Preference analysis on the online learning attributes among senior high school students during the COVID-19 pandemic: A conjoint analysis approach.

The COVID-19 pandemic has resulted in the shift from face-to-face to fully online learning. The purpose of this study was to evaluate the preference of senior high school students on online learning attributes during the COVID-19 pandemic by utilizing a conjoint analysis approach. Six attributes which consist of delivery type, assigned tasks, evaluation, virtual laboratory, interface layout, and delivery platform were simultaneously analyzed through orthogonal design. A total of 1189 senior high school students were collected via purposive sampling approach through the social media platform. The respondents voluntarily participated and answered 29 stimuli with 2 holdouts generated by using SPSS 25 utilizing a 7-point Likert scale. The results indicated that evaluation was found to be the most significant attribute and followed by virtual laboratory, delivery type, and delivery platform. Interestingly, multiple choice evaluation, not requiring virtual laboratories, mixed delivery type (synchronous with recorded lectures), and MS Teams as delivery platform were considered as the keys for the preference. This study is the first study that utilized a conjoint approach to analyze the senior high school students' preference on the online learning attributes during the COVID-19 pandemic. Finally, the conjoint approach can be applied and extended to evaluate the online learning attributes globally by utilizing the attributes and design created in this study.

Introduction

The COVID-19 pandemic has a big impact on the education sector across the world. It forced the educational system to shift from face-to-face to online which widely known as an online academic year or fully online learning (Morris et al., 2020). This fully online learning was implemented by the Government to minimize the infection rate of COVID-19 among students (Prasetyo et al., 2021a; Ocampo & Yamaguchi, 2020; Li & Lalani, 2020). The fully online learning implementation therefore increased the internet usage from an estimate of 56.3% of the global population to even higher of about 1500% (Nuncio et al. 2020). It even increased due to the implementation of fully online set-up due to the pandemic. Of the users, Asia was said to be the highest internet users of 50.4%, followed by Europe, and then Africa (Nuncio et al., 2020).

Even with the fully online set-up, e-Learning was being invested upon raising to about $18.66 billion USD in 2019 and is estimated to reach investment of $350 billion USD in 2025 (Li & Lalani, 2020). Nonetheless, Li & Lalani (2020) stated that the well-established e-Learning were limited to language applications, video conferencing tools, virtual tutorials, and online learning software. This shows that online learning came to full utilization unexpectedly.

Taking into consideration the Philippines, fully online learning has been implemented since March 2020. Thus, students have been experiencing this new type of educational setting for almost one year, including the senior high school students. Even internationally, fully online set-up was implemented to mitigate the pandemic (Li & Lalani, 2020). However, with the current set-up of fully online learning, students have difficulty coping when utilization of the online learning strategy is not met (Lin et al., 2017). With this coping difficulty, commitment to the student’s academic achievements affected the students’ satisfaction and motivation (Xiao & Wilkins, 2015).

Student’s motivation can be related to the social cognitive theory (SCT) and the self-determination theory (SDT). It is stated in the SCT that the social environment affects student’s motivation to learn (Schunk & DiBenedetto, 2020). SCT covers the student’s behavior, considering the environmental factor and personal factors (Bandura, 2001). In relation to SCT, SDT also tackles students’ behavior related to different factors (Butz & Stupnisky, 2017). These factors are autonomy, relatedness, and competency which could affect students’ wellness and in general, human development in an educational setting (Ryan & Deci, 2020). With that, there is a need to consider the students’ preference because it affects their motivation to continue pursuing online learning (Ryan and Deci, 2020, Kothe et al., 2018).

Although there were several studies that dealt with online learning (Santos et al., 2020; Pham et al., 2019; Li, 2019; Zamorano et al., 2019; Lin et al., 2017; Kuo et al., 2009), the students’ preference for online learning attributes during the COVID-19 is mainly underexplored. Prasetyo et al. (2020) only focused on the acceptance of students on the usage of Blackboard Collaborate using the extended Technology Acceptance Model. In addition, Pham et al. (2019) only focused on satisfaction of e-learning and loyalty among students. Furthermore, Li (2019) only focused on students’ acceptance and satisfaction on Massive Open Online Courses (MOOC). With that, preference such as the delivery type, virtual laboratory, delivery platform, assigned task, interface layout, and evaluation among students was mainly underexplored. Different methods can be utilized to analyze these preferences and one of the most widely utilized methods that can measure preferences is the conjoint analysis.

