Findings & Discussions

Findings can confirm or reject the hypothesis underpinning our study. These findings are obtained from our survey and analysed by using SPSS and SmartPLS.

Profiles Of Respondents

• Table 1: Demographic Profiles

Based on the demographic information in Table 1, it can be inferred that 32.9% of the respondents were males while 67.1% were females. This shows that the majority of the respondents are female respondents.

Among the total respondents, 83.2% of students are from age 19 - 21 and the rest of 16.8% of students are from age 22 - 25.

Besides, around 73.3% of respondents have enough experience as they have frequently learned in an online environment while 24.8% respondents stated that they have not much experience on online learning and only 1.9% respondents say they have none of experience in online learning.

Crosstabulation of Demographic Profile

• Table 2: Gender * Age Crosstabulation

According to Table 2, it shows that the highest respondents from both ages 19 to 21 and 22 to 25 are both female respondents with a total of 108 respondents which equal 67.1% while male respondents with only 53 respondents where it is equal to 32.9%.

• Table 3: Gender * Experience Crosstabulation

According to Table 3, the female students have all the highest respondents for the experience in online learning. The highest for enough experience in online learning comes from female respondents with a total of 51.6% while male respondents with only 21.7%. Besides, 13.7% of female respondents stated they have not much experience in online learning while for male it have only 11.2% of the respondents. With no experience in online learning, female respondents have a total of 1.9% while male respondents with 0.0%.

Structural Model Analysis

Structural model analysis is a multivariate statistical analysis technique used to examine relationship strength. This technique combines factor analysis and multiple regression analysis to examine the structural relations between items and measured variables.

• Hypothesised Model

Figure 2: Hypothesised Model

Figure 2 shows the hypothesis model of the original model and item. This is the first result of the data from the calculation of the SmartPLS based on the literature.

• Generated Model

In the Generated Model, the factor loading of each item that has value of less than 0.5 needs to be taken out from the model of framework and be deleted as it is a non-valid item. The factor loading for each of the items have to be more than 0.5. Factor loading is between each of the items and variables.

Figure 3: Generated Model for Variable Interaction

The first variable that has an item of less than 0.5 is Interaction. Item interaction 4 has a value of -0.357 and the question is “There were no opportunities for active learning in this course”. Next, interaction 8 has a value of -0.366 with a question of “There is a lack of teacher-student interaction in online classes”. Besides, interaction 9 also is a non-valid item as the item a have a value of -0.414 with the question “It is difficult to control group interaction during online classes” and lastly interaction 10 is the last of item from variable interaction has a non-valid item with a value of -0.124 with a question “Teacher- students disconnect is felt low in online classes compared to classroom method”. All of the non-valid items for the variable in interaction have to be deleted from the data analysis as it produce factor loading less than 0.5.

For our first variable which is interaction, only six items out of ten are valid. The valid items mostly discuss the frequent interaction of students in the course, regular communication with the instructor, promotion of interaction during learning activities, the opportunity for self-introduction in class, communication with other students within the course, and receiving ongoing feedback from classmates.

Figure 4: Generated Model for Variable Student Motivation

The second variable that has an item of less than 0.5 is Student Motivation. Out of the seven item, four of the items from this variable is non-valid items as the value is less than 0.5. Student motivation 1 has a value of -0.561 with the question “I do not feel motivated to participate in online class discussions” while student motivation 2 has a value of -0.571 with the question “I feel lack of motivation to take online classes”. Next, student motivation 4 has a value of -0.498 with a question “I am finding it difficult to adapt to the online teaching mode.” and student motivation 5 with value of -0.391 with question “I feel lazy and disinterested during online classes”.

Therefore, student motivation has only three valid items out of seven. Being easily distracted, producing perfect assignments, and working hard to get good grades even when the class is not liked are the valid items to measure this variable.

Figure 5: Generated Model for Variables Course Structure and Instructor Facilitation/Knowledge

The third variable, Course Structure, there are only six valid items out of eigth. Course structure 3 has a value of -0.310 with a question of “The layout of the course was disorganised” and the second item comes from course structure 4 with a value of -0.275 with a question “Course navigation was illogical”. Other items such as aligned learning outcomes and activities, clear course purpose, clear instruction of student participation, well-organized modules, the effective range of course materials, and interesting course materials are valid measurements for the variable course structure.

