05 02

Udacity statistics final exam answers

udacity statistics final exam answers

With my lab (MIT ALFA group) and various help from edX and MIT, I mined the data of the first offering of 6.002x: Circuits and Electronics (Spring 2012). Most of the analyses we have done are on http://bit.ly/MOOCdb.

Prior to the analyses, we had to find a right database schema that we could use for any MOOC. We called this schema MOOCdb, and it's also presented on http://bit.ly/MOOCdb.

Some figures for edX 6.002x: Circuits and Electronics (Spring 2012):

Course statistics: 6.002x had 154,763 registrants. Of these, 69,221 people looked at the first problem set, and 26,349 earned at least one point on it. 13,569 people looked at the midterm while it was still open, 10,547 people got at least one point on the midterm, and 9,318 people got a passing score on the midterm. 10,262 people looked at the final exam while it was still open, 8,240 people got at least one point on the final exam, and 5,800 people got a passing score on the final exam. Finally, after completing 14 weeks of study, 7,157 people have earned the first certificate awarded by MITx, proving that they successfully completed 6.002x. (stats given on Course Info - MITx 6.002x, and I sound the same when looking at the data)

Final grade distribution of students who received a certificate in 6.002x. Students who earned at least 60% received a C, at least 70% received a B, and who earned at least 87% received an A. Over 59% of students who received a certificate got an A in the class:

Percentage of registrees who received a certificate by country. Interestingly, Eastern Europe had the largest percentage of students who successfully completed 6.002x. Hungary led with 16.2% of students receiving a certificate. The U.S. had 5.1% completion rate:

The number of students who registered for 6.002x by country. The U.S. and India had the largest absolute number of students:

Percentage of students who registered for 6.002x of the total population of the country. Countries with populations under 50,000 were excluded from the list. Great Britain and Columbia had among the highest percentages with 0.0132% and 0.0134% respectively:

The number of students who registered each day since registration opened. Days are delineated by midnight UTC.

From this graph, we see three clear spikes of when users registered for 6.002x. The first and largest, on 02/14/2012, shows the largest number of students registered right after registration opened. 17248 students, about 11\% of the students, registered on that date alone. The next spike, around 03/05/2012, corresponds to the start of the course. The final, and smallest spike, around 05/02/2012, corresponds to a series of press releases announcing that Harvard will join MITx to create edX.

The number of observed events each day of the course Here, an observed event is a user accessing a 6.002x resource. A resource is any page of 6.002x content (including videos, forums, problems, etc.), differentiated by URL.

We see a series of sharp spikes, corresponding to the due dates of homeworks, labs, the midterm and the final. We also see a slow decay of resource access, showing the drop-off of students throughout the semester.

The duration of observed events, averaged by country. The duration of a resource is difference in timestamps between subsequent resource accesses (per user), with a maximum of 1 hour. A longer average duration would indicate students spending more time on each resource before moving on to another page of the site.

The average duration of observed events is fairly consistent across countries, ranging from around 39 to 73 seconds.

The cumulative time spent on resources, averaged by grade in the class. It shows a correlation between time spent and score, although A and B grades had relatively similar cumulative times.

Average number of submissions per country of students who received a certificate. Students in countries with a larger average submission number show less reluctance in submitting many times before getting the correct answer. Eastern Europe had the largest numbers of average submissions. We see a correlation between this and the percentage of the students who earned a certificate.

Observed events per submission per country of students who received a certificate. This number gives a measure of how long a student will spend on course material before submitting an assignment.

The number of forum posts each day after the forum opened. Days are delineated by midnight UTC. Posts by users who earned a certificate are overlaid in red.

In this graph, we see a large spike when the forum first opened. The largest spike in forum posts is on 06/11/2012, one day after the final exam. This most likely is due to students discussing the final and grades on the forum. We also see weekly cycle spikes in forum activity that match the due dates of the homeworks, labs and the midterm.

The number of wiki edits each day after the wiki opened. Days are delineated by midnight UTC. Edits by users who earned a certificate are overlaid in red.

Similarly to the forum posts, we see spikes corresponding to the due dates of homeworks and labs. We do not see much more activity on the wiki around the final or midterm.

An overlay of the number of forum posts the wiki edits. Interestingly, we see a 1-3 day delay of the spikes in activity between the two types of collaboration:

Average length of forum posts for those users who earned a certificate and those who did not. If a student never posted, he was ignored in this number. The average forum post length for students who earned a certificate is 20% longer than the average for students who did not.

Here is the database schema we designed and that we used to generate these graphs:

And some code snippets to show how we plot the graphs (I wrote a wrapper for MATLAB and a wrapper for Google Chart API in Python):



All these graphs are on http://bit.ly/MOOCdb (PDF and interactive versions to see the exact numbers).


  1. Yeqadawopiges

    Legit so sick of doing statistics I would literally opt to take my final TODAY if I could.

  2. Yuneqic

    Anyone at Cal State LA who can help me on this final Statistics Report?

  3. Lusokonacoze

    HONESTLY, fuck statistics. I"m still bitter over the final and the grade I got in the end I don"t blame her if she doesn"t want to study

  4. Foloyoye

    Can"t wait to be up allllll night doing my statistics final because I have no idea how to do anything on it

  5. Voxoharukoy

    96 on my applied probability and statistics final!!!! 18 points above the average!!!!!!

  6. Xarixovevoday

    Final Louisville Leopards JV Baseball Batting, Pitching, and Fielding Statistics:

  7. Tasebowoqetoxu

    So proud of myself for not going to the bar tonight, statistics final and a hangover = terrible terrible idea

  8. Qaxabacoca

    3 amazing presentations for the final statistics research project today. Looking forward to the rest of them next week!!

  9. Fiqeqecejekuy

    Statistics final today

  10. Gidenulom

    Some interesting and updated records statistics for there. Will Demetres and/or Kevin win those Final Parts for F3 HOH? We"ll see!

  11. Xiqeharohuyuhe

    My statistics final is tomorrow so please wish me luck

  12. Sikorodahes

    This dude in my statistics final was talking strategy to the girl next to him on how he was going to cheat off of her exam, like oh???

  13. Xofeyeqovepa

    When I thought my Statistics final has 2 parts to it for today and next week but turns out there"s only 1 final and it"s next week

  14. Xohifuh

    I got a 97% on my Statistics final I was super worried about, and I got an 100% on my Developmental Psychology final ^.^ very happy!

  15. Weseduxahoj

    No like my statistics final is a take home who wonna help a nigga out

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>