<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <title-group>
        <article-title>Towards Real-Time Analytics in MOOCs</article-title>
      </title-group>
      <contrib-group>
        <aff id="aff0">
          <label>0</label>
          <institution>Daniel T.</institution>
          <addr-line>Seaton</addr-line>
        </aff>
        <aff id="aff1">
          <label>1</label>
          <institution>Department of Electrical Engineering and Computer Science</institution>
        </aff>
        <aff id="aff2">
          <label>2</label>
          <institution>Department of Physics and RLE</institution>
        </aff>
        <aff id="aff3">
          <label>3</label>
          <institution>Frequency of Accesses Illustrates Activity</institution>
        </aff>
        <aff id="aff4">
          <label>4</label>
          <institution>Office of Digital Learning</institution>
        </aff>
        <aff id="aff5">
          <label>5</label>
          <institution>Participation: Attrition</institution>
          ,
          <addr-line>Tranches, and Total time</addr-line>
        </aff>
        <aff id="aff6">
          <label>6</label>
          <institution>edX and CSAIL, Massachusetts Institute of Technology</institution>
          ,
          <addr-line>Cambridge, MA 02139</addr-line>
          ,
          <country country="US">USA</country>
        </aff>
      </contrib-group>
      <abstract>
        <p>Massive open online courses (MOOCs) collect essentially complete records of all student interactions in a self-contained learning environment, with the benefit of large sample sizes. Building on our data mining of the first course in MITx (now edX) we demonstrate ways to analyze data to illustrate important issues in the course: how to distinguish browsers from certificate-earners, which resources were accessed the most and how much time was allocated by certificate-earners. Each topic is addressed via appropriate displays that, in future courses, can be updated in real time. Furthermore, we stress that analytics can provide useful information to teachers, to resource creators (authors), and to members of organizations trying to improve their MOOCs.</p>
      </abstract>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>-</title>
      <p>The number of accesses of various course components in 6.002x by those active
each day, which we call activity, is shown in Fig. 2. Homework and laboratory
activity has consistent periodicity with the number of unique users per day, while
lecture questions do not share this periodicity and also suffer a downward trend in
overall activity for the term. For learning based components, Discussion activity is the
only component sharing periodicity with the unique users per day, suggesting a
correlation with for-credit activity. These observations may imply that students
working on graded activities do not utilize many learning-based resources.</p>
      <p>Time represents the principal cost function for MOOC students, and it is therefore
important to study how students allocate time among available course resources. Of
the course components offered in 6.002x, Lecture Videos and Homework generally
took the most time each week. Discussion Boards (which were voluntary) represent
the next highest level of time allocation by students. It is also interesting that the
Discussion time trends upward relative to homework time for later weeks in the
course, suggesting increased use by students doing homework. A downward trend is
observed for Lecture Questions. Other course components have consistent time across
the course, but appear to have minimal activity.
4 References.</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          1. 6.002x: Circuits and
          <string-name>
            <surname>Electronics</surname>
          </string-name>
          . - https://6002x.mitx.mit.edu/
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          2.
          <string-name>
            <surname>Guzdial</surname>
            ,
            <given-names>M.</given-names>
          </string-name>
          (
          <year>1993</year>
          ).
          <article-title>Deriving software usage patterns from log files</article-title>
          .
          <source>Tech Report GIT-GVU-93-41.</source>
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          3.
          <string-name>
            <surname>Seaton</surname>
            ,
            <given-names>D. T.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Bergner</surname>
            ,
            <given-names>Y.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Chuang</surname>
            ,
            <given-names>I.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Mitros</surname>
            ,
            <given-names>P.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Pritchard</surname>
            ,
            <given-names>D.E.</given-names>
          </string-name>
          (
          <year>2013</year>
          ).
          <article-title>Who Does What in a Massive Open Online Course</article-title>
          ? In Press. - Communications of the ACM.
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>