Conjoint analysis is a multivariate tool to measure people’s preference and understand a group’s attributes based on evaluation of the complete set-up (Gracía-Pérez and Gil-Lacruz, 2018, Mok et al., 2010, Moore, 2004). It has been widely utilized by several studies in different fields such as choosing public e-services (Pleger et al., 2020), preferences on the vaccine (Motta, 2021) among university students (Seanehia et al., 2017), and even adolescent choice in park preference (Hecke et al., 2018, Veitch et al., 2017). When it comes to education setting, different studies also utilized conjoint analyses. Macindo et al. (2019) focused on the experiential learning preferences of nurses. Similarly, Factor and de Guzman (2017) also did a study on Filipino nursing students, focusing on students’ instructor preferences. Furthermore, Mok et al. (2010) focused their study on the better intellectual property curriculum specifically in Korea. However, these conjoint studies were conducted before the fully online learning set-up due to COVID-19 pandemic. Thus, a further study that utilizes a conjoint analysis approach would be very valuable particularly for determining the best combination of online learning attributes during the fully online set-up.

The purpose of this study was to evaluate the preference of senior high school students on online learning attributes during the COVID-19 pandemic by utilizing a conjoint analysis approach. This conjoint analysis approach would help in determining preference weights of importance to different set of characteristics. This study is the first study that utilized a conjoint approach to analyze the senior high school students’ preference on the online learning attributes during the COVID-19 pandemic. Finally, the conjoint approach can be applied and extended to evaluate the online learning attributes globally by utilizing the attributes and design created in this study.

Methodology

Participant

Through purposive sampling, 1189 senior high school students participated in this study. Table 1 represents the demographics of this study. From the descriptive statistics, 60.10% were male and 39.90% were female, 43.10% are in Grade 11% and 56.90% are in Grade 12. The majority of the respondents were between the age of 16 and 17 comprising 43.15% and 49.54%, respectively. As suggested by Sethuraman et al. (2005) and due to the COVID-19 pandemic, the responses were collected via Google forms distributed through different social media platforms.

Table 1

Demographics (n = 1189).

Characteristics	Category	N	%
Gender	Male	715	60.10
	Female	474	39.90
Grade Level	Grade 11	512	43.10
	Grade 12	677	56.90
	15	23	1.930
Age	16	513	43.15
	17	589	49.54
	18	64	5.380

This study was approved by Mapua University Research Ethics Committee. All participants were also required to fill out the consent form before answering the questionnaire. In addition, the survey was made open from February 2021 to April 2021 so students had equal opportunities in answering the online survey.

Conjoint design

Fig. 1 presents the research conceptualization of this study. The preparation stage involved the brainstorming and gathering of related studies to come up with the attributes that were considered in this study. We opted to use general attributes that are utilized by senior high school in E-learning. After considering the attributes and levels, the generation of the orthogonal design was created using SPSS following the suggestion of Kuzmanovic et al. (2011) and Hair (2010). Questionnaire development utilizing Google forms was the last preparation before collecting the data. The implementation stage involved a test run of the first 50 participants were collected to do a preliminary run (Hair, 2010). As suggested by Hair (2010), the correlation analysis with a value greater than or equal to 0.700 would determine if the attributes considered are adequate for the study. The preliminary result had a Person’s R-value of 0.947 which was greater than the cut-off value, therefore, the attributes considered were deemed adequate. The final stage involved full questionnaire distribution, generation of results, and analysis and interpretation.

Fig. 1

Conceptualization.

Table 2 represents the 6 online learning attributes that being analyzed in this study. The attributes were selected based on the possible online learning set-up during COVID-19 pandemic. Those 6 attributes were Delivery type, Assigned Task, Evaluation, Virtual Laboratory, Interface Layout, and Delivery Platform and were analyzed simultaneously through conjoint analysis with orthogonal design.

Table 2

Attributes for Online Learning Set-up.

