Measuring the impact of COVID-19 on distance learning for educational sustainability

Abstract The current health emergency has a significant effect on the educational community in unprecedented manners. Most of the educational activities are conducted via online to engage their students by keeping social distance. Most instructors and students face huge trouble using digital systems because they have no prior training about online learning systems. The aim of this study is to measure the student attitude toward online teaching during COVID-19. This study is different from existing literature, where online enrollment of courses is an option with a reliable learning management system (LMS) for students and teachers rather than an educational emergency. During this health emergency, we conducted an online survey to collect student response from 15 higher education institutions across Pakistan. The responses of 525 students were analyzed by using partial least square (PLS). We observed that the limited interaction and lesson plan have a significant impact on student attitude, while interaction with teachers has no impact on student attitude. In this study, we also provide a framework to use existing resources and LMS to improve the educational sustainability during COVID-19 pandemic.


Introduction
In 21th century, many health emergencies have occurred which highly affect the living and working environment. But, the COVID-19 pandemic has an un-precedent impact with the closure of whole educational, economic, social, and religious activities. Due to the high spreadness rate of this Novel Virus, Most educational institutions are close to keep the social distance possible and follow the standard operating procedures (SOPs) issued by World Health Organization (UNESCO, ABOUT THE AUTHOR Dr. Fariha Sohil, Assistant Professor at The Women University, Multan is working in the field of smart learning and trying to remove the educational barriers as much as possible. Dr. Muhammad Sohail, Assistant Professor at The University of Narowal is also working to facilitate and slove the social educational back locks. Our research group Sohail et al. believe in education for everyone.

PUBLIC INTEREST STATEMENT
This manuscript deals with the impact of online learning in the era of COVID-19. The sudden lockdown in underdeveloped countries creates the panic situation among learners and instructors because most learners and instructors have no prior experience of using digital media to receive/transfer instruction by using android applications. The detailed discussion about online learning based on comprehensive research methodological background makes it unique among others. 2021). Education for all is the one of possible ways for the development and sustainable growth of society. For the systematic engagement of students with educational activities, most of the higher education institutions start an online learning management system to save time and health by following SOPs (Sahu, 2020). For sustainable educational outcomes, students' attitude towards online learning and their inputs during online learning is very important because physical learning is replaced by online learning for the first time worldwide. Furthermore, COVID-19 pandemic can open a new dimension of educational activity. We are hopeful the current online learning mechanism may continue by merging new elements of educational activity after the COVID-19 pandemic.
Most educational courses are moved suddenly from the physical learning environment to online learning system (UNESCO, 2021). Most faculty members have no prior online teaching experience and they suffer from lack of training and proper use of online teaching and technical tools to deliver their lecture effectively (Scarborough, 2021). Most higher education institutions are not well equipped, they suffer from lack of proper administrative and technical support to conduct online classes (Mirzajani et al., 2016). An ideology is established between some researchers that the role of teachers in online learning and physical learning environments is similar (Wray et al., 2008). The recent literature about the role of teachers inside the classroom provides a valid point that the use of professional skills and interaction with students by a teacher in a physical learning environment is high as compared to online learning systems. The online learning system demands the professional and technical knowledge from teachers about virtual classroom software to manage and organize their class precisely. Most faculty members are able to prepare and deliver their lecture properly during online class (Lichoro, 2015). Downing and Dyment (2013) reported that most instructors feel that the online learning system is time consuming and the interaction between instructors to student and student to student is very low which creates the problem to cover their lesson plan.
In literature, various studies are available about the critical Success factors (CSFs) of online learning on professional online instructors. But very few studies are available on the use of CSFs for teachers having little or no professional experience to use online classes. The student satisfaction about online class, course design, instructor response, teaching style, instructor teaching skills, and online learning outcome are examined by Eom and Arbaugh (2011) and Eom and Ashill (2016). Furthermore, the participants in these studies already enrolled in online courses and expected online classes. The faculty members in these classes are well trained and equipped as compared to those who are conducting online classes during current health emergencies. The literature about the student educational sustainability about online class is very small and most of the factors are not manipulated significantly. The current health crises provide a chance to study student perceptions about physical and online classes that were never possible before. Therefore, the basic objective of this is to identify the change in learning attitude during online learning classes and their experience in the online system that impacts their overall satisfaction during COVID-19 crises.
The theoretical framework, research methodology, and discussion of statistical findings with conclusions are discussed in the coming sections.

