LACS Notes
6/23-24/2000
Taken by Kim Bruce
Attendees
included:
Bob Cupper (Allegheny), Allen Tucker (Bowdoin College), Joel Adams (Calvin College), Scot Drysdale (Dartmouth College), Dave Reed (Dickinson College), Henry Walker (Grinnell College), Max Hailperin (Gustavus Adolphus), Mike Jipping (Hope College), Mike Schneider (Macalester College), Amy Briggs (Middlebury College), Rhy Price-Jones (Oberlin College), Andrea Lawrence (Spelman College), Doug Baldwin (SUNY Geneseo), Charles F. Kelemen (Swarthmore College), Tom Whaley (Washington and Lee), Takis Metaxas (Wellesley College), Bob Noonan (William and Mary), Kim Bruce (Williams)
Emeriti:
Ted Sjoerdsma, Jeff Parker
Observers:
Claire Bates Congdon (Colby College), Daniel Scharstein (Middlebury College)
The turnout was outstanding. Most of the emeriti members who could not attend sent e-mail expressing their disappointment at not being able to come and wishing the group well. Unfortunately Doug Baldwin had to leave before the meeting formally began because of a medical emergency at home.
General themes included the difficulty of hiring, increasing enrollments, problems obtaining support staff, and retention of women. Discussion of systems support resulted in suggestions to get Deans to compare CS support staff with Biology, Chemistry, etc., with a justification of security rather than the usual safety concerns in the other sciences. It was also suggested that salaries for CS support staff be comparable with IT at the college (which has likely increased salaries dramatically recently).
Max mentioned that there is a program called "Preparing Future Faculty" at U. of Minnesota which attaches mentors to grad students. Mentors can be faculty at liberal arts colleges, who can try to help the grad students look at liberal arts colleges.
Henry Walker gave a report on forthcoming changes in the AP CS exam. An ad hoc advisory committee was set up about a year ago (headed by Henry) and attempted to get feedback from institutions (though had difficulty). The main feedback was that object-orientation needed to be taught somewhere in the first year. The committee has now recommended that the exam switch to Java-based from C++-based. The development committee has endorsed that decision, and the College Board is now considering that change. They will likely develop a detailed rationale for the Board. The new course will likely emphasize high-level thinking skills rather than syntax.
The earliest this change could take place is 2003-2004. It is interesting to note that there were over 20,000 students taking AP CS exams this year, compared with an average of about 11,000 when it was offered in Pascal.
Mike Schneider presented a discussion on the current status of Curricula 2001. The core curriculum is about to go into Stoneman specification, indicating that there likely won't be many changes after this.
A new decision made in June is to no longer attempt to construct a core curriculum applicable to all computing disciplines. Instead Volume I will be introductory, Volume II will cover CS, Volume III Computer Engineering, Volume IV Software Engineering, etc. Thus the current core specification (which Mike showed us) is only applicable to CS (although they hope the CE will have a lot of overlap, particularly in the first few years).
The meat of the report will be written by the Pedagogical Focus Groups (PFG's), which are described below:
PFG 1:
1.
Programming first
A. Algos first - algos, Data structures, pseudo-code
B. Specific language first: Imperative/OOP/functional
2.
Programming late model
A. Bottom up model: hardware,
circuits,
organziation, virtual machines
B. Breadth-first survey of CS
3.
Other approaches
A. Net-centric intro to CS
B. Integrate
discrete math /
programming model (math on demand model)
PFG 2: (Henry) Supporting Courses
1. Mathematics: 2 course discrete math model (1.5 units on 0.5 units on probability/stats)
DS1, DS2, + 1 - 2 additional math courses chosen to suit the needs
2. Science: one semester of lab science
3. Ethical/Social Issues: Considerable exposure
4. Communication Skills: one semester course outside CS that emphasizes writing skills
5. Foreign language: 3 semesters of language other than English or 2 semesters of language + 1 courses with a significant cultural or international perspective
6. Other recommendations: Study of an applications domain which is a user of computing. Much more on communications skills -- 4 semesters of courses w/significant writing component and oral discussion.
PFG 3: Core group
Group did nothing!
Eric and Michael did proof of concept. As did McGettrich.
Eric's was densely packed
Michael's was 8 courses - added extra material
Now there is a new chair. Perhaps they'll get something done.
PFG 4: Professional Practices
Defined terms, charter, scope, goals
Outlined main areas of concern:
· Business - industrial concerns
· Educational - accreditation - continuing ed, communication skills, professional skills such as teamwork, risk assessment, professionally oriented programs such as internships and co-ops.
· professional - licensing / certification; professional organizations; standards; tool usage; code of ethics and conduct.
· Governmental - computer law, ownership of intellectual property, computer crime
Developing educational materials for each area.
complete
course
models for practicums, internships,
coops
modules to be used in other courses
PFG 5: Advanced Elective Courses
virtually nothing done - very confused
Talked some about vertical vs. horizontal model
Talked some about "extracting common themes"
Probably will not do more than listing some fundamental advanced electives
Trying to get them to address models for undergraduate research / independents / honors / capstones
PFG 6: Computing Across Curricula
Charge: plan a process to figure out how to solve the problem
Steering committee very confused as to what they are supposed to do - just saying should be done later by someone else. Now going to be more specific.