Attributes	Characteristics
Delivery Type	Asynchronous
	Synchronous
	Mixed
Interface Layout	Weekly
	Course Outcome
Evaluation	Multiple Choice
	Computation
	Essay
	Identification
	Matching Type
	Project
Assigned Task	By group
	Individual
Virtual Laboratory	Virtual Laboratory
	None Required
Delivery Platform	Zoom
	MS Teams
	Blackboard Collaborate

The first attribute is the delivery type. The delivery types such as asynchronous, synchronous, and mixed were based on the protocol of the department of education (Bellafante, 2020, Bernardo, 2020). Asynchronous delivery type is where pre-recorded lectures and modules are being distributed to an interface being utilized in their respective schools. The students will be able to access the modules and lectures online at their convenient time (Lee ). Synchronous is where live teaching happens, following a set day and time throughout the week (Jan, 2020; Yang et al., 2019; Besser et al., 2020; Wolverton, 2018) while mixed delivery type is when synchronous classes are being recorded and uploaded after class to help students recall lessons (Lapitan et al., 2021, Aghababaeian et al., 2019, Yang et al., 2019). Moreover, the mixed delivery type was created for those who had difficulty with the internet connection to keep up with the lessons in class.

Second, the interface layout for learning materials was also considered. Fig. 2, Fig. 3 represents two possible layouts that is widely using the Blackboard Collaborate. Fig. 2 is the weekly layout wherein the lecture materials used for that week are uploaded while Fig. 3 presents the course outcome layout wherein any lecture materials covering that specific topic is uploaded. According to Christensen (2020), the interface layout is important to consider for an individual’s acceptance to a system design. This acceptance of the Interface Layout enhances the participation of the user in the system (Christensen, 2021).

Fig. 2

Weekly Layout.

Fig. 3

Course Outcome Layout.

Third, the attribute measured in this study was the assigned tasks being given to students. Students would either prefer to do tasks individually (Schultze et al., 2012, Kirschner et al., 2009) or within a group (Dobao, 2012, Kirschner et al., 2009). Students are being assigned different tasks such as homework, projects, and other outputs. Usually, the teachers assign these tasks individually or by the group. As indicated by Kirschner et al. (2009), in giving individual tasks, students tend to try to learn everything to be able to accomplish the assigned tasks. For group works, students tend to focus only on their assigned parts only (Kirschner et al., 2009).

Fourth, evaluation was also measured as attributes in this study. There were 6 different types of evaluation considered: multiple choice, identification, essay, computation-based, project-based, and matching type. First, multiple choice-based evaluation ( Fig. 4) is an assessment wherein a question consists of potential answers and only one is correct (Butler, 2018). Second, identification-based evaluation is when students are asked to input the correct answer. Third, essay-based evaluation is when students compose explanations of a topic given to measure the level of understanding and knowledge. Fourth, computation-based assessment is where students show their solution to derive to an answer. Fifth, project-based evaluation is where students create a prototype that is related to the course to measure the level of knowledge that they have gained throughout the course. Lastly, matching-type based evaluation ( Fig. 5) is a form of assessment that pairs two columns. Column A is the question and students are asked to match it to column B for the answers (Treser, 2015).

Fig. 4

Multiple Choice Assessment (Butler, 2018).

Fig. 5

Matching Type Assessment (Treser, 2015).

Fifth, with online learning, virtual laboratory was also considered. According to the study of Estriegana et al. (2019), laboratory activities should be engaging and fun. With the lack of hands-on laboratory activities, it is best to consider the virtual laboratories utilized in the fully online learning set-up.

Lastly, the delivery platform was also measured. In fully online learning, either Zoom (Mahr et al., 2021), MS Teams (Pal & Vanijja, 2020), or Blackboard Collaborate (Prasetyo et al., 2020) are common delivery platforms during online learning. Thus, these delivery platforms were also considered in this study.

Statistical Analysis

SPSS 25 was utilized to generate the conjoint analysis with orthogonal design. A total of 27 stimuli were generated by the SPSS with 2 holdouts, creating 29 total stimuli for this study (Ong et al., 2021a). The orthogonal design was selected to ensure the reasonable number of stimuli that were evaluated by the participants. Table 3 represents the 29 stimuli in this study. These 29 stimuli were evaluated by 7-points Likert scale ranging from 1 as “Strongly disagree” and 7 as “Strongly agree”.

Table 3

Stimulus.