Theoretical framework and statistical hypothesis
The quality of education in a society is analyzed by the student perceptions and learning outcomes (Garnjost & Lawter, 2019). The research model in this study considers the student attitude about their learning outcomes and their level of satisfaction during an online learning environment. Most students are focused about their learning outcome, it provides a path of success and achievement, and also encourages the sense of competency, which are the key outcomes of self determination, motivation, and engagement (Ryan & Deci, 2000), thus changing their focus and behavior (Sørebø et al., 2009;Vierling et al., 2007). Furthermore, the proposed research model considers the student interaction with their classmates and their instructors. The human (student) behavior about motivation and their personality in social sciences deal with the self-determination theory (Deci & Ryan, 2011), which suggests human interaction with each other is the basic need for social, cultural satisfaction and attitude. Thus, the self-determination theory is used as a theoretical framework to understand the online learning problems more precisely (Chen & Jang, 2010).
The variable of interest is based on commonly used elements in practical and vast variety of online multiple dimensions. Majority of them are obtained from the model and there of online learning, while some of them utilize existing learning methodologies on online learning. From all of them, the satisfaction of students and the learning outcomes are based on following factors, as: (1) Interaction during learning process Eom and Ashill (Eom & Arbaugh, 2011) extracted different learning models from previous literature and used them to explain the characteristics and benefits of online learning. These models are constructive in nature for learning (Piaget, 1977;Vygotsky, 1978), virtual learning system (VLS; Piccoli et al., 2001) effective model, and framework for technical learning support (TLS; Alavi & Leidner, 2001). The basic structure of learning models is the knowledge as opposed to being transferred from instructor to students. The learning process is more effective when students find their learning material at their own time and pace. The self motivation and regularization level are used to conceptualize the online learning model. The VLS and TLS are the technical side. The VLS provides the framework and technical support for the learning process via an online system. This infrastructure provides a reliable function for the assessment and evaluation of student performance and attractive communication tools, which motivate the students and instructors.
Eom and Ashill (Eom & Arbaugh, 2011) explain online learning as the open learning system for three entities, these are Students, faculty members, and VLS. They interact with each other without time limitation and learning environment to optimize the student satisfaction and their outcomes. TLS defines the AV-aids during the online learning environment to build an effective teachinglearning environment. TLS is defined as the learning system in which learners fulfill their learning needs, interact with their classmates and instructors via virtual classroom technology (Alavi & Leidner, 2001). The VLS and TLS establish the connection between students, instructor, literature, and evaluation for student performance, which are major factors, which affect the students' learning outcomes and their satisfaction. The measurement of degree of satisfaction about the learning process (Ho & Dzeng, 2010) is very important to evaluate the sustainability education system (Maki et al., 2000). In figure 1, the conceptual framework for the proposed research model is defined. Statistical hypotheses for the relative comparison are discussed in the later subsections.

Interaction
The community framework (CoI; Garrison et al., 2000) was adopted by (Gurley, 2018) to study the necessary elements required for an ideal learning environment in a blended and online-learning system and their impact on the course quality, student's achievements, and their satisfaction about learning system and perceived relative learning outcomes (Dereshiwsky, 2013). The CoI is a social learning process in a blended learning process and online learning system. This model include following main components (Gurley, 2018), which are (1) Teaching Attitude (2) Social Appearance

