|
Computer Science
Prepared by John Tarjan,
CSU Bakersfield, Lead Discipline Faculty for Computer Science
Summary of Identified Issues 
Several key issues emerged. It is interesting to not that
an almost identical list of issues came up at all four regional
meetings. They are discussed below.
Topics to be Covered
in the Calculus Series
Faculty were fairly united in the belief that most important
outcome from taking the calculus series would be mental discipline
and mathematical/logical reasoning. Performance in calculus
is one of the best predictors of performance as a Computer
Science major. Almost all four-year faculty agreed that computer
science majors should take the series designed for engineering
majors.
Topics to be Covered
in Discrete Mathematics
There was less congruence in terms of topical coverage in
the discrete math course. Boolean algebra and proofs were
mentioned the most frequently.
Topics to be Covered
in the Physics Series
There was relatively little agreement on topics. The faculty
agreed that computer science majors should take the same physics
taken by science and engineering majors (possible to substitute
chemistry). Four-year faculty agreed that it should be calculus-based.
Two-year faculty were less concerned about a calculus prerequisite.
It was felt that exposure to rigorous scientific thinking
was more important that specific topics, although electricity,
etc. may be more relevant topics for computer science majors.
Skills Required
to do Well in a Computer Science Program
As mentioned above, the mental discipline required to succeed
in a rigorous calculus series and a rigorous science series
is considered a good predictor of success in a computer science
program. Study skills, English fluency, logical reasoning
and abstraction were frequently mentioned as other requirements
for success.
Prerequisites/Course
Sequencing
There was a great deal of variation between the three
systems as to course sequencing and prerequisites. Part of
the problem arises because many of the computer-related coursework
taught at the community colleges is taught to an audience
much broader than computer science majors.
Baseline Curriculum
The "typical" lower-division course pattern for
computer science majors was treated at every meeting. It was
somewhat surprising to the facilitator how much overlap there
was at the lower-division level, despite concerns about rigor
and coverage in those courses. Faculty at each of the regional
meetings independently came up with a very similar "baseline"
lower-division curriculum to be recommended to computer science
majors.
Identified Trends/Future
Directions
The faculty identified several trends and future directions
in computer science.
The Language Problem
There was a good deal of pessimism that the problem of different
programming languages could ever be "solved." It
would be very difficult to get different programs with different
emphases to standardize on a language for Computer Science
1 and Computer Science 2 at one point in time, much less to
get programs to change that language in unison.
Evolving Nature
of Computer Science
It was agreed that since computer science is such a dynamic
field, whatever comes out of these efforts will require that
future monitoring, discussion and revision would constantly
be needed.
CAN
No computer science courses are currently CANned, including
Discrete mathematics. Future groups may wish to address this
issue.
Comments from Statewide Meetings and the General
Field
Curricular Changes
without Adequate Notice to Community Colleges
Many community college faculty indicated that insufficient
articulation was taking place with four-year campuses. Computer
science faculty initiate needed changes but often do not fully
appreciate the amount of lead time necessary to implement
changes at the community college level.
Importance of
Rigor in Lower-Division Coursework
A fairly common concern is that transfer students have not
been exposed to the same level of rigor in their lower-division
coursework that native four-year students have been. Many
faculty felt that the best indicator of success in the computer
science major was performance in a rigorous physics and calculus
sequence. Most felt that these disciplines helped develop
the critical thinking and reasoning skills essential for computer
science students.
Difficulty Community
Colleges have Offering Several Levels of a Similar Course,
Even Though the Objectives of the Students May Vary Widely
Most community colleges are capable of offering rigorous
sections of Computer Science 1, Computer Science 2, etc. However,
most of the students taking computer science courses at this
level are not intending to transfer to four-year computer science
programs. They may be seeking a certificate, an A.S., or to
transfer to a Computer Information System program. Only a
minority ever transfer to a computer science or a computer
engineering program. It is likely economically infeasible
to offer sections just geared to computer science or computer
engineering majors.
Confusion of
CIS, CS, CE Curricula by Community College Students
Many students begin at community colleges with an interest
in computing, with no firm idea of what type of program they
eventually will seek to transfer to, if
indeed they do transfer. This makes the problem of generalist
computer science courses that may be less than optimal for
computer science transfers even more difficult to overcome.
Faculty are put in a very difficult position trying to gear
coursework for students while only being able to offer a limited
number of sections.
Shortage of Instructors
Given that the starting salaries of some students who have
only a certificate are sometimes above those of instructors,
it is becoming increasingly difficult to attract and retain
computer science faculty at the community colleges. The problem
is mirrored at the CSU campuses and to a lesser extent at
UC campuses.
Need for "Remediation/Bridge
Courses" upon Transfer to a UC Campus
Experience has shown at many four-years that even students
who have done relatively well in articulated lower-division
coursework are not sufficiently prepared upon transfer to
be successful in upper-division coursework. Much of this situation
can be attributed to use of different programming languages
at different campuses, but clearly not all of it. Computer
Science 1, Computer Science 2, etc. taught at the community
colleges often do not cover as many topics as those at the
four-years. Several campuses have addressed this problem by
developing "bridge courses" for transfer students.
Most of these students "catch up" fairly well by
taking these courses even though, or perhaps because, many
topics may be repeated.
Use of Different
Programming Languages Across Campuses
Everyone recognized this as a problem. No one had a proposal
to overcome it.
