Course Evaluation

[Steve W]

Hello All!  

I've just received results, compiled by the ROSE school, of student evaluations from our spring 2013 course in Reliability Analysis.

I don't want to make all of these results available to all, to avoid violating your understanding of confidentiality.  I do, though, want to make a few general comments.

First, many people expressed gratitude for the course, its content and teaching, etc.  For all of these remarks, I am extremely grateful.

There are also a number of constructive comments.  One of the most prevalent was the suggestion that the course content could be useful separated into 2 ROSE courses.  This seems quite sensible to me.  One student suggested to include some of the introductory prob/stat stuff in "the math introductory course".   I don't know this particular course, so won't comment specifically on this.  I would say, though, that a "probability for engineers" course is best taught by someone with practical experience applying this stuff to real engineering problems.  If taught instead by a math or statistics department, I've found there's a chance that the material will be dry or overly academic.

Another common suggestion was for a course project.  This is a request I hear often as I teach, though I've not yet tried to design such a project.  My main concern here is whether the knowledge gained, through such a project, is justified given the "cost overhead" in terms of time spent (1) introducing the problem, (2) the needed software, and (3) the time required for students to create either oral and/or written presentations of the material.  I'm open to suggestions on this.

Finally, a number of people remarked on the course content -- specifically, that it went beyond the normal seismic context (e.g., into wind/wave applications).  Some students suggested greater presentation of seismic applications -- perhaps involving Prof. Bazzurro more often.  I think this is a fine idea.  I would only say, though, that I don't feel too apologetic for giving you access to a broader world of applications.  One simple reason for this is that wind/wave applications are what I know best -- and I think that an instructor is required to give you the most realistic applications he/she can, even if they aren't what the students care most about.  More importantly, Structural Reliability is a course with a special property -- it applies the same methods/algorithms across a wide range of engineering problems.   I hope you've come to appreciate these analogies: input conditions described probabilistically, response describe through random vibrations, and the combination/integration through various math methods (e.g., direct integration, FORM, and simulation).  Maybe I should emphasize these analogies better if I get to teach this the next time.

Again, thank all of you for your comments -- I recognize and appreciate your time and efforts with this.  And best of luck in your future efforts -- Steve

Personal Website: http://sites.google.com/site/stevewinterstein

Company Website: http://sites.google.com/site/probabilitybasedengineering