(3) Cognitive Ideology
Most researchers provide a valid conclusion about the existence of a relative relationship between these components with student satisfaction and their learning outcomes. Thus, the key responsibility of the instructors is to engage their students effectively in an online classroom, encourage student's participation, and assist students via direct learning (Harnegie, 2015).
The effective interaction plays a key role in physical learning, blended learning (Physical learning and online learning sessions), and online learning sessions, respectively. Social appearance means the ideology and thought the learner establishes by communicating with their age fellows, classmates, and from their community where they live (Meter & Stevens, 2000). Number of researchers suggest different methods of interaction in an educational environment such as, learner to learner interaction, interaction of learning with their instructor and their course material (Anderson, 2021(Anderson, , 2003Bernard et al., 2009;Kanuka, 2011). Most of the researchers emphasize the high impact of interaction of learners with their classmate and instructor in a learning process. The physical or online interactions promote the thinking ability (to observe critically), problem solving skills, and social understanding. These interactions provide cognitive support for learners to understand their course material more appropriately and reduce social distance and psychological stress. Interactions encourage student participation in the learning environment and strengthen student higher-order knowledge by us cognitive engagement (Muirhead & Juwah, 2004). Duncan-Howell (2010) and Matzat (2013) point out the need for regular interaction of learner and instructor during online class, these interactions establish the feeling of security and self-confidence among learners and their sense of social bond.
In the last few decades, several studies have been conducted to measure the impact of interactions and student satisfaction, and their learning outcomes. Few of them provide inconsistent results about the relationship. But most studies show that the interaction between learner and instructor commits their student with their course material and strengthens their ability to get high learning outcomes (Arbaugh & Benbunan-Fich, 2007;Jaggars & Xu, 2016). The learner to learner interaction does not promote student satisfaction and learning outcomes. Furthermore, Arbaugh and Rau (2007) suggest that the mode of interaction (learner to learner or learner to instructor) encourages learner participation and strengthens their ability to perform well in a learning environment. The interactions between students are helpful to predict student satisfaction. The major possible reason for inconsistent findings is the mode or level of interaction between participants. Limited interaction between learners creates a feeling of disconnection between instructors and their classmates. Recently, researchers use the mode of interaction to measure the quality of education. Due to this realization, in this study we use the constructive interactions to provide a clear and picture of meaningful content delivery to each online participant during COVID-19 crises. We construct following statistical hypothesis, as:

Hypothesis 1a (H 1 a):
There is a positive correlation between online student to student interaction and satisfaction level of students.

Hypothesis 1b (H 1 b):
There is a positive correlation between online student to student interaction and educational outcomes.

Hypothesis 1c (H 1 c):
There is a positive correlation between online student to instructor interaction and satisfaction level of students.

Hypothesis 1d (H 1 d):
There is a positive correlation between online student to instructor interaction and educational outcomes.

Virtual classroom software
The key responsibility of the instructor is to facilitate, engage, and deliver their online lecture significantly. In online learning there is no physical connection between learners and instructors

Figure 2. Role of Course Facilitator
and teachers need to adopt different teaching methodologies to interact with their students in online classrooms and minimize the social distance (Gillett-Swan, 2017). The availability of authentic technical support (ATS) and modern virtual class software provide an opportunity to deliver the course content precisely (Banna et al., 2015). The ATS facilitate the faculty member for timely response to student emails, timely assessment and grading system, and submit their observation report, while other learning sessions are less effective (Martin & Bolliger, 2018). Berge (2008) suggested the Instructors Role model (IRM), which changed the role of instructor from subject specialist to course facilitator and categorized it into four different categories in figure 2.
Many studies are available which investigate different aspects of course facilitator. Hosler and Arend (2012) reported the specific role of course facilitator, a dedicated facilitator engaging their class significantly, providing feedback timely, assessment and evaluation techniques are up-tomark, and enclosing student participation. (2015) develop a measurement scale for assessing the online teacher role and attitude during online classes, and utilize this scale to examine student perceptions about online and blended learning systems. This model includes the five constructive components, course planner, open participation, social facilitator, technical support, and evaluation organizer. Learn to receive timely responses from their instructor rather than delay (Kleij et al., 2012). Moreover, Arbaugh (2010) reported that the online teacher has two major responsibilities, (i) teaching and (ii) timely feedback about learner queries. These factors directly affect the student learning outcomes and their educational sustainability about online learning. The teaching attitude includes direct instructions about course material to establish meaningful learning outcomes. The timely online feedback about student queries create the feeling among students that they are connected with their instructor, institution, and course material and also minimize the social distance during COVID-19 crises. We construct the following statistical hypothesis, as:

Hung and Chou
Hypothesis 1e (H 1 e): The availability of reliable virtual classroom software positively related to student satisfaction.

