Attendance & Withdrawal patterns

Students attendance and retention is a major concern contributing to a successful MOOC experience \cite{Hew2014}especially that to many learners, the problem is about fitting the needed study hours into their busy schedules, and being on the same page as the course progresses. 
In this 3 months course, we note that out of the 73 registered participants, the attendance rate was higher at the beginning of the course (~85%), then it dropped progressively towards mid-June and early July (Figure \ref{287326}). This accompanied a steep  increase of the withdrawal curve which can be related to 2 factors: 1) the Linux module, which is understandably challenging as the material naturally requires a bit different mode of thinking than what wet lab biologist are typically used to.  2) culturally, the IBT course started just a few weeks before the holy month of Ramadan coinciding with end-of-year/semester holidays in many universities and colleges. As can be inferred from Figure \ref{189578}A, during the few starting weeks it was easy for students (53% of the total IBT participants) both full-time (44%) and part time  (9%) and other academic staff (22% of the total IBT participants) alike to follow up with the progress of material and activities in the IBT. 
However, once these holidays were over, those IBT participants (collectively, 75%  of the total IBT participants) needed to be back to full time working hours and classes in their respective institutes, making it harder for them to attend IBT sessions in person and timely submit their assessments or work towards their tests. It is also clear from Figure \ref{287326} that the withdrawal pattern plateaued after this point in time, such that in total 5 participants withdrew from the CBSB classroom and 14 from the Main Library classroom, making up a total of 19 participants (26% of the total intake of participants for the 2017 IBT).
A remarkable observation seen from Figure \ref{189578}B is that the majority of the IBT participants (73%) have graduated from the University of Khartoum. Furthermore, all the local course staff have studied at the University of Khartoum. This could have been the reason why ~63.6% of the participants heard of the course through friends, 27.3% from their supervisors or mentors, and the remaining 14.5% through social media (Email, Facebook and Twitter) (Supplementary 9). This also suggests that local circles of friends/ acquaintances were already in place before the course had actually started. This support system in place could explain why for those participants who stayed in the course to the end, 51 out of the 54 remaining  participants (94%), they  did actually satisfy the requirements of the course, and only 3 participants (6%) failed. 
Given the high success rate of participants in both classes, it is more interesting to look for factors that contributed to a participants withdrawal from class (as opposed to failure). We investigate such predictive models based on the demographics of the class in the Discussion (section \ref{105477}) ( Figure \ref{364861}supplementary 11,6 ). We note however that there are unmoderated personal effects; like participants having to travel with their families, either for holidays or accompanying a close sick family member for medical treatment abroad; or for some work commitments that could not be waived. 

Participants' perceptions & expectations

The data on the Sudan IBT 2017 participants' the Sudan IBT 2017 participants' perceptions and expectations were collected from the 3 surveys (supplementary 2, 3, 4 ) disseminated at the start of the course, the mid-point and the end of the course as shown at the times indicated by the S, E and M boxes in Figure \ref{287326}, respectively.  Out of the 73 participants, only 33 (45%)  filled all the 3 surveys (Supplementary 10), while 15 (21%) never filled any.
The experience of the IBT course is both unique and new to the participants considering its blended multi-delivery learning model  \cite{Gurwitz_2017}, and also its extended 3-months time span.  Our surveys aimed to check the alignment between the participants' expectations and the course scope \cite{Via_2011}. On a labeled five-point Likert  scale from very uncomfortable to very comfortable, we asked participants about their perspectives in terms of their prior experience level in each IBT course module (start survey); the extent to  which the content was appropriate, and the level it met their expectations (in the mid-course and end surveys, for each module taught until that point). 
Participants' perceptions for each of the 6 modules of the IBT in its 2017 iteration largely followed the same trend (Supplementary 17), hence they are illustrated as an average of responses across all modules in  Figure \ref{197355}.  Not surprisingly, most participants were largely unfamiliar with the various modules (with the 75th percentile of responses below neutral familiarity level),  and this was especially true for the Linux module given their background ( Supplementary 6). To the contrary, we see higher satisfaction levels (with the median of the averaged responses at a level above Comfortable in Figure \ref{197355}) in terms of satisfaction with the appropriateness of the taught material and meeting participants expectations.