Introducing virtual classrooms for undergraduate physiology teaching during the COVID-19 pandemic: acceptance by students and subjective impact on learning

Abstract

The COVID-19 pandemic and worldwide lockdowns brought major changes in education systems. There was a sudden obligatory shift toward utilization of digital resources for teaching and learning purposes. Medical education, specifically physiology teaching, comprises hands-on training in the laboratory. It is challenging to offer a course like physiology in a virtual format. The objective of this study was to assess the effectiveness and influence of virtual classroom technology on online physiology education in a sample size of 83 first-year MBBS undergraduates. A questionnaire comprising questions related to technology accessibility and utilization, comprehensibility and effectiveness of instructions, faculty proficiency, and learning outcomes was administered to the group. The responses were collected and analyzed. Validation through principal components and factor analysis showed that online teaching is not very effective and has a limited application in the physiology education of undergraduate MBBS students. Our study also revealed that virtual physiology teaching of undergraduate medical students during the COVID-19 pandemic had a moderate level of effectiveness.

NEW & NOTEWORTHY In the present qualitative study, we have conducted and validated an online physiology teaching platform at a medical college to continue medical education during the peak times of the COVID-19 pandemic and prolonged lockdowns. Furthermore, we have evaluated the effectiveness of online physiology teaching through multidimensional feedback from undergraduate MBBS students. It is experimental evidence of inadequate sustainability, moderate efficacy, limited application, and poor first-hand experience gained by the students in virtual physiology teaching in a preclinical and clinical setting.

INTRODUCTION

The suspension of classroom teaching in all educational institutions was enforced as a measure of social distancing to control the spread of the COVID-19 pandemic. India was among the first nations to enforce a complete and all-encompassing lockdown. Institutions were forced to teach a traditional in-person physiology course in a virtual format. We found ourselves in a situation that required us to promptly transition our meticulously designed curriculum to an unfamiliar mode of delivery. Because of the uncertainty and new challenges arising from the limit reached by internet infrastructure, it was necessary to adapt to the “new normal” (1). The use of digital resources for teaching and learning became obligatory in most fields of education (2, 3). Physiology courses require significant laboratory work and practical training, which is challenging to replicate in a virtual setting. When the nationwide lockdown was imposed because of the COVID-19 pandemic, the traditional modes of teaching were disrupted, and online modes of teaching emerged as alternative options to ensure the continuity of education.

The study was designed to validate the effectiveness of online physiology teaching through the virtual classroom format. The purpose of this study was threefold: first, to determine whether online physiology teaching could be a substitute for traditional classroom instruction; second, to ascertain students’ views on the virtual teaching; and third, to examine the impact of online physiology instruction on the overall academic performance of first-year MBBS undergraduates.

METHODS

A cross-sectional, questionnaire-based observational study was designed. After approval from the Institutional Ethics Committee, All India Institute of Medical Sciences, New Delhi (IEC-578/06.08.2021), the study was initiated after obtaining informed consent from the participants.

An online questionnaire was shared as a Google form along with participants’ information among the first-year undergraduates of a medical college. The questionnaire was administered between June 2020 and August 2020, after their second assessment but before the final professional examination. Participation was completely voluntary and not linked to their curriculum and assessments. The responses did not include any information that revealed the identity of the students. Eighty-three medical undergraduates (∼78% of the total strength of the class) gave informed consent for the study. The study was aimed at the validation of the effectiveness of online physiology teaching for undergraduate medical students in different aspects of medical education such as overall interaction, attitude, learning outcomes, and use of technology with structured feedback-based questions. The duplicity of the feedback was taken care of by having their college roll number (mandatory) and phone number (optional) as primary and secondary identification of the participants, respectively.

The questionnaire included 43 questions (Table 1); of these, 27 questions were Likert type. Using questionnaires, we assessed the students on various aspects of learning outcomes such as “students’ use of technology” (4 items), “online classroom experience” (11 items),” interaction with faculty” (6 items), “students’ interactions among themselves” (2 items), and “learning outcomes” (4 items). In the Likert items, 23 were positive and 4 were negative worded (Table 2). Each question was unambiguous, self-descriptive, and complete in itself. The Likert 5-point scale questions were agreement type (15 questions), frequency type (7 questions), performance type (4 questions), and intensity type (1 question).

Table 1.