Hypothesis 1d (H 1 f):
The availability of reliable virtual classroom software positively related to online learning outcomes.

Course outline/plan
The cognitive model shows that the learning process is more effective, when the course is designed as per the requirements of the student. The teaching style clicks the wide variety of learner style, which results in more satisfied learning outcomes. The available latest technologies make it possible to deliver course material with more attractive digital features e.g., AV-aids. From online school thoughts, Martin et al. (2019) found that the online course design, online teaching method, online assessment and evaluation tools, and timely feedback are the key components for the effective online learning process. The selection of these components is obtained for the existing literature. The designing of course content as per scholarly guidance is a basic component for the satisfied teaching process and perceived learning outcomes (Moore, 2016). Moore and Kearsley (2011) shows that students with cognitive thoughts are more active learners and understand prior knowledge. Keller (1983) found different elements of satisfaction during online class, the availability content and the course content is designed as per the requirements of the students i.e. student centered approach (Johnson & Johnson, 2002).
Consequently, the course plan and course material provide different elements that influence students perceptions and learning outcomes. We consider the following statistical hypothesis.

Hypothesis 1 g (H 1 g):
The design of course content is positively correlated with student satisfaction.

Hypothesis 1 h (H 1 h):
The design of course content is positively correlated with student learning outcomes.

Student behavior
In the field of education, motivation and encouragement is the fundamental step of student learning and educational learning outcomes. The course satisfaction directly correlated with student satisfaction (Fujita-Starck & Thompson, 1994) and outcomes (Eccles et al., 1993). Another important factor, which has been commonly applied is the self-determination in the motivation theory (Chen & Jang, 2010). Furthermore, this theory is utilized as the average of study in various primary factors of the teaching and learning environment (Guay et al., 2008). Based on this theoretical background, many researchers reported that level of self-determination and self-motivation strengthens the student satisfaction and their outcomes (Noels et al., 2000;Pae, 2008).
The existing literature provides the valid conclusion about the existence of negative attitudes due to low motivation and self determination among students (Connell & Ryan, 1985;Peled et al., 2019). In current health crises, the online learning environment provides an opportunity to support teaching pedagogy for student satisfaction, better interaction, and provide learning material beyond their time and space. In this research model, we believe that these components facilitate students to improve their satisfaction and educational outcomes. Thus, this environment positively changes the student behavior toward online education and improves the sustainability toward sustainable education. For these situations, we formulate following hypothesis, as:

Hypothesis 1i (H 1 i):
The positive behavior toward online education is positively correlated with student satisfaction.

Hypothesis 1 j (H 1 j):
The positive learning attitude is directly correlated with student satisfaction.

Hypothesis 1k (H 1 k):
The positive behavior toward online education is positively correlated with student learning outcomes.

Hypothesis 1 l (H 1 l):
The positive learning attitude is directly correlated with student learning outcomes.
Recently, Baber (2020) found that the learning outcomes have a positive impact on the student satisfaction in a large geographical study by an excellent effort for learning performance in an online learning system. In favor of student satisfaction, we construct the following statistical hypothesis, as:

Hypothesis 1 m (H 1 m):
The student satisfaction is directly correlated with student learning outcomes.

Research methodology
The research methodology is categorized into main parts, as shown in figure 3.

