Use of WhatsApp®, a Simple and Widely Distributed Social Media, for Distance Teaching during COVID-19 Pandemic: Experience and Perception from a Sub-Saharan African Setting

Background: In the midst of the COVID-19 pandemic, to palliate to the lockdown and cover academic programs, the Faculty of Medicine and Pharmaceutical Sciences of the University of Dschang (FMPS-UDs) in Cameroon has implemented e-learning using WhatsApp®. We aimed to describe the opinion of students and lecturers about it after its implementation at the FMPS-UDs. Methods: We designed a uniform scheme to teach using WhatsApp® during the university lock-down, and both the students and the teaching staff of the FMPS-UDs received clear information on the implementation. At the end of the online-teaching period of two and a half months prior to the reopening of the university, we surveyed our students and teaching staff. Sociodemographic characteristics and opinions about e-learning were collected using a standard questionnaire. Results: We enrolled 229 students and 40 lecturers of the FMPS-UDs. Students reported a poorer internet connection (p<0.001, p-homogeneity< 0.001) despite an increased expenditure related to internet use. Electronic devices were broadly used before the implementation of mobile learning. The use of course materials was more challenging among students because of the size/format of lecture notes and the internet connection/cost (all p<0.05). Perception of discipline compared to classroom-based lessons was similar in students and lecturers (all p>0.05). While lecturers were mainly [n(%)=16(40%)] more comfortable conveying the contents of their lectures, students tended to be less prone to actively participate [n(%)=137(59.8%)]. The motivation and satisfaction of the latter group toward e-learning were modest compared to classroom-based lectures while their feedback about the organization was positive. Conclusion: E-learning using WhatsApp® could be an effective alternative to conventional classroom-based lessons in SSA. The use of a blended-learning program including classroom-based sessions could help improve its limitations.


Background
The novel Coronavirus disease  outbreak has led to a lock down in almost all countries of the world. By March 18, 2020, 107 countries had implemented national school closure as a major preventive strategy [1]. In Cameroon, Central Africa, the government response strategy to the coronavirus pandemic included the closure of all public and private training institutions of the various levels of education, from nursery school to higher education, including vocational training centers and professional schools [2].
In this context, the Faculty of Medicine and Pharmaceutical Sciences of the University of Dschang (FMPS-UDs) had to set up a strategy to deliver academic programs and to ensure the continuity of training. To comply with the need of social distancing during the COVID-19 pandemic, internet-based teaching was the most convenient strategy. Given the fact that the faculty was not prepared to face this challenge and considering the urgent need to nd a solution, the choice felt on WhatsApp® that would be used as a transmission channel. This social media offered the advantage of being widely used by both teachers and students before containment measures. It also offered a minimum of functionalities in terms of transferring text, video, sound and image les to users within classes made up of groups. Finally, it offered the opportunity to be used both on smartphones and computers without the need of heavy logistics (servers) [3,4,5]. Moreover, it has been shown in another context to be more applicable in settings where there are challenges to obtain permanent or regular optimal internet bandwidth or speeds [6].
The rapid implementation of this teaching strategy via WhatsApp® made it possible to continue the academic program as soon as from March 18, 2020, the day Cameroonian universities closed. The purpose of this work was to describe how mobile-learning was put in place and the perception and di culties encountered both by students and faculty members of the FMPS-UDs in the implementation of this remediation strategy to cover the academic program [7,8,9].

Study setting and population
A cross-sectional study was conducted among students and lecturers of the FMPS-UDs, from April 26 to May 4, 2020. This faculty is one of the establishments of the University of Dschang located in the city of Dschang, in the West Region of Cameroon since 2017. The University of Dschang is one of the main universities in Cameroon with more than 600 lecturers and almost 30, 000 students. The population included in this study was recruited among students and lecturers (permanent and non-tenure track faculty members) of the FMPS-UDs. By the time of conduct of the study, the FMPS-UDs had 40 lecturers and 698 students. All participants provided informed consent. All methods were carried out in accordance with relevant guidelines and regulations Description of the WhatsApp® e-learning scheme The implementation of e-learning was done using the same lectures' schedule prior to the lockdown. Virtual classes pertaining to the various study levels were created through WhatsApp® groups. Those platforms were used by all lecturers who were invited to join the class at the day and time of the planned lecture. Lecture notes were rst sent in Portable Document Format (PDF), word or power point format and lessons were given via voice-mails. Interactions (questions and answers) during the lesson used the same channel or short message service as an alternative communication medium.