List of questions

Question
1. Which gadget do you use to access online classes?
2. Do you have internet connectivity for access to online classes?
3. Do you have an exclusive gadget available to access online classes?
4. The technologies used in this course worked the way it was intended.
5. I had some problems logging into the class with my assigned link.
6. I find the online classes conducive to the way I like to learn.
7. The online classes are helpful in learning the topic.
8. I would like the duration of online classes to be longer.
9. How much of your time during online class was spent actively listening to the faculty?
10. How much of your time during online class was spent in other activities like email, Facebook, Instagram, etc.?
11. Would you prefer availability of a discussion forum on WhatsApp or Facebook with faculty and other students?
12. Which medium would you like to be added for enhanced learning (multiple items can be selected)?
13. Motivation to learn online is as high as in face-to-face classroom.
14. The incentive to participate and interact in class is
15. It is easier to be absent in online classrooms.
16. It is easy to get distracted in online classrooms.
17. I was able to understand the concepts in online teaching mode.
18. The online mode supported my learning.
19. Online mode should be the only mode of learning.
20. I prefer face-to-face traditional classrooms to online learning.
21. There is not much difference in online and face-to-face classrooms.
22. How much of your interaction occurred with the faculty as compared to face-to-face classroom?
23. Facial expressions and gestures of teacher in face-to-face classroom enhance learning experience.
24. The faculty in the online classroom was able to identify me during interactions.
25. The faculty was active and engaged with the students.
26. There was adequate opportunity to interact online with the faculty.
27. The faculty replied to the queries/comments raised in class.
28. There was adequate opportunity to interact online with other students.
29. Feasible online study groups
30. Scope of self-learning
31. Discussions and interactions with other students
32. Summarizing acquired knowledge to help learning of other students
33. Consolidating and managing my own learning
34. Online learning is: Monotonous/Varied
35. Online learning is: Passive/Active
36. Online learning is: Dull/Exciting
37. Online learning is: Boring/Interesting
38. Online learning is: Bad/Good
39. Online learning is: Taxing/Nontaxing
40. Online learning is: Knowledge Based/Skill Based
41. Online classes are: Ineffective/Effective
42. Online classes are more of: Facts/Ideas
43. Online classes are: Unstructured/Structured

Table 2.

Average score on 5-point Likert scale of the factors and variables of the study

Question	Factors	Variable Characteristics	Mean	SD	Communalities
1	The technologies used in this course worked the way it was intended.	Technology	3.39	0.778	0.515
2	I had some problems logging into the class with my assigned link.	Logging	3.05	1.070	0.247
3	I find the online classes conducive to the way I like to learn.	Conducive to learn	2.93	1.156	0.733
4	The online classes are helpful in learning the topic.	Helpful in learning	3.16	1.099	0.738
5	Duration reverse scored	Duration	1.93	1.080	0.344
6	How much of your time during online class was spent actively listening to the faculty?	Active listening	3.39	0.824	0.554
7	Motivation to learn online is as high as in face to face classroom.	Motivation	2.41	1.288	0.612
8	The incentive to participate and interact in class is	Incentive	2.73	1.013	0.518
9	It is easier to be absent in online classrooms.	Absence	2.36	1.077	0.590
10	It is easy to get distracted in online classrooms.	Distraction	1.86	1.061	0.632
11	I was able to understand the concepts in online teaching mode.	Understanding	3.31	0.999	0.742
12	The online mode supported my learning.	Support in learning	3.16	1.053	0.801
13	Online mode should be the only mode of learning.	Sole mode	1.83	1.248	0.635
14	I prefer face-to-face traditional classrooms to online learning.	Conventional classroom	2.08	1.118	0.708
15	There is not much difference in online and face-to-face classrooms.	Difference	1.93	0.960	0.190
16	How much of your interaction occurred with the faculty as compared to face-to-face classroom?	Interaction with faculty	2.23	0.941	0.603
17	Facial expressions and gestures of teacher in face-to-face classroom enhance learning experience.	Facial expressions	2.12	1.130	0.464
18	The faculty in the online classroom was able to identify me during interactions.	Identify me	2.51	0.992	0.590
19	The faculty was active and engaged with the students.	Faculty engagement	3.18	0.857	0.612
20	There was adequate opportunity to interact online with the faculty.	Interaction with faculty	2.81	1.006	0.748
21	The faculty replied to the queries/comments raised in class	Queries	3.40	0.910	0.504
22	There was adequate opportunity to interact online with other students.	Interaction with students	2.70	1.079	0.569
23	Online study group	Study groups	3.25	1.069	0.573
24	Self learning	Self learning	2.78	1.083	0.345
25	Discussions and interactions with other students	Group discussion	2.46	1.151	0.532
26	Summarizing acquired knowledge to help learning of other students	Help others learning	2.18	1.112	0.736
27	Consolidating and managing my own learning	Managing own learning	2.89	0.959	0.788

Questions 9, 10, 14, and 17 are negatively worded on Likert scale. Questions 1–4: student use of technology; questions 5–15: online classroom experience; questions 16–21: interaction with faculty; questions 22 and 23: student interactions among themselves; questions 24–27: learning outcomes.

