can conduct analysis on complex models with formative and/or reflective constructs.¹⁵ Furthermore, this is exploratory research to study the complex relationships that exist among the variables. Under such circumstance, PLS-SEM is appropriate.¹⁶

Validity and reliability

Table 2 presents the descriptive statistics including the means, standard deviations, Kurtosis, Skewness, and PLS loadings of the measurable items of the research model. The indicator reliability values ranged from 0.762 to 0.950. The indicator reliability was proved because all the loadings exceeded the proposed level of 0.70.¹⁴ Table 3 shows the Cronbach’s Alpha (CA) and the Composite Reliability (CR) values of all constructs. The internal consistency reliability was verified because CA and CR values were above 0.7.¹⁷ The values of Average Variance Extracted (AVE) were higher than 0.5, so the convergent validity was confirmed.¹⁷ The “square root” of AVE of each construct was higher than the correlations among the constructs, so the discriminant validity was proved.¹⁸

Table 2.

Means, standard deviations, and PLS loadings.

Item	Mean	SD	Kurtosis	Skewness	PLS loading
Comp1	4.859	1.314	0.410	–0.318	0.882
Comp2	4.795	1.192	0.736	–0.238	0.871
Comp3	4.813	1.153	0.871	–0.174	0.774
INT1	4.477	1.189	1.155	–0.349	0.762
INT2	4.658	1.111	1.301	–0.280	0.776
INT3	4.750	1.058	1.400	–0.138	0.776
INT4	4.673	1.073	1.477	–0.175	0.776
INT5	4.757	1.107	1.135	–0.132	0.854
BI1	4.776	1.184	1.257	–0.380	0.941
BI2	4.738	1.181	1.084	–0.278	0.950
BI3	4.788	1.132	1.413	–0.316	0.939
PE1	4.584	1.243	0.545	–0.153	0.926
PE2	4.615	1.224	0.637	–0.116	0.930
PE3	4.732	1.168	0.895	–0.211	0.913
ATT1	4.816	1.185	0.560	–0.057	0.844
ATT2	4.566	1.154	0.859	–0.086	0.892
ATT3	4.521	1.135	0.875	–0.040	0.888
FC1	4.875	1.163	1.225	–0.422	0.913
FC2	4.868	1.167	1.310	–0.414	0.917
FC3	4.959	1.132	1.216	–0.332	0.890
EE1	4.938	1.170	0.740	–0.251	0.857
EE2	4.876	1.181	0.551	–0.143	0.911
EE3	4.855	1.170	0.675	–0.158	0.908
EE4	4.749	1.165	0.834	–0.234	0.861
IDI1	4.518	1.184	1.119	–0.291	0.906
IDI2	4.380	1.254	0.743	–0.171	0.906
IDI3	4.605	1.217	0.981	–0.211	0.908

Note: (1) Comp: compliance; IDI: identification; INT: internalization; ATT: attitude; EE: effort expectancy; PE: performance expectancy; FC: facilitating conditions; BI: behavioural intention.

Table 3.

Reliability, validity, and corrections of the constructs.

	Cronbach’s α	CR	AVE	ATT	BI	Comp	EE	FC	IDI	INT	PE
ATT	0.847	0.907	0.765	0.875
BI	0.938	0.960	0.889	0.760	0.943
Comp	0.796	0.881	0.712	0.753	0.740	0.844
EE	0.907	0.935	0.783	0.745	0.731	0.713	0.885
FC	0.892	0.933	0.822	0.740	0.724	0.735	0.739	0.907
IDI	0.892	0.933	0.822	0.752	0.724	0.749	0.687	0.785	0.907
INT	0.893	0.922	0.703	0.790	0.807	0.735	0.720	0.704	0.708	0.839
PE	0.913	0.945	0.852	0.809	0.767	0.754	0.801	0.758	0.794	0.738	0.923

Note: AVE: average variance extracted; CR: composite reliability.

Testing of hypotheses

A bootstrapping analysis via SmartPLS was run from 1133 responses to 5000 samples, for the purpose of measuring the path coefficients among variables of the research model. Figure 2 below presents the testing results of the research model.

Figure 2.

The results of PLS analysis for the second-order model.

According to the results of PLS analysis, the p-value of SI on ATT is less than 0.05. And the p-values of ATT, SI, EE, and PE are less than 0.05 too, so the hypotheses H1, H2, H3, H5, and H6 are accepted. However, the p-value of FC is higher than 0.05, so H4 is rejected. The value of R-square for the research model is 73.09%. The value of the coefficient of social impact (β = 0.557, p-value < 0.001) is higher than the coefficients of attitude (β = 0.082, p-value = 0.027), effort expectancy (β = 0.117, p-value = 0.004), and performance expectancy (β = 0.108, p-value = 0.014) on BI.