Data collection
The sample size of 600 undergraduate students is selected from the 15 higher education institutions, 12 public sector institutions and 3 private sector institutions. All the participants attend physical learning (Spring semester-2020) and online learning environment (Fall semester-2019) in the educational year 2019-2020. The target population were identified by personal relation, social and educational network, then, the online survey performed was sent via email and hyperlink Google Form to obtain their response. The pre-test of this online survey is conducted on 50 students belonging to 4 public sector and 1 private sector higher education institutions. After pretest, few minor changes regarding wording of the questionnaire were made by the authors. Participants in pre-test respond to all questions (no question in this survey sensitive in nature) in online surveys.
In the online survey (see Table 1), 64.67% are female and 35.33% are male participants. Out of 600 sampled units, 58.67% are senior students (enrolled in 5 th and above semester) and 41.33% are fresher's (2 st and above semester). The research tool consider their educational belonging (faculty), 38.67% belong to faculty of science, 27.67% belong to faculty of arts and humanities, 11.50% belong to faculty of business, 09.33% belong to faculty medical science, 04.67% from sport sciences, and 08.17% belongs to others faculty like diploma, certificates, etc.
To assess the reliability and validity of the research items by the four educational experts, who were reported on the theoretical validity of items. The factor analysis (FA) using PLS to test the measurement of the research model. The results of FA were evaluated on the reliability, discriminant validity, and convergence of research items. In Table 3, the item loading, Coranbach's alphas (CA), composite reliability (CR), and average variance extracted (AVE) is given. It is found that the CA and CR of all the items is above than 0.7, as the accepted criterion of reliability (Nunnally & Bernstein, 1994). Furthermore, the factor loading of all items is above 0.7 and VAE for all constructs is greater than 0.5. So, it is illustrated that all the items satisfy the convergence validity (Fornell & Larcker, 1981).
We also express the correlation and the average square root of VAE in Table 4. It is observed that the AVE of all items is greater than its correlation value (Gefen & Straub, 2005), the reliable discriminant validity is also found at stage. The statistical findings of all the research items attain valid psychometric features. Table A shows the mean and stanadard deviation of research questions to understant the overall pattren of the respondents.

Statistical finding
Bootstrapping approach is followed in the PLS algorithm to obtain re-sampling (up to sample size) to evaluate the relative performance of the proposed model (Chu et al., 2015). The significance of each path is analyzed by using the t-test. Te = e coefficient of each path is shown in figure 4. The fitness of model (model selection) measures are provided by PLS. Among all of them, Standardized Root Average Square Error Residual (SRAR) seems to be a reliable tool to measure the fitness of a model for the sample of size 600 students with low value of bias. If SRAR value is equal to zero, mean perfect fit and less than 0.08 mean good fit. For our proposed model the SMAR value is 0.063. Table 5 illustrates the significance level of the test hypothesis. Out of 13 statistical hypotheses, 11 hypotheses are significant at 1% level of significance. The hypothesis H 1 a, and H 1 f are insignificant. H 1 a is the student interaction with their classmate that has no impact on learner satisfaction in the online learning education. During COVID-19 crises, students do not interact with each other effectively, so it does not affect their satisfaction level about online education.
The hypothesis H 1 f has an insignificant relationship with perceived satisfaction level and their performance, H 1 f is insignificant in nature. The limited or poor technical support during the online learning environment damages the level of students' satisfaction. Most instructors and students have no/limited prior experience to use an online learning environment. Online teaching required some  special skill to use online equipment and tools in an online classroom. The sudden educational breakdown creates the problem for everyone belonging to the educational community to move their classes online without any prior training and experience. It is still a challenging task to engage students effectively during the online class. Furthermore, e-cheating during online exams, exam paper security, and strict time limit for paper submission create trouble for the online users.