Need for Continuous
Intersegmental Meetings
The meetings were seen as very valuable by faculty from all
segments. Given the issues identified and the dynamic nature
of computer science, it was felt that these meetings should
somehow become institutionalized.
Students should take Calculus for Engineers
and Scientists and Majors' Physics
These sequences are good training for the mind. They should
be rigorous. The mental discipline, critical thinking, etc.
required are essential for success upon transfer.
Different Programming
Language Standards Cause Serious Problems for Transfer Students
This came up again and again. No approaches to this problem
would seem too forthcoming given the diverse nature of programs
and faculty.
CSAB/IBET Accreditation
Standards and the ACM/IEEE Curriculum Guidelines
These standards can serve as the departure point for discussion
of curriculum, skills sets, etc. They can be very helpful
in efforts such as IMPAC.
CS Program Impaction
Many computer science programs are impacted. Not all four-year
or two-year programs have an incentive to increase the number
of majors, making facilitating transfer seem perhaps less
important to faculty.
Computer Science
is Not a Four-year Program
Native student routinely require five years or more. Community
college students with weak math skills and job commitments
may be looking at 7 or 8 years.
Some Students at Community Colleges May Need three to four-years
of Coursework (especially mathematics) Prior to Transfer
Confusion of
Computer-Related Disciplines and Programs
There is confusion among many new students as to their ultimate
goal within the field of computing. Some may desire certification,
some belong in CIS programs, a minority of those without clear
direction will end up as transfer computer science majors.
Lower-Division/Upper-Division
Coursework
There is some variation in the designation of lower/upper-division
coursework. Several courses and topics were treated differently
across the four-year campuses.
Recommendations for the Discipline
- Faculty at four-year institutions should post current
course syllabuses on the web to allow for timely notification
of changes in content/approach/texts.
- Faculty at two-year institutions should work to make sure
the courses recommended to transfer students are appropriate
to prepare them for the appropriate four-year program (CIS,
CS, etc.). Computer Information Systems and Computer Science
programs differ so widely in approach and required coursework
that students not taking the "appropriate" courses
will suffer from unnecessary and/or insufficient preparation.
For example, a computer science transfer student who took
computer information system foundation coursework would
likely be delayed more than a year due to math and science
coursework.
- An intersegmental computer science curriculum group should
be established to keep the dialogue on lower-division curriculum
going after IMPAC runs its course. Curriculum and associated
issues will continue to change and evolve.
- Communication between community colleges and UC campuses
should be increased. There are currently very few formal
or informal lines of communication between faculty. IMPAC
has provided a much needed avenue for discussion. Perhaps
some agency could provide funding for service-area community
colleges and UC campuses to get their faculty together for
transfer and articulation discussions on a periodic basis.
- Future groups should determine which courses should apply
for CAN status and should review and recommend any existing
descriptors as well.
Recommendations
for Support Courses
- The one-year physics series should be calculus-based and
have a laboratory (the same applies to chemistry, if selected
as an option). Computer science majors should take the same
sequence as science and engineering majors. This point was
stressed by almost all four-year computer science faculty
and most community college computer science faculty. The
physics faculty were all in agreement.
- Computer science majors should take the same calculus
series as science and engineering majors. There was a fairly
strong consensus on this point across computer science and
math faculty.
- The discrete mathematics course should contain: functions,
relations, and sets; basic logic (including Boolean algebra
and 1st order predicate calculus); proof techniques (including
proof by construction, proof by induction and proof by contradiction);
the basics of counting; graphs and trees; and discrete probability.
The math faculty were very solicitous in asking for topics
and approaches that would be desirable. It is unclear how
much of the input from the computer science faculty will
actually be implemented in new/revised sections of discrete
math courses taught through math departments.
Topics for Further
Discussion
We recommend that next year's group take up the following
tasks:
1. Circulate the recommended baseline curriculum as broadly
as possible and solicit feedback.
2. Come to closure on the baseline curriculum.
3. Work with ICAS to form an Intersegmental computer science
Curriculum Advisory Committee to keep this process going and
any recommendations up-to-date.
4. Explore the possibility of developing CAN descriptors
for Computer Science 1, Computer Science 2, Machine Architecture/Assembly
Language, and Discrete Mathematics (or whatever the final
configuration of baseline transfer courses looks like).
5. Keep the dialogue open between systems and try to formalize
contacts on a regional basis.
Recommendations
Forwarded/to be Forwarded to:
CAN: Work with the discipline
to explore possible CAN descriptors for Computer Science 1,
Computer Science 2, Machine Architecture/Assembly Language,
and Discrete Mathematics.
ASSIST: Develop a report of
the current state of articulation in computer science across
four-year institutions and distribute it to next year's group.
CSU CSIS Chairs Council: Review
the baseline curriculum and get comments to the lead discipline
faculty member as a means to keep the council informed and
ensure system-wide feedback.
Outreach presentations
made by members of this group:
| Organization |
Date/Place |
Presenter's Name |
Number Present |
|
| CSU Business Assessment |
4/20/01 |
Tarjan |
25 |
|
| Meeting, Cal Poly, Pomona |
|
|
|
|
The purpose/progress of IMPAC was reviewed. These faculty/administration
assessment leaders were informed of the CIS and business clusters
beginning next year and were asked to encourage participation
from their respective faculties.
|