Combination	Delivery Type	Interface Layout	Evaluation	Assigned Task	Virtual Laboratory	Delivery Platform
1	Asynchronous	Course Outcome	Project	Individual	Virtual Laboratory	Blackboard Collaborate
2	Synchronous	Weekly	Essay	Group	Not Required	Zoom
3	Asynchronous	Weekly	Identification	Group	Virtual Laboratory	Zoom
4	Synchronous	Weekly	Computation	Group	Virtual Laboratory	Blackboard Collaborate
5	Asynchronous	Course Outcome	Computation	Group	Virtual Laboratory	MS Teams
6	Synchronous	Weekly	Multiple Choice	Individual	Virtual Laboratory	MS Teams
7	Mixed	Course Outcome	Computation	Group	Virtual Laboratory	Zoom
8	Mixed	Course Outcome	Multiple Choice	Group	Not Required	Blackboard Collaborate
9	Mixed	Weekly	Identification	Individual	Virtual Laboratory	Blackboard Collaborate
10	Asynchronous	Weekly	Computation	Individual	Not Required	MS Teams
11	Synchronous	Weekly	Computation	Group	Not Required	Blackboard Collaborate
12	Synchronous	Weekly	Essay	Individual	Virtual Laboratory	Zoom
13	Asynchronous	Weekly	Project	Individual	Not Required	Blackboard Collaborate
14	Synchronous	Weekly	Multiple Choice	Group	Virtual Laboratory	MS Teams
15	Mixed	Weekly	Essay	Group	Not Required	MS Teams
16	Synchronous	Course Outcome	Matching Type	Individual	Virtual Laboratory	Blackboard Collaborate
17	Asynchronous	Weekly	Essay	Group	Virtual Laboratory	Blackboard Collaborate
18	Synchronous	Course Outcome	Project	Group	Virtual Laboratory	Zoom
19	Mixed	Course Outcome	Essay	Individual	Virtual Laboratory	MS Teams
20	Asynchronous	Course Outcome	Essay	Group	Virtual Laboratory	Blackboard Collaborate
21	Synchronous	Weekly	Matching Type	Group	Virtual Laboratory	MS Teams
22	Mixed	Weekly	Multiple Choice	Group	Virtual Laboratory	Blackboard Collaborate
23	Asynchronous	Course Outcome	Multiple Choice	Individual	Not Required	Zoom
24	Asynchronous	Weekly	Multiple Choice	Group	Virtual Laboratory	Zoom
25	Asynchronous	Weekly	Matching Type	Group	Virtual Laboratory	MS Teams
26	Mixed	Weekly	Computation	Individual	Virtual Laboratory	Zoom
27	Mixed	Weekly	Matching Type	Group	Not Required	Zoom
28	Synchronous	Course Outcome	Identification	Group	Not Required	MS Teams
29	Mixed	Weekly	Project	Group	Virtual Laboratory	MS Teams

Results

Table 4 and Table 5 represent the utilities and the average importance score of the online learning attributes during the COVID-19 pandemic respectively. Based on these tables, Evaluation was found to have the highest score (41.20%), followed by Virtual Laboratory (22.55%), Delivery Type (17.88%), Delivery Platform (13.27%), Interface Layout (4.388%) and Assigned Task (0.708%). For the evaluation, Multiple-choice evaluation was considered the most preferred. Moreover, students do not prefer virtual laboratories. For the delivery type, the students preferred a mixed delivery type wherein classes are conducted on a specific day and time and the lectures are being recorded. This helps students recall the lessons after recordings are uploaded. For the delivery platform, students prefer to utilize MS Teams rather than Zoom and Blackboard collaborate. Lastly, students preferred to work within a group and considered course outcome layout as the preferred interface.

Table 4

Utilities.

Attributes	Preference	Utility Estimates	Std. Error
	Asynchronous	-0.061	0.019
Delivery Type	Synchronous	-0.143	0.019
	Mixed	0.204	0.019
Interface Layout	Weekly	-0.043	0.014
	Course Outcome	0.043	0.014
	Multiple Choice	0.449	0.028
	Computation	-0.350	0.028
Evaluation	Essay	-0.124	0.028
	Identification	-0.215	0.036
	Matching Type	0.179	0.036
	Project	0.061	0.036
Assigned Tasks	Group	0.007	0.014
	Individual	-0.007	0.014
Virtual Laboratory	Virtual Laboratory	-0.219	0.014
	Not required	0.219	0.014
	Zoom	0.063	0.019
Delivery Platform	MS Teams	0.097	0.019
	Blackboard Collaborate	-0.160	0.019
(Constant)	4.180	0.017

Table 5

Averaged Importance Score.