To assess the magnitude and direction of attitude toward online learning, 10 questions (questions 34–43) based on the Osgood differential semantic scale were also used, as given in Table 1 (4). The direction of the attitude was measured by agreement or disagreement with the statement, and the strength of the attitude was assessed by the degree of agreement or disagreement of the participants. Each response to bipolar adjectives of opposite nature was scored from 1 to 7 on a different rating scale, making the maximum possible score 56. An attitude score was obtained by summing the responses of the questions. The Osgood differential semantic scale was used to correlate the responses to their midterm scores already undertaken.

Kaiser–Meyer–Olkin (KMO) test was used for sampling adequacy and to measure the suitability of the data for factor analysis in terms of the proportion of variance in the variables that probably are due to underlying factors. KMO value (normal value 0 to 1) was assessed; KMO value for our study variables was between 0.8 and 1.0 (KMO = 0.82), which indicates adequate sampling and meritorious enough for further analysis (N = 83). Communalities or h² is the sum of squared factor loadings for the variables and measures internal variability in the factors. Bartlett’s test of sphericity was used to test the null hypothesis. Correlation matrix indicated whether the variables of the student feedback in our study were unrelated to each other and whether there is an adequate redundancy in the variables and therefore unsuitable for structure detection of the study. A value < 0.05 indicated that factor analysis could be worthwhile for our datasets. Our sample size was <300, so the average communality of the retained items has to be tested for internal consistency in the variables. An average value > 0.6 is acceptable for a sample size < 100. Communalities of all the factors are given in Table 1. Communalities of ∼70% of the variables of our study were in the acceptable range, i.e., ≥0.5, which suggests adequate internal consistency among the variables used in the study. Checking internal consistency was required to proceed with factor loadings and factor analysis and performance of principal component analysis.

With factor analysis we clustered similar variables into the same factor to identify underlying variables using a data correlation matrix. Principal component analysis of the Likert-type questions was performed with factor extraction methods. The reliability of the questions was examined with Cronbach’s α, which provided the simplest way to measure whether or not a scoring scale is reliable. Cronbach’s α has a range between 0 and 1, according to which a moderate score of Cronbach’s α was found in our study (Cronbach’s α = 0.46). The rotation sums of squared loadings for the total variance explained of the components were the correlation between each variable.

In the present study, an extraction method-based principal component analysis was performed. Using factor extraction, we encompass determining the least number of factors that can be used as core variables of the interrelationships among the set of variables in the feedback questions of learning outcomes. The Kaiser criterion (eigenvalue criterion) and the scree test were used to determine the number of initial unrotated factors to be extracted from the data set (Table 1). Eigenvalues were the ratio of common variance in the feedback responses and their specific variance explained by a specific factor extracted for the set of questions framed for various domains of learning objectives. The correlation between online learning effectiveness score based on the Osgood differential semantic scale and the midterm marks obtained by the candidate was calculated by Spearman correlation.

Statistical Analysis

The data analyses was performed with SPSS v.25 (SPSS Inc. Statistical Software, Chicago, IL) and Graph Pad Prism 8.01 for Windows (GraphPad Software, San Diego, CA). Factor analysis and principal component analysis were performed with SPSS v.25. Shapiro–Wilk test was used to check for the normality of the data, and Spearman correlation was used to test the strength of association between variables.

RESULTS

Distribution of gadgets and connectivity of the students’ use of technology for attending online classes were among the non-Likert questions (Fig. 1). Approximately 68% (56 students, ∼2/3) of students had regular internet access, and most of the students (42%) were attending the online lectures on their mobile phones. Furthermore, 56% of the students always had an exclusive gadget available to access online classes (Fig. 1). Mean response scores, standard deviations, and communalities of the responses of all the variable characteristics of the Likert scale are given in Table 2.

Figure 1.