Study population
Students and lecturers of the FMPS-UDs were approached regardless of study type and level for students and independently of administrative position or teaching subject for the lecturers. Clear information was given to all students and faculty members about the study purpose, adequate lling of questionnaires and participation modalities. All participants who responded to the questionnaire were included. Procedure Data were collected using an anonymous online survey administered in both o cial languages (French and English) using Google Forms (Alphabet Inc., California, USA), supplementary materials 1 and 2. Data complied with the terms and conditions of Google Forms. In an initial step, we conducted a pilot survey to optimize data collection forms. Responses to the pilot survey were excluded, and all issues raised during that preliminary phase of the study were addressed. We then distributed the survey link to students directly via the various class representatives and lecturers via a WhatsApp® (Facebook Inc., California, USA) group. The class representatives disseminated the survey links in the respective o cial students' WhatsApp® groups. WhatsApp® was already the primary mean of formal communication among students and lecturers of the FMPS-UDs. We collected the total number of students in each group to calculate the participation rate, link dissemination was con rmed by screenshots, and reminder messages were sent every week in the class groups to ensure a reasonable participation rate.

Variables
Using standard questionnaires, data were collected on sociodemographic (age, gender study level) characteristics, the challenges and the opinion of the study population about internet-based distance learning. Survey questions were a mix of open, close, and multiple-choice questions.

Statistical analysis
The data collected were processed on Excel 2016 (Microsoft Corp., Washington, USA). We analyzed data using Stata 16 (StataCorp, Texas, USA). Figures were generated with GraphPad Prism 8 (San Diego, California, USA). Quantitative data are presented as median (interquartile range [IQR]) and qualitative data as frequencies and proportions. Group comparisons were performed with Chi-square or Fisher's exact tests. Variations of the ow of internet access, expenditure and connection time for academic purposes after implementation of internet-based lectures were assessed using Bowker's test for table symmetry. Statistical signi cance was adjudicated using Stuart-Maxwell test for marginal homogeneity.
The threshold of signi cance was set at 0.05.

Study population and internet access before online-based lectures
The study population included 229 students and 40 faculty members with median age (IQR): 20 (19)(20)(21)(22) and 39 (36-43) years, respectively. This yielded a participation rate of 32.8% and 100% respectively among students and faculty members.
Female students were more represented than their male counterparts, n = 140 (M/F ratio 0.64) while lecturers were more men than women, n = 29 (M/F ratio 2.6).
The large majority of students were registered for biomedical sciences [n(%) = 100 (43.7)] followed by medical studies [(n%) = 84(36. 7)]. Students in their rst and second years of studies were more represented [n(%) = 77(33.62) and n(%) = 82 (35.8), respectively]. All students had internet access before the implementation of distance learning. The same was true for lecturers who had overall better internet access than students (overall p < 0.001), Table 1. While students reported signi cantly poorer internet connection during the implementation of online courses than before (p < 0.001, p-homogeneity < 0.001), internet access was similar among faculty members before and during e-teaching (p = 0.549, p-homogeneity = 0.351). Age is expressed as median (interquartile range), *clinical biology and public health, **overall p < 0.001.

Devices used for internet access, nancial and time burden of e-lectures
At the time of the survey and implementation of online lessons using WhatsApp®, one lecturer and seven students did not have a smartphone. However, they had a laptop or a notebook to comply with teaching duties and to follow lectures respectively. Both students and faculty members mainly used either a

Challenges and opinion of respondents about internetbased lectures
Despite a relatively good internet connection, students and lecturers faced di culties either to download or upload lecture materials via the social media network.
This challenge was more encountered among students and was mainly related to the les' size and format, internet connection and cost (all p ≤ 0.015).
Opinions of both study groups were similar about discipline during lessons (all p > 0.05), Table 2. While lecturers' feeling was that they were more comfortable answering questions during distance learning