Importance Values	Score
Delivery Type	17.881
Interface Layout	4.3880
Evaluation	41.200
Assigned Task	0.7080
Virtual Laboratory	22.554
Delivery Platform	13.268

Table 6 presents the ranking of the different combinations considered in this study. The ranking is based on the most preferred (ranking 1) and the least preferred (ranking 29). It is seen in the table that combination 8 is the most preferred followed by combination 4 as the least preferred by students during the online learning set-up.

Table 6

Stimulus Rank.

Combination	Delivery Type	Interface Layout	Evaluation	Assigned Task	Virtual Laboratory	Delivery Platform	Total	Rank
1	Asynchronous	Course Outcome	Project	Individual	Virtual Laboratory	Blackboard Collaborate	-0.343	20
2	Synchronous	Weekly	Essay	Group	Not Required	Zoom	-0.021	13
3	Asynchronous	Weekly	Identification	Group	Virtual Laboratory	Zoom	-0.468	23
4	Synchronous	Weekly	Computation	Group	Virtual Laboratory	Blackboard Collaborate	-0.908	29
5	Asynchronous	Course Outcome	Computation	Group	Virtual Laboratory	MS Teams	-0.483	26
6	Synchronous	Weekly	Multiple Choice	Individual	Virtual Laboratory	MS Teams	0.134	8
7	Mixed	Course Outcome	Computation	Group	Virtual Laboratory	Zoom	-0.252	18
8	Mixed	Course Outcome	Multiple Choice	Group	Not Required	Blackboard Collaborate	0.762	1
9	Mixed	Weekly	Identification	Individual	Virtual Laboratory	Blackboard Collaborate	-0.440	22
10	Asynchronous	Weekly	Computation	Individual	Not Required	MS Teams	-0.145	16
11	Synchronous	Weekly	Computation	Group	Not Required	Blackboard Collaborate	-0.470	24
12	Synchronous	Weekly	Essay	Individual	Virtual Laboratory	Zoom	-0.473	25
13	Asynchronous	Weekly	Project	Individual	Not Required	Blackboard Collaborate	0.009	10
14	Synchronous	Weekly	Multiple Choice	Group	Virtual Laboratory	MS Teams	0.148	7
15	Mixed	Weekly	Essay	Group	Not Required	MS Teams	0.360	4
16	Synchronous	Course Outcome	Matching Type	Individual	Virtual Laboratory	Blackboard Collaborate	-0.307	19
17	Asynchronous	Weekly	Essay	Group	Virtual Laboratory	Blackboard Collaborate	-0.600	28
18	Synchronous	Course Outcome	Project	Group	Virtual Laboratory	Zoom	-0.188	17
19	Mixed	Course Outcome	Essay	Individual	Virtual Laboratory	MS Teams	-0.006	12
20	Asynchronous	Course Outcome	Essay	Group	Virtual Laboratory	Blackboard Collaborate	-0.514	27
21	Synchronous	Weekly	Matching Type	Group	Virtual Laboratory	MS Teams	-0.122	15
22	Mixed	Weekly	Multiple Choice	Group	Virtual Laboratory	Blackboard Collaborate	0.238	5
23	Asynchronous	Course Outcome	Multiple Choice	Individual	Not Required	Zoom	0.706	2
24	Asynchronous	Weekly	Multiple Choice	Group	Virtual Laboratory	Zoom	0.196	6
25	Asynchronous	Weekly	Matching Type	Group	Virtual Laboratory	MS Teams	-0.040	14
26	Mixed	Weekly	Computation	Individual	Virtual Laboratory	Zoom	-0.352	21
27	Mixed	Weekly	Matching Type	Group	Not Required	Zoom	0.629	3
28	Synchronous	Course Outcome	Identification	Group	Not Required	MS Teams	0.008	11
29	Mixed	Weekly	Project	Group	Virtual Laboratory	MS Teams	0.107	9