Heavily influenced by NRC report "Fluency with Information Technology"
Goals and Rationale - impact of IT on all other disciplines. Need to function as citizen in today's high tech society. Skills vs. concepts
Curriculum development process: Communications/interactions with client departments. course design. course evaluation and assessment
Types/modes of courses: General fluency. Multidisciplinary courses for related group of departments. Specialized or single-disciplinary service courses.
Bob:
80% of students now have their own computers.
Choose software so students can use from their rooms.
Terminal emulation helps.
JDK can run on any.
UNIX emulations possible for some platforms (Cygnus)
[Charles tries to encourage students to work in lab instead so can get legal help. Helps build esprit de corps. Others agreed.]
Mike
Students showing off with hot machines cause problems - intimidate other students rather than share.
Some of these students expect similar environments in college lab.
Kelemen talked about the GEN proposal to Swarthmore, Williams, etc.
Bruce talked about the Williams committee deliberating on distance learning and intellectual property issues.
Takis also discussed the GEN presentation at Wellesley
Parker discussed his experience teaching distance learning at Harvard Extension.
Streaming video of the instructor seems to be important to students taking these courses.
Chat rooms don't seem to work well, e-mail better.
Lecture adds to notes and text.
We came up with the following recommendations on changes to the CS2001 proposal
CUT
IS2,3 cut from 8 to 4 hours (-4)
IM3, cut 6 to 2-3 (-3)
NC, cut from 15 to 10 (-5)
SE 1-5, cut 35 to 18 hours (-17)
SP, cut from 16 to 10 hrs (-6)
HC1, cut 6 to 3 (-3)
GR1, cut (-5) add 3 hours of principles of graphics
Total cut -40
ADD
Prog paradigm (alternative) +9
Models of computing AL5-7, +6
AL3 (+3)
PL8 in place of PL2: net change (+4)
Add 22 hours
Total change: -18 hours
Rationale
We like the categories.
Change balance between principles and techniques
Recognize value of reducing the core
Recognize that SE will have its own core which will add to this minimal core
Reduce core by about 1/2 course
Mike and Max will try to write this up. Charles and Scot will look at AL5-7.
Kim discussed a library developed at Williams which supports the use of graphical objects, event-driven programming, and concurrency in CS1. They are very excited about how well it has worked and are going to be developing materials further for others to use and hope to begin writing a text soon. More information is available at http://www.cs.williams.edu/~cs134/CCLI/intro.html
Bob gave an eloquent history of the 17 year history of the liberal arts CS consortium, including questions for the future. He will be sending a copy to everyone.
Briefly, LACS began in summer of 1993, in conversations between Cupper and Norm Gibbs, newly moved to Bowdoin. There were 9 attendees at the first meeting in the summer of 1984 at Bowdoin. In the first several years we generated a model curriculum for liberal arts programs, and engaged in lengthy discussions of the role of labs, both of which resulted in papers. We also served as a general support group for faculty suffering the difficulties of trying to establish computer science at their home campuses.
In 1991, questions arose as to whether there was any reason to continue as many of the initial issues bringing us together had been resolved. We decided to continue, and a rush of other issues have kept us engaged since (including our revised curriculum in response to the Curricula 2001 report).
Issues that seem always to be with us are workload, keeping pace with technology, and recruiting. Bob reminded us that we are a group organized as individuals rather than institutions, and that seems to make us more effective as a working group. Bob also said many other things, but it will be better to read Bob's original remarks.
Allen added his feelings on value of LACS on other topics, e.g., as outside reviewers, intellectual contributions, etc. Henry, Kim, Charles, Ted, Max, and others chimed in with their memories and statement of how valuable LACS has been to them.
Curricula had impact on schools outside of consortium.
Useful clout w/ Deans, etc.
Takis made suggestions for changes to the questionaire, drop the minimum size for CS question. Instead add question on how many students go to grad school, gender composition, etc. Rhys will take charge of making those changes for next year.
We also discussed ways of getting useful salary comparison data as the CRA data does not include our type of school, and a general survey of all non-Ph.D. schools would also not likely be representative of our kind of school. It was mentioned that CUPA has disciplinary salaries.
We discussed as to whether there might be some way to get CLAC to do it. Henry, Scot, and Takis will work on this.
We ended by giving thanks to Allen for his excellent job of hosting the workshop and to Rhys for his organization and running of the workshop.
It was noted that Dave Reed and Rhys Price-Jones are moving to new institutions. As a group of individuals, there was no question that both would retain their membership. At this time it was decided not to expand membership to invite a replacement from Dickinson or to add any other faculty. Instead it was felt that a membership size of 15 to 20 was good. If it drops below 15 then we might want to think about making additions.
The meeting next year will be at Williams if we can get the dates worked out. It now appears that the meeting will most likely be in July or even early August.