Thank you everyone who handed in a course evaluation in the first week of October. This document tabulates the results. The evaluation asked for comments on
| challenging | 13 |
| learning the material | 10 |
| interesting examples and applications | 6 |
| theoretical and design parts of course | 3 |
| Deitel + Carrano | 2 |
| C++ classes | 2 |
| learning about compile-link process | 1 |
| throw and catch | 1 |
| lectures | 23 |
| quizzes helpful | 15 |
| interaction and explaining code in lecture | 6 |
| pace good | 5 |
| small lecture sizes | 1 |
| readings written on board at start of lecture | 1 |
| lectures helpful for homework | 1 |
| learning in labs each week | 15 |
| assignments in pieces | 7 |
| homework | 3 |
| Homework #4 | 5 |
| SUD | 2 |
| UNIX | 2 |
| working together | 1 |
| I like (name omitted) | 13 |
| helpful staff | 12 |
| the web page | 8 |
| quick e-mail help | 5 |
| working at home | 1 |
| dropping lowest grades | 1 |
| Everyone drops what?! | 1 |
| organization | 1 |
| lab consultants | 1 |
| too big a leap from CS 152 | 8 |
| Textbook (Carrano) | 4 |
| programs too small | 2 |
| out of date | 1 |
| not following textbook | 1 |
| linked lists are scary! | 1 |
| homework unrelated to lecture | 9 |
| lecture unorganized | 4 |
| quizzes | 3 |
| lecture too long | 2 |
| debugging ``on-the-fly'' in lecture | 1 |
| too fast paced | 1 |
| lab assignments too long | 9 |
| first half of lab | 4 |
| basing homeworks or labs on (unfinished!) homework | 3 |
| makefiles unexplained | 3 |
| labs unclear | 3 |
| UNIX | 2 |
| not clear what homeworks require | 19 |
| lots of work and little time! | 13 |
| too hard | 8 |
| Homeworks #1-4 required too many things at once | 4 |
| makefiles | 4 |
| bugs in ``solutions'' (or IO) | 3 |
| homeworks complicated | 3 |
| Homework #1,2 | 2 |
| frustration | 1 |
| staggering of sections | 5 |
| homework lags behind lecture | 4 |
| lab TA unsure about what is covered in lecture | 2 |
| coordination between lectures confuses students | 1 |
| homework is unfair | 1 |
| not enough help or office hours | 8 |
| I don't like (name omitted) | 7 |
| slow grading | 2 |
| Explain or help with homework (more) in lab | 8 |
| Explain homework in lecture | 8 |
| Continue to not be like previous semesters | 4 |
| Course booklet (like CS 152) | 3 |
| Make lab directly related to lecture | 3 |
| Provide solution executables | 2 |
| Let us work with whole programs | 2 |
| Open discussion forum on web | 2 |
| More review for quizzes and midterms | 2 |
| Single class-wide due-date | 2 |
| More hands-on time in lab | 2 |
| Go over quizzes afterwards | 1 |
| Use Borland, Watcom or MS | 1 |
| Tutoring help | 1 |
| More explanation of Makefiles | 1 |
| Explanation of distinction between different files: ``.cc'', ``.h'', ``.o'' files, executables. | 1 |
| Provide big picture (concepts) | 1 |
| Give models for us to write code on in homework | 1 |
| ``Brainwash'' us into starting assignments early | 1 |
| One thing at a time in lecture | 1 |
| Explain form of expected solution (esp. HW 1,2) | 1 |
| Explain gdb | 1 |
| When are we going to cover inheritance ? | 1 |
| Make sure TA's know material | 1 |
| Weight exams heavier | 1 |
| Pretest Homework | 1 |
| Lab manual | 1 |
The biggest difficulty seems to be the homeworks: making sure they are unambiguous in what is required, yet giving (potentially ambiguous) guidance on how to proceed. I assume that some of the comments here come from the way in which homework was graded. We will not be able to get rid of all ambiguity here since the grading must be subjective. But I believe we have a way to go before reaching that point. I, in particular, will make an effort to talk in lecture about practical issues related to the homework.
There are a number of concerns about how the different due dates for homework play out. The Thursday labs have to turn in homework long before the rest of class, and the later labs have a lag from lecture. I'm not sure the alternative (a single due date) is much better, because then people have labs either far before or right before the deadline. Thus the lab cannot be directly related to homework, because then the information would either be too early or too late for some people. There is also the concern that a single deadline (as in CS 152) would cause a massive use of miller right before the due date.
There were many positive responses about the quizzes, unusual for something like this. People said they liked knowing right away whether they understood something and needed to review or not. On the other hand, some people suggested having all quizzes ungraded. For next semester, I am considering having pass/fail grades together with makeups for those who failed (or missed taking the quiz).