Table 7 represents the reliability results of the study. Based on this table, The Pearson’s R correlation showed a value of 0.992 (a value close to 1) and 0.000 significance test indicating that the result is acceptable (Hair, 2010). Hair (2010) discussed that Pearson’s R is utilized to determine the profile of the attributes considered. The design created could be validated from all responses with a value close to 1.00. Moreover, Table 7 shows Kendall’s Tau value 0.892 and Kendall’s Tau for Holdouts with a value of 1.000. With 2 holdouts considered, a perfect value of 1.000 showed that the participants were consistent with how the questionnaire was answered (Ong et al., 2021a, Ong et al., 2021b). This shows that the internal combination has an accepted internal validity (Prasetyo et al., 2014). Hair (2010) also discussed that Kendall’s Tau is utilized to determine internal consistency among the response of the respondents. With a value close to 1.00 (≥ 0.70), this result also showed the overall consistency within the responses of the senior high school students (Ong et al., 2021c).

Table 7

Correlation.

	Value	Significance
Pearson’s R	0.992	0.000
Kendall’s Tau	0.892	0.000
Kendall’s Tau for Holdouts	1.000

Discussion

The combination that was most preferred was having a multiple-choice based evaluation (0.449), without virtual laboratory (0.219), mixed delivery type (0.204), utilization of MS Teams as the delivery platform (0.097), course outcome as the interface layout (0.043), and by group assigned task (0.007) with a total possible value of 1.019. The least preferred combination involved having a computation-based evaluation (−0.350), with virtual laboratory (−0.219), synchronous delivery type (−0.143), utilization of Blackboard Collaborate as the delivery platform (−0.160), weekly as the interface layout (−0.043), and individually assigned task (−0.007) with a total possible value of − 0.922.

From the results seen in Table 6, the averaged importance score showed that Evaluation was the attribute students focused on when it comes to fully online learning set-up. This is supported by the study of Albay and Eisma (2021) wherein they stated that evaluations are highlighted to be very important in developing the competence and skills of students in the 21st century. From which, having a multiple-choice evaluation was preferred by the students in this study. Multiple choice-based evaluation is an assessment wherein a question consists of potential answers wherein only one is correct (Butler, 2018). This was the most preferred way of evaluation because multiple-choice type of evaluation provides students with clues leading them to the correct answer. This happens when a question (stem) is provided for example 5-item choice (Primary choices or secondary choices, as seen in Fig. 6).

Fig. 6

Multiple choice (Butler, 2018).

Butler (2018) stated that multiple-choice type would engage students in strategic guessing. This leaves students narrowing down plausible correct answers as compared to other types of evaluation such as identification or essay – wherein students need to put in the correct answer without any form of a guide. Multiple-choice type of exam would be deemed easy for students to have a higher grade in the evaluation as compared to other types of evaluation. The results of this study are also supported as the second-highest attribute among the evaluation is matching type. Matching type evaluation also has choices, leading the students to be able to eliminate answers and choose or guess what may be deemed the correct answer. However, Zhu et al. (2020) stated that automated feedback or quick response to evaluations such as formative writing would engage students with the said evaluation.

Virtual Laboratory was the second-highest important attribute. Based from the result, most students do not prefer virtual laboratory. According to Sneddon and Douglas (2013), laboratory activities are interesting to students when they have a sense of fun, a sense of application of the different concepts, and theories discussed in their lecture classes. With the online learning set-up, students do not feel the actual scenario of doing the experiment which supports the negative implications of students when it comes to virtual laboratories (Ong et al., 2022). Moreover, Estriegana et al. (2019) explained that student’s enjoyment in doing laboratory activities is the crucial aspect of preference. The study added that the model of virtual laboratories needs to be modified to keep the engagement of students to enjoy and appreciate virtual laboratories.

