Participants' perceptions & expectations

By design, the IBT course is taught over 3 months, to provide an introductory level knowledge and skills to basic bionformatics users.  Besides providing relevant and high standard content, the core IBT team and various instructors have placed great emphasis on defining clear learning objectives and outcomes prior to each of the taught modules \cite{Via_2011}, and maintained the logical structure of the course despite removing the structural proteomics module in the 2017 iteration \cite{Gurwitz2017}
Therefore course participants had clear expectations out of each module, and we see largely positive indicators in their responses about all the taught 6 modules (Figure \ref{197355}, and supplementary 17). This satisfaction about the course is also reflected in the high success rate of both classrooms in the University of Khartoum node(excluding early withdrawals, the success rate, i.e. percentage of participants satisfying the course requirements is 94%, see also section \ref{105477}). Considering that the majority of the participants are in their MSc or PhD level (Figure \ref{189578}C, supplementary 11) in a domain related to genetics or molecular biology (supplementary 6), it is expected that some of them had some exposure to concepts and uses of  Databases; however, less familiarity with Genomics and related topics, because there is none or limited postgraduate degrees in bioinformatics currently in Sudan.
In terms of utilizing local and remote classroom resources, we make a few observations. While we see that our participants were comfortable networking with each other, we see less interactions with participants from other classrooms. In part, we can attribute this to the demographics of the course participants, as was elaborated in section  \ref{105477}, in terms of the highest awarding institute of each participant and also how the majority of participants heard of the course (supplementary 10). While a few minutes were spared at the beginning of each session for the scattered classrooms to introduce themselves via the videoconferencing system, mconf, this was ineffective in linking our participants with the other classrooms (Figure \ref{636892}) because often times the connection would be too noisy to yield meaningful conversation. The same issue often arises during the discussion time at the end of each session where the instructor is typically available and answering questions live. The alternative our participants would employ in this case is discussing issues through the forum or the chat rooms available through the learning management system, Vula. We do indeed see more engagement in these avenues towards the end of the course in Figure \ref{636892}.
Another element that could explain the high success rate of the 2 classrooms run in the University of Khartoum, is the local help provided to the participants through the local teaching assistants who were themselves alumni of the 2016 iteration of the IBT. From one perspective, testimonials from successful alumni have been shown to improve the sense of belonging in MOOCs, and their by their by learners' performance \cite{Kizilcec2017}. This is especially true as the 2017 iteration of the IBT included training sessions for the local team in each classroom on how to best facilitate the course. These training sessions employed the  Mental contrasting with implementation intentions (MCII) model  \cite{Kizilcec_2017} in equipping the local staff with best strategies to facilitate the course by asking them to set goals for the course, and then predict future challenges and devising solutions for them should they occur. While not all identified challenges were within direct control (see section \ref{420328}),  collectively looking at problems beforehand gave a sense of confidence to the local staff. It should also be mentioned that the majority of our volunteering staff already had prior teaching experience (7 out of the 8 TAs, in addition to the class PI). The responses from the TAs do indicate that they benefited from the previous IBT experience, and also that they were comfortable with facilitating such a class with different demographics (supplementary 18). On the personal level, we also see the TAs highly satisfied with the experience, even though they had other pressing commitments  playing at the same time as the IBT (supplementary 19).
A final element in examining the IBT experience of 2017 is the non-cognitive factors pertinent to the course participants. These factors, unexamined in section \ref{105477}, include the expected load incurred by participating in the IBT, which many found to be overwhelming when compared with commercial 1 week bioinformatics courses, that have no exams or assignments. This perception has exacerbated especially with the Linux module as can be seen in Figure \ref{287326}. Another contributing factor may be the language of the course. Sudan is officially an Arabic speaking country, though university educated graduates have some competency in English. However, one teaching style commonly employed, especially at the undergraduate level, is to deliver teaching curricula in Arabic, while maintaining the English keywords and having English-based handouts (Supplementary 20). This means that for many of the participants, the experience of complete course delivery in English can be seen daunting. Collectively, these factors reasonably explain the large withdrawal of the course before the mid-point (26% of the total participants- also see Figure \ref{287326}).