Non-Likert questions assessing student use of technology. A: gadgets used for attending online classes. B: regular internet connectivity for access to online classes. C: had an exclusive gadget to access online classes. Pie chart distribution is the proportion of students in a class size of 83. Besides nominal responses, a blank response option was given in the questionnaire (B and C).

Maximum and minimum feedback response scores on the 5-point Likert scale were 4.89 and 1.83, respectively. The range of the items’ feedback response score was 3.06. Ratio of maximum and minimum was 2.67, and mean feedback response score on the Likert scale for all the variables was 4.69. For Bartlett’s test of sphericity of the dataset approximate chi square was 1,242.857 (degree of freedom = 51), and significance was 0.000.

Out of the initial loading of 1.00, extractions of the principal component variables are given in Table 3. Eigenvalues along with the extraction sums of squared loadings (both cumulative and percentage variance in them) are given in Table 3. Eigenvalues > 1.72 were selectively treated as principal components, and correlation of the crude components was done with a transformational matrix based on the rotational sum method: Varimax with Kaiser normalization (Table 3).

Table 3.

Factor analysis and principal component analysis

Components	Eigenvalues	% Variance	Extraction Sums of Squared Loadings, cumulative %	Rotation Sums of Squared Loadings
Supported Learning	9.098	33.695	33.695	5.941
Interaction	2.270	8.409	42.104	3.778
The Only	1.946	7.208	49.312	2.836
Active Listening	1.755	6.500	55.811	2.514

Eigen threshold for the principal components was set at 1.72.

There was no correlation (r = 0.019) between the Osgood differential semantic scale and the midterm marks obtained by the candidate; furthermore, it was found to be nonsignificant (P = 0.86) (Fig. 2). The relationship between components and eigenvalues of principal component analysis is shown in Fig. 3 in the form of a scree plot. Principal components were extracted based on their eigenvalues and the squared sum of their loadings (Table 3). Then components were correlated; sums of squared loadings cannot be added to obtain a total variance. Four components were designated as principal components based on the principal components and intercomponent matrices. The component transformation matrix for the same is given in Table 4. The principal components were

Figure 2.

Correlation between online learning effectiveness score based on Osgood differential semantic scale and midterm marks obtained by the candidate. Correlation coefficient was calculated by Spearman r and was found to be statistically nonsignificant (r = 0.019, P = 0.86).

Figure 3.

Scree plot: component-wise eigenvalue distribution. Relationship between components and eigenvalues of principal component analysis based on the candidates’ collective response.

Table 4.

Component transformation matrix: rotation method: Varimax with Kaiser normalization

Components	Supported Learning	Interaction	The Only	Active Listening
Supported Learning	0.742	0.524	0.345	0.237
Interaction	−0.517	0.137	0.388	0.750
The Only	0.182	0.037	−0.819	0.542
Active Listening	−0.386	0.840	−0.242	−0.294

• Component 1: The online mode supported my learning (r = 0.827; factor: Supported Learning).

• Component 2: There was adequate opportunity to interact online with the faculty (r = 0.789; factor: Interaction).

• Component 3: Online mode should be the only mode of learning (r = 0.696; factor: The Only).

• Component 4: How much of your interaction occurred with the faculty as compared to face-to-face classroom (r = 0.638; factor: Other Activities)?

DISCUSSION

The present study was aimed at understanding the role of virtual teaching (online mode of classes) in overall learning by undergraduate medical students during the COVID-19 pandemic amid nationwide prolonged lockdown. Eighty-three undergraduate medical students of a medical college answered a questionnaire consisting of 43 items about their experience of online classes, and the responses were recorded and analyzed online. The online mode of teaching was introduced for the very first time for medical students by Izet Masic in 2008 (5). It initiated a series of studies designed to assess the feasibility and effectiveness of online classes for medical students. Among the various components analyzed by Likert scale, four components emerged as principal components, which are also the primary outcome measures of the overall assessment by the students (6). There are many differences between virtual and physical classroom teaching. In a physical classroom, a teacher/instructor uses “immediacy cues,” eye contact, smiles, tone of voice to welcome students and support their contributions, in the absence of which online learners experience isolation as no one is there to convey all the nonverbal messages of support; nor can the students have the physical presence of other learners for the social support that is very crucial for active learning (7).