Discussion
In this report, we share our experience on the implementation of mobile-learning and the perception and di culties encountered both by students and lecturers at the FMPS-UDs. Mobile-learning was used as a remediation strategy to cover the academic program during the university lockdown due to the COVID-19 pandemic in Cameroon. Our results suggest that teaching equipment was largely available among students and faculty members. The major challenges related to the implementation of distance learning using WhatsApp® in our study population were to access internet and to share teaching materials. Despite a pretty good satisfaction about the organization, this teaching alternative warrants complementary measures to be as attractive as conventional classroom-based learning method for students in our study population.
Smartphone ownership rates were high among students and teachers. Several studies have found a high availability and use of smartphones among medical students in many countries [10]. This was an important prerequisite for the implementation of mobile-learning. Almost three out of four students spent at least four hours a day on WhatsApp® while this proportion corresponded to 43.6% and 51.9% of students at Albaha and Dammam universities, Saudi Arabia, respectively [11]. This extended time on WhatsApp® is not necessarily linked to the implementation of mobile-learning since students use this application also for the purpose of social exchanges. As discussed, by Alkhalaf MA et al., this can be a source of addiction, sleep disorders and poor academic performance [11,12]. Therefore, in the case of concomitant use for social and academic purposes, students should be reminded of the risks associated with prolonged use of this application.
In a resource-limited setting such as that of our study, one needs to consider the increase in the internet connection budget. This could be a limiting factor for the participation of students in online teaching. However, the expenditure must be weighed against the reduction of other costs, those related to transport for example, which could have been reduced due to the lockdown. Besides the cost, the quality of internet connection before and during the lockdown is an additional important factor to consider. A slow Internet connection wastes time, increases the cost of connection and can be a source of poor scoring of elearning for both students and teachers. Indeed, one of our observations is the challenge of downloading and uploading course materials. Although both students and lecturers increased their expenses related to internet use, this did not positively impact the quality of the connection. Considering the di culties encountered by students and teachers in managing online courses, the major challenges were linked to the quality and cost of the internet connection. Among students, internet speed and related expenses were a challenge for 87.8% and 82.1%, respectively. These di culties were signi cantly greater for students than lecturers (both p < 0.001). In a Liberian study about introducing e-learning in medical education, the authors found that limited bandwidth was one of the main limitations of distance learning [13].
Among students, 96 (41.9%) felt that there was less discipline during online lessons than during face-toface lectures. The difference in the perception of discipline from the perspective of students and that of teachers 21 (52.5%) was not signi cant. While it is true that online teaching gives an impression of freedom (since it can be followed from home), participants should be self-disciplined and teachers should regularly remind students the rules applicable to this speci c teaching method. Indiscipline here can take more subtle forms such as not logging in at the right time or logging out during class without a grounded reason [14]. In our study population, students and lecturers were rather satis ed with the modalities and organization of e-learning at the FMPS-UDs during the lockdown. However, the majority of surveyed students declared to be less satis ed, less motivated and less prone to actively participate in lectures compared to traditional classroom-based lessons. This could be related to the eld and level of study of students included in the present survey. Moreover, the intrinsic nature of e-learning which closely relates to students' affect can not be underestimated [15,16]. In fact, the vast majority of included students were undergraduate students. One could expect less propension to independent work, more need of contact with the tutors and peer interaction in this population [17]. Besides, our data show that 87.0% of the study population thought that add-on classroom-lectures would improve e-learning using WhatsApp®. Although consistent with previous reports, this needs to be explored in subsequent studies considering as endpoints students' and lecturers' satisfaction and the impact on students' performance [18].
Our survey has limitations that need to be acknowledged. The study was conducted in one-third of the overall students' population at the FMPS-UDs.
This could suggest a selection bias related to the relatively low response rate. However, the absolute number of responses should be acknowledged. Furthermore, there is no formal consensual threshold sample size to be acceptable in such studies [19]. Also, postgraduate students are underrepresented in our study population. In this subset of students, one would expect more autonomy to appraise and to follow online lessons. The strengths of our study include the high response rate of lecturers (100%) and the survey of a heterogeneous group of students considering their years and elds of study. However, the impact of level and type of study on the need of complementary lectures using conventional learning method needs to be explored in subsequent studies.

Conclusion
In the context of the COVID-19, social media-based e-learning could be an applicable alternative to conventional classroom-based lessons. However, solutions to improve students' satisfaction and motivation need to be sought. This could potentially be addressed by de ning a blended-learning strategy where a speci c number of hours of classroom-based courses, to be identi ed, could be implemented in limited groups of students. The effectiveness of this strategy to optimize delivery and assimilation of academic programs and to impact academic performances needs to be investigated.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics declarations:
Ethics approval and informed consent to participate The study was approved by the institutional board of the University of Dschang and the Faculty of Medicine and Pharmaceutical Science. Participation in this study was voluntary and informed consent to participate was a prerequisite to the survey.

Consent or publication
Participation in this study was voluntary and informed consent to participate was a prerequisite to the survey. A study information sheet explaining that the data could be used for the purpose of publication was included at the beginning of the survey.