For JusticeX, edX has records of 79,787 students who registered for the course. In most ways, these registrants are similar in demographic characteristics to students in the other four early HarvardX large-scale courses. The registrants were about 60% men; nearly half were in their 20s, and as a group they were highly educated. In terms of their highest completed degrees, over 70% of students had earned at least a four-year college degree. Of all registrants, 37% had a bachelor’s degree, an additional 28% had a master’s or professional degree, and an additional 5% of registrants had earned a doctorate.
Figure 1. Gender distribution of JusticeX registrants (n=72,980; 6,807 missing) and registrants from first four other 2012–2013 HarvardX large-scale courses (n= 311,080; 28,447 missing).
Figure 2. Educational attainment distribution of JusticeX registrants (n=72,599; 7,188 missing) and registrants from other HarvardX large-scale courses (n=309,617; 29,910 missing).
Figure 3. Age distribution of JusticeX registrants (n= 72,089; 7,698 missing) and registrants from other HarvardX large-scale courses (n=308,934; 30,593 missing).
As with other HarvardX courses, JusticeX students came from all over the world. Nearly a third of registrants with identifiable locations came from the United States, but many students also came from India, Brazil, China, and over 150 other countries around the world.
Table 2. Counts and percentages of students country of residence among 72,990 with identifiable locations. (6,797 missing)
Participants began enrolling in the course as soon as registration opened, and they continued doing so even beyond the end of the course. From the opening of registration until course launch, registrations averaged 518 people per day. After the course launched, an average of 220 people registered per day.
Figure 4. Cumulative enrollment for JusticeX (n=79,787).
Of those who signed up for JusticeX, the degree and kind of participation in the course varied considerably. In other studies of large-scale online courses, researchers have clustered students based on assessment patterns (students who complete the first assignment, students who complete all assignments, etc.). In JusticeX, however, the emphasis of the course was much more on participation and access than on assessment or certification. Indeed, the course team considered their multiple choice assessments to be the “least bad” of the assessment options available to humanities instructors, and they placed little emphasis on quiz completion. Students were encouraged to watch videos, participate in discussions, and complete practice and for-credit questions as suited their own interests. There are multiple meaningful ways to participate in JusticeX.
To illustrate these diverse pathways through the course, Figure 5 shows a scatterplot of participant grades versus a simple metric of course activity: the number of the 24 “chapters” or courseware sections the students opened and viewed.
Figure 5. Scatterplot of grades versus chapters viewed for JusticeX registrants (n=79,787).
This scatterplot of activity illustrates the diverse ways that students used the JusticeX courseware. In the top section of the figure, we see that most participants who earn a certificate viewed most or all of the course content, but some certificate earners viewed only part of the course, and a few passed the course by only viewing the required quiz and exam chapters. Some of these users may already have been proficient with the course material and only wanted expeditious certification of their proficiency. We describe this as “optimizing” behavior for certification. However, most certified students are in the top right, exhibiting patterns akin to those of a “completionist,” students who get a perfect score and view all of the content.
In the bottom right section of the figure, we show those who did not earn a certificate, but viewed more than half of the course chapters. We selected this cutoff to identify students who viewed a substantial proportion of the chapters, and we describe these students as having “explored” the curriculum. In the bottom right of the figure, we circle students who exhibit behavior consistent with “listeners”—borrowing the MIT term for a course auditor—who viewed every chapter and answered zero for-credit questions correctly. In the bottom left, we show students who viewed fewer than half of the chapters, and we describe these participants as having “viewed” the course. In the bottom left, we circle those students who never opened the courseware at all after registering for the course. The coverage of data across this scatterplot reflects the very diverse ways in which people engaged with the JusticeX courseware.
Motivated by this variation, we defined four subsamples of participants to explore in this series of HarvardX course reports. In JusticeX, Registrants consist of all 79,787 students recorded by edX as having registered for the course. Viewers are the 50,044 registrants who opened up the courseware. The courseware is the section of the course site containing the content of the course: the lecture units, videos, poll questions, and problems. Explorers are the 8,415 registrants who viewed at least half of the chapters in the courseware. Certificate Earners are the 5,442 registrants who earned a certificate in the course. As the example of the optimizer shows, not all certificate earners are explorers. To clearly distinguish among these nested groups, we identify (in Figure 6) four mutually exclusive and exhaustive sets of students who “only registered, “only viewed,” “only explored,” and were “certified.”
Figure 6. JusticeX participants clustered into four disjoint sets (not to scale) (n=79,787).
While these groups are unevenly distributed in numbers, they are similar in demographic composition. Compared to the overall gender distribution that was 60% male, certified students were 62% male, and “only registered” students were 58% male. “Only explored” students were slightly more educated than other groups, with 72% having a Bachelor’s degree or higher, and they were slightly older, with a median age of 31 compared to the overall median age of 28. Broadly, however, demographic characteristics did not strongly correlate with particular patterns of activity.
Figure 7. Distribution of females and students with Bachelors degree in four disjoint groups of JusticeX students (n=79,787).