Next, one out of nine items under the instructor facilitation variable is eliminated as it is an invalid item to be used to measure the variable. The non-valid item which comes from instructor facilitation or knowledge is item IF1 with a question of “The instructor's feedback on assignments was not constructive”. Other items such as instructor feedback and supervision, students get to learn from feedback, instructor facilitation, responsiveness, and interaction with students are all considered valid items to measure the variable as all of them have factor loading that is greater than 0.5.

Figure 6: Generated Model for Variables Student Perceived Learning and Student Satisfaction

The next variable comes from Student Perceived Learning and Student Satisfaction with each of them has one non-valid item. Student Perceived Learning 5 has a value of -0.001 with a question of “I learned less in the course than I anticipated” while for student satisfaction 6 has a value of 0.051 with a question “There is lack of work satisfaction while taking online classes”.

For the next variable student perceived learning, eight items are valid out of nine. Promotion of achievement of student learning outcomes, pleased learning, useful skills for future, enhanced understanding, contribution to professional development, exploration of innovative teaching methods, the consciousness of learning skills, and well-learned presented class materials are the valid items for measuring this variable.

The last variable is student satisfaction. One out of eight items is invalid, leaving another seven items as valid measurements for the variable. The valid items are consist of satisfaction in the level of student interaction, the satisfaction of online classes, satisfying online classes experience, satisfaction with the instructor and content, satisfying overall online learning experience, and the online course met my needs as a learner.

• Re-specified Model

Re- specified model is more than one time of generated model. Re-specified Model is after eliminating all of the items that are less than 0.5 from the framework.

Figure 7: Re-specified Model

Based on figure 7, we can identify the highest student perceived learning based on the path coefficient. The higher path coefficient, the better it is. In this figure 7, it shows that Course Structure has the highest path coefficient with a value of 0.500 followed by Instructor Facilitation or Knowledge with a value of 0.266. Then, Student Motivation has path coefficient with a value of 0.126 and last is Interaction with a value of 0.114. All of the variables have a positive influence for the path coefficient as the value between variable to variable is positive values and does not have a negative value. In Figure 7 also, it shows the r square variable of Student Perceived Learning and Student Satisfaction. For Student Perceived Learning, the value of r square is 67.6% where in this model of framework, it is able to explain the student perceived learning of 67.6% out of 100% of the phenomena of student perceived learning while for student satisfaction the value of r square is 58.6%. This model of framework is able to explain the student satisfaction of 58.6% out of 100% of the phenomena of student satisfaction.

• Bootstrapping Model

Figure 8: Bootstrapping Model

For the Bootstrapping Model, the value of factor loading; item to variable and path coefficient; variable to variable have to be more than 1.96. We find that interaction (T-value = 2.285), student motivation (T-value = 2.926), course structure (T-value = 5.940), and instructor facilitation (T-value = 4.127) have significant positive relation with student perceived learning (p < 0.05) and student perceived learning (T-value = 21.596) have significant positive relation with student satisfaction (p < 0.05). Based on figure 8, it shows that all of the factor loading and path coefficients have a value more than 1.96 and this has been confirmed that all of the factor loadings and path coefficients are significant.

Results of Hypothesis Testing

• Table 4: Hypothesis Results

It shows in Table 4 that all of the results of the regression for the hypothesis are accepted. The higher the path coefficient in the Re-specified Model, the better it is. There is a strong influence for course structure together with instructor facilitation or knowledge on the student's perceived learning outcome. While for interaction and student motivation, it has a moderate influence on student perceived learning outcome. The student perceived learning also positively affects student satisfaction in online classes during the pandemic of COVID-19. All of the variables have a t-value more than 1.96 for the factor loading and path coefficient in the Bootstrapping Model. The item and variables that obtain more than 1.96 are confirmed as significant. It indicates that due to COVID-19, all of the variables have significant effects on education.

Findings related to the relationship between the variable interaction and student perceived learning, hypothesis H1, show a significant relationship in this case. Loneliness might hinder the students’ learning process (Sit et al., 2005 as cited in Kaufmann &Vallade, 2020). Students’ interaction such as initiating communication between students and lecturers, participating in activities that promote interaction, and receiving feedback from classmates influences student perceived learning which will then influence student satisfaction. According to Kaufmann and Vallade (2020), this element, among other factors, becomes critical in the development and delivery of high-quality online courses. Also, researchers and instructors may use moment-to-moment interactions to identify learning opportunities as mentioned in Smit et al. (2021).