Discussion
Out of 13 statistical hypotheses, 11 found to be significant with the student satisfaction and their perceived educational outcome. The student stratification about online learning improves the learning attitude and leads to a sustainable educational environment. However, the focus of our research to find out major factors influences the learning outcomes in online learning. From our model, the interaction between students and their course content highly influences the obtained educational outcomes. The results show that if the aim of the researcher is to improve the student's perception about online education and their behavior toward it, we must try to improve their educational achievements. There are two possible paths (Eom & Ashill, 2016) to enhance these results: • Better equipment for online interaction and encouraging peer-learning • Re-design the course content and promote self-learning skills during online classes Furthermore, the technical and administrative support, online learning equipment, regular interaction between students via virtual room to share and communicate with each other improve online attitude. Online instructors should facilitate their students by promoting self-learning skills, encouraging online interaction, and providing timely course content and feedback to make the learning process easy. Instructors also promote online class groups on social media for better communication with each other. Course content and design is the major contributor to improving online learning attitude. This is also the key responsibility of the instructors to prepare course content/material and make its easy access via online. Instructors must adopt easy and smooth teaching methodology and make sure students have real-time participation during online classroom (Tsay et al., 2018). Additionally, instructors prepare a peaceful, clear, and organized online course content for the effective delivery of online lectures by using Notepad, Ms Office, Latex, or any other digital software, which are easily accessible for everyone. Most virtual classroom software are well equipped with attractive features like sending reminders for assignment submission deadlines and e-assessment of performance.
The virtual classroom software does not have the significant effect of student learning performance but is significantly correlated with degree of satisfaction. The availability of more teaching options in online classes for instructors and active response for the administration/technical department improve the student satisfaction about online education (Ladyshewsky, 2013). The online grading of student assignments and their exam paper are conducted as per the design of the course content.
The interaction between instructor and student is likely to be improving student satisfaction and their educational outcomes. Due to social distancing, physical interaction is impossible but the virtual and social media interaction e.g., Facebook and Whatsapp build the sustainability of online education. Students can place their query in these groups and find their solution at one click if the instructor provides timely feedback. For the large class size, online learning is a cost and time effective tool as compared to physical learning, because everyone gets proper attention and replaces their queries on social groups and online class chat boxes easily.
Another major component that we find from our research model, the effective participation and utilization of online teaching tools by faculty members is only possible if the online training sessions are conducted on a regular basis by the higher education institutions. The current health emergency provides an opportunity to continue the current online system in future for distance education programs (Al-Kumaim et al., 2021). Purchase of an online system can improve sustainability in the education system and decrease the workload on faculty members by e-assessment system in future (Broadbent et al., 2020).
On the other hand, the online system engages our young generation with a digital community and provides a plate forum to share their ideas with each other by e-interaction. During online education, a student might be enrolled in multiple courses because they save their traveling cost and time, dressing cost, fuel charges, food expenses, etc. to enhance their professional skills (Bocanet et al., 2021). For sustainable educational development, the traditional classroom pedagogies in online education are an interesting field in future educational research (Buil-Fabregá et al., 2019;Caird & Roy, 2018).

Limitation of the study
As like to other social science studies, our research has some limitation, these are discussed later.

Self-Selected samples
In this study most of the samples are selected on the educational and social network. We tried to enhance the quality of the survey by using diverse samples and prior literature at the design stage, unidentified samples, find-out target participants by using personal connections. Additionally, we use a caption fill option during the online survey to obtain the accurate answer "Fill the Caption: 2 × 4 = . . . " (Chu et al., 2015). If the participant is unable to write the correct answer, she/she will be unable to submit their response.

Lack of proper infrastructure
There is a lack of infrastructure to measure the change behavior about online learning and their learning attitude in COVID-19 crises. However, we consider a strict infrastructure to develop a framework for measuring the change in attitude toward online education and learning behavior.
During this research we take all the possible measures to ensure reliability and validity of research items. We perform the content analysis and also obtain the expert opinion from the different higher education experts of all the research items used in this online survey . We reverify that the research item achieves all the psychometric features by evaluating item loading, CA, CR, AVE and correlation with all items (Chu & So, 2020).

Geographical restrictions
Due to limited resources, all the participants belong to the province Punjab, Pakistan. A large study is required across the country to study the cultural diversity and generalization of statistical findings.

Conclusions
COVID-19 crises create unprecedented changes in the living and working style So et al., 2021). No one knows about the recovery from this health crisis but the continuity of educational activity is essential to engage the young generation in positive activities. The research design explores the learning perception about online education and changes the learning attitude, just after all the educational institutions can move their activities via online to keep the social distance. It is a challenging task to predict the impact of any communication disease in future. For the sustainable development of society, the complete blackout/closure of educational institutions is very costly. Thus, it is the fundamental need of society to start online education until the physical learning environment is not possible.
The most important finding of this study is the significance of student interaction with their instructor and course outline/plan. During online learning, the interaction of students with their classmates has no effect on their learning outcomes. At the end of this research we provided some valuable recommendations for the significant improvement in educational sustainability based on statistical findings.

Recommendation
This research design provides a framework for the research and teaching community to prepare their course content and teaching methodology as per the requirement and availability of virtual classroom software. Therefore, there are various research questions about the sustainable online education in higher education institutions i.e. technical skills of online users (teacher and students), availability of technical support during online class, and culture diversity among users. There is huge space to study the online learning attitude and learning behavior for the long-run educational sustainability for feature research.