Table 4 shows the total effects of ATT, EE, FC, PE, and SI on BI.

Table 4.

Total effect of each factor on the behavioural intention.

	BI
ATT	0.0817*
EE	0.1169**
FC	0.0425
PE	0.1084*
SI	0.6263***

*p-value < 0.05, **p-value< 0.01, and ***p-value < 0.001.

Discussion

This study reveals that the behavioural intention to adopt OTPs can be predicted by the proposed model (R² = 0.731). Social impact, attitude toward behaviour, effort expectancy, and performance expectancy are the determinants of user behavioural intention toward using the OTPs. The total effects of SI, ATT, EE, and PE are 0.626, 0.082, 0.117, and 0.108, respectively. This implies that it is important for colleges and related institutions to create a social atmosphere to encourage students to adopt OTPs, so it helps to promote the use of OTPs for online and learning effectively during the Covid-19 pandemic period. It shows that students suppose the design of an OTP should be interface friendly, easy to use, and convenient. Of these four drivers, SI is the strongest predictor of user behavioural intention toward using the OTPs for college students. This shows that SI plays a critical role in predicting college students’ behavioural intention toward using the OTPs.

Besides, this study seeks to determine significant factors influencing college students’ attitude toward using OTPs. The results reveal that SI consists of three dimensions (compliance, identification, and internalization), from which, internalization is found to have the largest reflective loading on SI (β = 0.929, p-value < 0.001). Compliance and identification are observed to have almost the same reflective loading on SI (β = 0.891, p-value < 0.001; β = 0.888, p-value < 0.001, respectively).

Moreover, the effect of social impact on students’ attitude toward OTPs is 0.848. Thus, students’ attitude is affected by social impact. Students should understand the fact that if they do not use the OTPs, they may not acquire adequate academic credits. Since students usually attend classes through remote access simultaneously, complete and present group assignment via the same online tutoring platform, they are inclined to behave properly as they are afraid of losing face in front of their classmates.

As time goes on, functions of the OTPs have been improved increasingly, such as virtual classroom, meetings initiating, text and voice interaction, screen sharing, etc. Online teaching has been popularized and become a routine. OTPs become highly acknowledged with the fundamental change of students’ psychological perception shift toward OTPs. Students tend to think that the value system of OTPs is consistent with theirs. This study verifies the assumption and reflects present conditions in Mainland China. Although the pandemic situation is under control, risks still exist. College students are more and more accustomed to using OTPs to learn their courses.

Conclusions

In conclusion, the main purpose of this research is to figure out the determinants of students’ acceptance of OTPs during the pandemic period. This study examines OTPs systematically and used the SIT and UTAUT as two fundamental theoretical frameworks for examining the five determinants (SI, ATT, EE, PE, FC) of OTPs and their impacts.

The study results show that SI, ATT, EE, PE are determinants of students’ behavioural intention toward using the OTPs. The results of the second-layer model further indicate that SI consists of three dimensions (compliance, identification, and internalization) positively and significantly affects the ATT and BI toward OTPs.

However, this study has a few limitations and the findings should be read discreetly. Firstly, this study was conducted online particularly for college students in Mainland China. Future research should involve other regions, countries and university students, etc. Secondly, this study explores SI as one of the most significant determinants of using OTPs. However, SI can be divided into objective factors and subjective factors that may influence the process of compliance, identification, and internalization. Future studies are proposed to focus on what happens during the process and how the process happens. Thirdly, this study collected data from students who used no less than one kind of OTPs for research. This study didn’t compare the effects of different OTPs. Further studies are recommended to investigate this area.

Finally, online education should not only satisfy students’ acquisition of knowledge, but also become a broad stage to stimulate curiosity and create potential. Two crucial issues are brought up, one is whether the online tutoring scheme is effective to stimulate college students’ intention to use the OTPs. Another issue is how to carry out effective interactive design in online teaching so that students can change from passive observers to active thinkers. This study only tested the former issue and concluded that experience in China could be used for reference. However, the latter issue remains to be studied.

Footnotes

Author's Note

Jun Wu is also affiliated with Guangzhou PIRT Innovation and Entrepreneurship Research Institute.