Findings related to the relationship between the variable student motivation and student perceived learning, hypothesis H2, show a positive relationship. This finding counters the finding by Bulić and Blažević (2020) in the past research where they both recommended an inverse relationship between student motivation and online teaching. According to Brandmiller, Dumont, and Becker (2020), teacher expectations of students' achievement have a close relationship with student motivation. It means that students with higher motivation will be able to see the importance of learning thus motivates them to achieve greater results. In our research, we find out that being easily distracted during class, turning in perfect assignments, and working hard to earn good grades even though not liking the class, plays a role in student motivation, which will then influence student perceived learning.

Findings related to hypothesis H3 which is the relationship between course structure and student perceived learning shows the highest significant relation between the variables. This means that among four of the independent variables that influence student perceived learning, this variable influences student perceived learning the most. Items such as aligned learning outcomes and activities, clear course purpose, clear instruction of student participation, well-organized modules, the effective range of course materials, and interesting course materials are seen to be influencing the variable course structure which will then influence student perceived learning. Findings from previous research conducted by Heilporn et al. (2020), also found that having well-structured and well-paced blended-learning courses proved to be crucial for student participation. Student's behavioral and emotional participation was boosted by explicitly explaining how the blended-learning course was structured and students' anxiety and negative emotional responses were decreased by clear contact at the start of the class, and their interest in events was increased.

Findings related to hypothesis H4 which is the relationship between instructor facilitation or knowledge and student perceived learning shows a positive relationship between the variables. There is also a relatively good positive relationship between teaching presence and perceived learning in research conducted in 2020 by Caskurlu et al.. In our research, items such as instructor feedback and supervision, students get to learn from feedback, instructor facilitation, responsiveness, and interaction with students are all influencing the variable instructor facilitation. Feedback allows students to have a deeper view of the learning goals and, as a result, manage their goals as mentioned in Chan and Ko (2020).

The last findings are related to hypothesis H5 which is the relationship between the mediating variable, student perceived learning, and the dependent variable, student satisfaction. The relation between the two variables is a significant relationship and it proves that student perceived learning does influence student satisfaction. The variable student perceived learning is influenced by other independent variables. However, the variable student perceived learning itself also has its own set of items which are the promotion of achievement of student learning outcomes, pleased learning, useful skills for future, enhanced understanding, contribution to professional development, exploration of innovative teaching methods, the consciousness of learning skills, and well-learned presented class materials. All of these items influence the variable student perceived learning which is also influenced by four other independent variables, affecting the outcome of the dependent variable which is student satisfaction. According to Chan and Ko (2020), enabling them to make more informed decisions will increase their learning satisfaction.

Therefore, to conclude, all five hypothesis in our research are significant, valid, and has positive relationship between variables.

Achievement of Research Objectives

First, this research would benefit the teachers by allowing them to identify the factors influencing students’ satisfaction during online learning. Using a quantitative approach, this research looked at the variables that influence the students' perceived learning and students' satisfaction in an online learning environment. Teachers will benefit from this research as it will help them to visualize the relation of one factor to the other and how it will affect the students’ perceived learning as well as students’ satisfaction. In other words, they are provided with a greater sense of the student’s point of view. Therefore, this research allows the educators to revise their study plan to increase their students’ satisfaction and student’s perceived learning. This will aid them to improve the quality of learning in an online environment.

Second, the findings of this study reveal that instructor knowledge or facilitation and course structure are important factors affecting the student perceived learning and student perceived learning does influence student satisfaction. Therefore, after knowing these facts, educators who want to increase their students’ satisfaction will now know the first approach they have to take which is to increase the student perceived learning. This is because student satisfaction is affected by student perceived learning and to elevate the engagement on student perceived learning, one must focus on the two most important factors which are course structure and instructor facilitation or knowledge. A well-structured course will definitely help students to keep track of their learnings. It might also be easier for them to understand the course as a well-structured course will take a step-by-step approach. On the other hand, instructor facilitation or knowledge will definitely increase student perceived learning as they will receive proper guidance. Thus, student satisfaction can be elevated through this method.