The third attribute that the students considered was the preference to have Mixed Delivery Type during the online learning set-up. Mixed delivery type is when synchronous classes are being held, at the same time the lectures are being recorded and uploaded after class in their respective delivery platforms. This helps students to review the whole lecture after class or before evaluations. Shah et al. (2013) indicated that students would tend to prefer the pre-recorded lectures and learning materials to study in their own pace. On the other hand, Aghababaeian et al. (2019) stated that either asynchronous or mixed delivery type would result to no significance when it comes to students’ evaluation outcomes. However, in terms of preference, our study highlighted that mixed delivery type was still more preferred than asynchronous. This can be supported by a report of Domingo (2020). Since the respondents are from the Philippines, internet connectivity is a problem. Domingo (2020) stated that the Philippines ranked 32nd among other Asian countries when it comes to internet speed. With the whole country simultaneously using the internet, it bounds to have usual problems. This also is one reason to support why students preferred mixed delivery type – to be able to view the recorded lectures in times of connectivity problems. This finding is also applicable to countries with low internet speed.

The Delivery Platform was the fourth important attribute that the students considered. The students highly preferred the utilization of MS Teams, followed by Zoom, and the students did not prefer the Blackboard Collaborate. MS Teams is an application where students can create groups, chat, upload materials, collaborate by pairs or by a group with video conferencing and screen-sharing, and recordings. Pal and Vanijja (2020) stated that students preferred this delivery platform due to ease of use and got a high usability in the System Usability Scale. This leads to the high preference among students; however, they cannot do even simple chatting in Zoom when the session is not active; also, blackboard collaborate (Prasetyo et al., 2020). Mahr et al. (2021) stated that Zoom has functions such as live chatting, collaboration, and file sharing, however, security such as personal information may be easily obtained. However, students only utilize Zoom for live collaboration and live lectures using an account provided by the university. This is the reason why among the attributes, Zoom was preferred to be the second delivery platform following MS Teams.

Interestingly, results showed that Blackboard Collaborate was seen to be the least preferred by students even with high-security measures. With the result of Prasetyo et al. (2020) showing Blackboard Collaborate having a high satisfaction among students, the highlighted variables in their study were actual use, features, perceived usefulness, and perceived interactivity. This showed that Blackboard collaborate does not have a feature for students to easily activate collaboration without the instructor. It also does not have chatting features and easy file sharing like MS Teams and Zoom. Even with messaging across different classmates are available, it does not have the notification features on computers and is not as easy as simple chatting.

The two attributes, the Interface Layout and Assigned Tasks had low consideration on the preference. On the interface layout, students prefer a compilation per course outcome rather than weekly. This is because all materials for that specific course outcome is under one folder. This helps students follow what the coverage of the course outcome is. Mohammadi (2015) and Cidral et al. (2018) stated that functionality, efficiency, responsiveness, and flexibility lead to satisfaction of using a system. The course outcome layout would have ease of usage wherein all materials such as PowerPoint files, documents, recordings, and other lecture materials are uploaded in one area. This enables the students to access the learning materials with ease (Hornbæk & Hertzum, 2017).

As the least important attribute, results showed that students preferred to work within a group rather than individually. According to the study of Ruetzler et al. (2011), students who are highly inclined in technology or information work well with teams. Since the respondents are in a generation of highly developed technology, results showed that students prefer to work in groups than individually. Moreover, Fjelstul (2007) showed that working in groups is the second most desired attribute. However, the study of Kirschner et al. (2009) showed that students would learn more from doing tasks individually rather than doing it in groups. This is because the knowledge when doing the assigned task individually forces students to review and deepen their understanding of a topic. Within a group, the students would tend to disseminate the assigned task and eventually focus on their respective assigned task (Kirschner et al., 2009).

Overall, the combination set from the preference of the senior high school students has both advantages and disadvantages. Aghababaeian et al. (2019) mentioned that either mixed or asynchronous classes would not hinder the learning processes of the students since the score would not have a significant difference. However, the study showed that students should be responsible for managing their time in learning the lessons (Aghababaeian et al., 2019). Multiple-choice as evaluation could be implemented but with responses that need cognitive processes is highly suggested. It could also be suggested to have a multiple-choice based evaluation with multiple responses or the requirement for answer justification (Butler, 2018). In addition, virtual laboratory activities should be engaging to promote the interest of students. Delivery platforms such as MS Teams where students could easily communicate, learn, share files, and collaborate is highly preferred. For the interface layout, it was found that categorizing lecture materials in one area was more preferred by students for easy access. Moreover, results showed that assigned tasks were not part of the preference of the students. Since fully online learning promotes learning efficacy among students, an individual task should be given in priority rather than working within a group to enhance students’ understanding (Kirschner et al., 2009).