Declaration of conflicting interests

The author(s) declared no potential conﬂicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported in part by the Youth Program of The Education Science Planning Department in Guangzhou under Grant 202012523, Major Projects of Guangdong Provincial Department of Education under Grant 2017GWTSCX028, Innovation and Entrepreneurship Project of Guangzhou Education Bureau under Grant 2020KC018, Innovation and Entrepreneurship Project of Guangzhou Education Bureau under Grant 2019PT102, Major Scientific Research Platform Cultivation Project of Guangzhou Panyu Polytechnic under Grant 2017Y01PY.

ORCID iDs

Jijuan Cao

Ting Yang

Ivan K-W Lai

Jun Wu

References

Aini

Rahardja

Hariguna

The antecedent of perceived value to determine of student continuance intention and student participate adoption of ilearning. Proc Comp Sci 2019; 161: 242–249.

Lai

IKW

Lai

DCF.

Student acceptance of electronic schoolbag systems: an empirical study in China. In: Cheung S.K.S., Fong J., Fong W., Wang F.L., Kwok L.F. (eds) The proceedings of the international conference on hybrid learning and continuing education. Berlin, Heidelberg: Springer, 2013, pp.334–344.

Lai

IKW

Liu

The consumer acceptance of smart product-service systems in sharing economy: the effects of perceived interactivity and particularity. Sustainability 2019; 11: 928.

Radovan

Kristl

Acceptance of technology and OTP impact on teachers’ activities in virtual classroom: integrating UTAUT and CoI into a combined model. Turkish Online J Educ Technol-TOJET 2017; 16: 11–22.

Yang

Zhong

Mok

, et al. 2018. The acceptance of “flash class” – mobile mini-lessons through WeChat. In: Cheung S., Lam J., Li K., Au O., Ma W., Ho W. (eds) The proceedings of the international conference on technology in education. Singapore: Springer, pp. 101–108.

Venkatesh

Morris

Davis

, et al. User acceptance of information technology: toward a unified view. MIS Quart 2003; 27: 425–478.

Kelman

HC.

Compliance, identification, and internalization three processes of attitude change. J Conflict Resol 1958; 2: 51–60.

Ajzen

Fishbein

1980. Understanding attitudes and predicting social behaviour. Englewood Cliffs, NJ: Prentice-Hall.

Lakhal

Khechine

Relating personality (big five) to the core constructs of the unified theory of acceptance and use of technology. J Comput Educ 2017; 4: 251–282.

10.

Davis

FD.

Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quart 1989; 13: 319–340.

11.

Sutton

Harrison

AW.

Validity assessment of compliance, identification, and internalization as dimensions of organizational commitment. Educ Psychol Meas 1993; 53: 217–223.

12.

Ringle

Wende

Becker

J-M.

2015 SmartPLS 3. Boenningstedt: SmartPLS GmbH, http://www.smartpls.com ( accessed 2021, ).

13.

Bass

Avolio

Jung

, et al. Predicting unit performance by assessing transformational and transactional leadership. J Appl Psychol 2003; 88: 207–218.

14.

Hair

Hult

GTM

Ringle

, et al. 2016. A primer on partial least squares structural equation modeling (PLS-SEM). Newcastle Upon Tyne, UK: SAGE.

15.

Hair

Ringle

Sarstedt

Partial least squares structural equation modeling: rigorous applications, better results and higher acceptance. Long Range Planning 2013; 46: 1–12.

16.

Wong

KKK.

Partial least squares structural equation modeling (PLS-SEM) techniques using

SmartPLS. Market Bull 2013; 24: 1–32.

17.

Bagozzi

On the evaluation of structural equation models. JAMS 1988; 16: 74–94.

18.

Fornell

Larcker

DF.

Evaluating structural equation models with unobservable variables and measurement error. J Market Res 1981; 18: 39–50.

19.

Cao

Shang

Mok

, et al. The impact of personal innovativeness on the intention to use cloud classroom: an empirical study in China. In: Cheung S., Jiao J., Lee LK., Zhang X., Li K., Zhan Z. (eds) The proceedings of the international conference on technology in education. Singapore: Springer, 2019, pp.179–188.

Is online education more welcomed during COVID-19? An empirical study of social impact theory on online tutoring platforms

Abstract

Keywords

Introduction

Experimental details

Literature review

Social impact theory

Unified theory of acceptance and use of technology

Research model

Data collection

Results

Validity and reliability

Testing of hypotheses

Discussion

Conclusions

Footnotes

Author's Note

Declaration of conflicting interests

Funding

ORCID iDs

References