“Online mode supported my learning” is the first principal component. Averaged feedback response of the participants is between neutral and agreement that online mode/virtual classrooms supported their learning. Other studies from previous literature have also endorsed this viewpoint (8). “There was adequate opportunity to interact online with the faculty” is another principal component regarding the ability of students to interact with the faculty during the class. Only a limited (25–50%) opportunity was there, where students could interact with the faculty during the class. However, some students also declared that some interaction was possible with the faculty in an online class (9–14). The third principal component was “Online mode should be the only mode of learning”; only ∼6% of the students showed preference for online mode as the sole mode of learning physiology, and most of the students disagreed and preferred other modes such as physical mode as a more productive mode of learning, which seems to be a more acceptable mode for everyone in the medical field. “How much of your time during online class was spent actively listening to the faculty?” is the last principal component extracted. Students’ feedback to it suggests that only some of the times (25–50%) did students actively listen to the lecture during online classes, indicating a counterproductive way of learning and a very difficult situation to be assessed, detected, or controlled by teachers. A previous study, based on these principal components, also showed that online classes are a favorable mode of learning but required discipline on the part of the students, as they were prone to use other means of entertainment during the unsupervised online classrooms (15). The recent introduction of several virtual modes of communication to day-to-day life such as mobile phones, laptops, digital notebooks, and other gadgets has already eased our ability to communicate and dispose of information on a large scale. Moreover, the younger generation is more accepting and comfortable in using these virtual modes of attending lectures and taking physical notes in the classroom. However, there are certain drawbacks, such as

• 1. Assessment of attentiveness
• 2. One-to-one interaction with faculty
• 3. Lack of feedback to the teacher, which could improve the interaction
• 4. Enjoyment achieved by group learning/activities

Although it may be tempting to simply rely on some of the available technology to impart education in a large group of students, the direct conversion of a 50- to 75-min in-person teaching experience to an online-only format has been shown to be detrimental to student attention and knowledge retention (16).

In our study Cronbach’s α of standardized items was 0.91, which showed that the items used in the study were moderately dependable on the outcomes related to the feedback. Out of 28 Likert-type questions, 2 questions, “How much of your time during online class was spent actively listening to the faculty?” and “How much of your time during online class was spent in other activities such as email, Facebook, Instagram, etc.,” were found to be similar, so the latter was excluded from the factor analysis.

Furthermore, students may run into issues in connecting to an online class because of the lack of technology available to them, including a slow connection, no connection, or nonavailability of a computer (17). Unfortunately, these hurdles are commonly faced by students from economically weaker sections or living in remote areas. One suggestion for overcoming these issues is by using an asynchronous approach to teaching. This style uses many “modes” of teaching, including audio only, downloadable files, and/or a direct online presentation. These multimodal styles are often recommended as a method of getting around many of the hurdles that technology can place in front of online learning because of their flexibility (18, 19). Additionally, just as many educators are struggling with learning a new way of interacting, students may also be technically challenged when faced with new and foreign software. In the present study, major feedback from the students was that all the lectures should be saved and available online in case they miss the sessions. Most of the students wanted more interactive sessions and suggested including some student presentations since online classes do not have a real student-teacher interaction and hence the classes become monotonous. Since medical educators have been pushed inevitably to rely on technology-based learning (2, 3), they should not only embrace it but also develop and evaluate its sustainability and application in a preclinical and clinical setting. A study conducted in Nepal suggested that training teachers and students about online classes might create effectiveness of e-learning (20). Furthermore, the government needs to provide free internet services to remote areas and students of lower economic conditions since a large margin of the population are struggling with economic burdens (20). Meanwhile, the students, whose medical education is stuck in the pandemic time, should realize that there is no bigger teacher than first-hand experience.

Conclusions

Our study suggests that virtual physiology teaching of undergraduate medical students after the COVID-19 pandemic has only a moderate level of effectiveness. Validation through principal components and factor analysis has revealed that the application of online teaching is not an effective method owing to certain limitations compared with the traditional classroom teaching methods. Classroom teaching ensures better quality of learning and student-teacher interaction in medical education.

DATA AVAILABILITY

Data will be made available upon reasonable request.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

N.A., B.M., K.K.D., and R.B. conceived and designed research; N.A., B.M., K.K.D., and R.B. performed experiments; N.A., A.K., B.M., and K.K.D. analyzed data; N.A., A.K., and R.B. interpreted results of experiments; N.A. and A.K. prepared figures; N.A., A.K., B.M., and R.B. drafted manuscript; N.A., A.K., B.M., and R.B. edited and revised manuscript; N.A., A.K., and R.B. approved final version of manuscript.