15.871 - 15.874 15.871/874
Fall 2007
Instructors: John Sterman, Paulo Goncalves
TAs: Taylor Moulton, Kawika Paul Pierson, Nirav B Shah, Qi Zhang
Lecture: A: 10.00 - 11.30; B: 13.00 - 14.30 (Both sections in E51-345)
Class Info:
15.871, Introduction to System Dynamics, is a half-semester (6 unit) course. 15.874, System Dynamics for Business Policy, is a full-semester (12 unit) course. During H1, the two courses meet concurrently and are the same. In the second half of the semester, 15.874 continues.
Announcements
Reminder: Only One Recitation
Announced on 29 November 2007 4:47 p.m. by Taylor Moulton
Collecting homework if you are not an MBA
Announced on 13 November 2007 12:15 p.m. by Kawika Pierson
Recitation Consolidation
Announced on 02 November 2007 2:55 p.m. by Taylor Moulton
General HW FAQ
We are still seeing causal loop / stock and flow diagrams not following the basic rules. As a reminder, the most common issues are:
(1) missing polarity on arrow head,
(2) not closing important loops,
(3) not labeling loops appropriately with B/R and a name,
(4) causal links cannot influence a stock directly (only its flow), and
(4) units of stocks and flows need to be consistent.
Announced on 08 October 2007 2:17 a.m. by Taylor Moulton
HW#01 FAQ
So far, the class has done well on the first assignment. The TAs are grading diligently and we will let you know when we are finished. In the meantime, here are some common issues we would like to make clear:
B2.a. When members of the population are infected at the start of the simulation, we get "S" shaped growth (not just exponential).
B2.b. There is no mechanism to remove people from the infected stock of the simple SI model, so we will always infect our total population for allowable values of contact frequency.
B2.d. It is crucial to see that a false interpretation of the system might be concluded if too narrow a time horizon (such as 50 days) were used. One might conclude there is only exponential growth and not realize there is an effective carrying capacity.
C2.c. The single most important element missing from our system dynamics model is the behavior modification that reduces infectivity (mask use, etc.) and reduces contact frequency (quarantine, etc.) during the epidemic. Also note that SARSDATA shows "reported" cases and many early infections went undiagnosed for a short period.
C3. It is unreasonable to imagine a policy of "quarantine" in the case of HIV control. We were expecting comments on education and safe sex.
Announced on 21 September 2007 12:03 a.m. by Taylor Moulton