Theoretical contributions

From a theoretical standpoint, the result of this study could be a powerful tool in understanding the attributes of students when it comes to fully online learning. Since this is the first study that utilized conjoint analysis in the online learning attributes, it could be the basis to enhance students’ satisfaction when preference is addressed. With the collaboration of learning platforms and media usage, the findings of this study could help in how it could be delivered effectively.

This study could be used as the basis for further exploration when it comes to student satisfaction and motivation to academics when correlated to performance. This approach could be beneficial to universities considering the preference of students to rethink the possibility of simply doing the traditional learning online. It is noteworthy that traditional learning is different from fully online learning delivery, therefore, this analysis provided the perception of students who experienced a full online academic year. As indicated by Ryan and Deci (2020) and Kothe et al. (2018), there is a need to consider the preference of students because it affects their motivation to continue pursuing online learning and would eventually finish education.

Practical applications

The results of this study were able to measure the preference of senior high school students among different attributes applicable in a fully online learning set-up. In a university, the students are considered as the customer. Therefore, the preference of students may be used as a promotion to engage students to enroll in a university. This could be utilized as a marketing strategy by different universities to have students enrolled in fully online learning mode. The emphasis on the findings of this study could also help build reputation and promotion for long-term student satisfaction, hence, increase profit by increase the enrollment population.

Most substantially, the focus on delivery type, improvements in the delivery platform utilization, and the virtual laboratory could pave the increase of engagements of students. Moreover, the evaluation and assigned task may be highlighted during classes to have word-of-mouth promotion. This could promote students’ motivation to use and could enhance learning during fully online learning. Lastly, the outcome of this study does not limit universities during the COVID-19 pandemic but could consider the preference of students to continue offering traditional learning and fully online learning, even blended learning.

Limitations

This study considers different limitations. First, this study measured only the preference among senior high school students and not the performance of the students itself. Future research can correlate preference and performance by utilizing a structural equation modeling approach. Second, this study measured only senior high school students and it is proposed to measure the preference of the attributed for college and graduate school students. The attribute may result to different preference since the level of understanding and priorities are different among different age groups (Spencer et al., 2020). To elaborate on the result, a suggestion is to correlate also the preference and performance among the different educational levels using structural equation modeling (Chuenyindee et al., 2022, Prasetyo et al., 2021b). Moreover, the differentiation of age and gender, together with the experience of utilizing online classes or E-learning before the pandemic may play a big part in the results of the study. It would be suggested that future researches consider these variables as an extension of this study. Lastly, the study was conducted during the COVID-19 pandemic. Different attribute measures may be seen after the COVID-19 pandemic when the educational setting goes back to the traditional way of learning. Students may have a different perspective when both fully online learning and traditional learning is available.

Conclusion

With the COVID-19 pandemic resulting to the shift from face-to-face to fully online learning, it is encouraged to consider the students’ preference with the fully online learning to enhance motivation leading to high satisfaction and learning. The purpose of this study was to evaluate the preference of senior high school students on online learning attributes during the COVID-19 pandemic by utilizing a conjoint analysis. It was found out that from six attributes which consist of Delivery type, Assigned Tasks, Evaluation, Virtual Laboratory, Interface Layout, and Delivery Platform were simultaneously analyzed through orthogonal design, evaluation was the most considered attribute among senior high school students.

The attributes that highlighted the preference of students were to have mixed delivery type, multiple choice evaluation, without virtual laboratory requirements, and MS Teams as the learning platform. Even with low attribute evaluation of Interface Layout and Assigned Task, it is advisable to prioritize individual works as it enhances the students’ learning compared to group works. The 29 stimuli showed consistency with the Kendall’s Tau for Holdouts value of 1.000. The approach can be applied and extended to evaluate the online learning attributes in other countries. This study can also be a strong foundation for the marketing strategy of the online learning implementation in different universities. Finally, when preference of students is considered, this heightens their motivation and therefore increases students’ motivation. This will lead to an increase in enrollment among universities.

Author statement

All authors have made substantial contributions to all of the following: (1) the conception and design of the study, or acquisition of data, or analysis and interpretation of data, (2) drafting the article or revising it critically for important intellectual content, (3) final approval of the version to be submitted.