untitled

Report on ABET Activities
Computer Science Program, University of Virginia

Author: Tom Horton (horton@virginia.edu)
Date: April 26, 2002

This progress report describes activities in the last year regarding achieving ABET/CAC accreditation for the computer science degree. It summarizes the current state of defining our objectives and outcomes, and our definition of a process for continually monitoring program quality and making changes to improve quality. The report also summarizes inputs gathered from stake-holders in the last year, and how we have or will respond to them. The report ends with a description of some of the concerns we have identified when we compare our program to the Criteria of ABET/CAC guidelines.

CS ABET Web Site

We have created a Web site to store all documents and information about our ABET activities. The site is currently at:
http://www.cs.virginia.edu/~horton/abet/

We will attempt to keep this site a "living" repository of all information created and developed for ABET. CS faculty and others can access information here to get latest copies of reports, objectives, time-lines, current concerns, etc. (Some documents may be password-protected.)

Objectives and Outcomes

There has been a lot of discussion in the CS Undergraduate Committee this year regarding the definition of program objectives and outcomes. This has not yet been resolved in a satisfactory manner, and this is an area of concern.

In 2000-2001, the department officially adopted the ABET/EAC required outcomes "A-K" as the outcomes for our CS program. We now believe this it not satisfactory, and we have agreed to adopt a new set of objectives and outcomes that includes computer-science specific statements. We also recognize the importance of making our objectives and outcomes consistent with those of the Computer Engineering program (as much as possible). Many courses we teach are central for both degree programs, and one can imagine that we could double our workload in collecting information etc. where program objectives and outcomes drive an activity.

A draft of a proposed revision based on the current CpE objectives and outcomes is here:
http://www.cs.virginia.edu/~horton/abet/CS-objectives-2-10-02.pdf
This has not been accepted by the CS department but is our committee's current working draft. We think that the computer science specific items in Section 2 need to be revised. Our committee also wonders if the list could be simplified in some manner while keeping it effective for ABET purposes.

Discussion of program objectives and outcomes has been slow and frustrating. Our committee finds this a difficult task, and questions how useful defining these objectives and outcomes at this level are in terms of improving quality of our program. The committee has asked several times "how is this directly useful to us?" The discussion has also at times shifted to whether we should describe our current program or what we wish our program to be (which leads into a discussion of future curriculum revision).

Because of this situation, we cannot provide a mapping of courses to program outcomes that was request. It is not clear that we can require individual faculty to map their course objectives to program objectives in their end-of-year course reports that SEAS will require this May.

In summary, there is a problem here that would be of less concern if an accreditation visit was not so imminent. Arguing and discussing goals is not a bad thing, but in order to carry out the assessment activities (and "close the loop") between now and our visit requires us to resolve this soon.

Stakeholder Input and Department Reactions

We have had several meetings with stakeholders in the last 15 months. Major results are reported in this section. We have not yet analyzed course-specific feedback or the longitudinal survey information supplied to us by SEAS.

Meeting with Current Undergrads, Fall 2001: The chair, associate chair, and I attended a meeting organized by the student computing group, the ACM, last fall. About 25 CS and CpE majors attended. We showed them the ABET/EAC standard outcomes "A-K", and asked for their reaction. We then asked for general feedback about our program. Here is a summary of the major points:
- Students reported that they did not feel they work on multi-disciplinary teams (outcome D).
- Some students reported their TCC courses did not do much with outcome F, professional and ethical responsibility. This seemed to vary among students.
- Students felt that CS101 in its current form did not meet the needs of students who had programmed before coming to UVa. They questioned why we had one course that attempted to meet the needs of students at all levels of experience.
- Other comments were made about specific courses etc.
Department reaction: We are re-examining CS101 because of this and many other inputs. We have identified the issue regarding professional and ethical issues after our own assessment. (More on this in the next section.) If SEAS-wide courses (ENGR162 and TCC) do not support multi-discplinary team activities, then at this time we are not sure how to handle this.
(Note: CS programs are accredited under ABET/CAC criteria, and outcomes A-K are part of ABET/EAC criteria. This is would not be an issue for CS except that we adopted A-K as our program outcomes. It would be an issue for the CpE majors that CS and ECE share.)
Meetings with Corporate Partners Group: In our February 2001 meeting with our industry advisors and partners, we presented the ABET outcomes A-K to the group, asked for feedback, and asked for more general comments about our program and our graduates. Here is a summary of the major points:
- They requested a course in Internet applications, which might include topics like distributed computing (Javabeans, XML, SOAP, CORBA, etc.) and would require the integration of multiple technologies.
- The previous was a more specific version of a general need for a fourth-year level course in which students would work on large systems, incorporating existing components and writing new code, have more of computer systems view than a software development view, etc.
- Include more design, quality assurance, business issues, ethics and intellectual property
- Many wanted to make sure students had significant experiences writing and speaking effectively (writing good memos, emails, giving good techical presentations).
- They asked about the distinction between the CS and CpE degrees. They wondered if we could supply tracks where students specialized within the major.
Department reaction: The need four a fourth-year course that focuses on large systems has been discussed by faculty in the department, though no concrete plans for a new course or courses have been reached. There has also been some discussion of tracks (at least a software engineering option, which would include more focus on quality assurance, design, testing, etc.) As noted above, we have recognized the need for more professional and ethical topics. Plans to meet this need are presented below.

We also had a similar meeting with our corporate partners in Fall 2002 that focused on research. (One member did ask if we had reacted to the suggestions they made during the spring meeting.)

Current Concerns Related to ABET/CAC Criteria

As noted earlier, CS programs are accredited according to ABET/CAC criteria, which differ from the ABET/EAC criteria that all engineering disciplines must meet. The ABET/CAC criteria are similar to the old CSAC/CSAC criteria before that organization merged with ABET. In particular, they include more low-level criteria that must be met, including some prescriptive requirements. These are divided into seven Categories, and under each of these are a set of Standards that represent more specific statement of what a program should do.

This spring we have reviewed the ABET/CAC Categories and Standards and made a first pass at determining where our CS program might have problems. In this section we will list the areas of concern. Note that "Objectives and Assessments" is one of the seven Categories, so our department's discussion this spring about this topic is not wasted. But we could fail to be accredited by not meeting any of the other six Categories.

Objectives and Assessments: The ABET/CAC standards in this area are less than those required by ABET/EAC, so if we meet what SEAS requires for all program then we will not have a problem.
Student Support: We should be able to meet the intent for this Category for ABET/CAC by showing that we satisfy each of the Standards listed in the ABET/CAC criteria. The only possible problem is Standard II-2, which relates to class size. We will have to argue that we provide enough undergrad and grad TAs for upper-division classes larger then 30 students and for large lower-division classes.
Faculty: We should be able to meet the intent for this Category for ABET/CAC by showing that we satisfy each of the Standards listed in the ABET/CAC criteria.
Curriculum: There are a number of Standards in this Category that we must address:
1. Standard IV-3 states: The curriculum must include at least 30 semester hours of study in humanities, social sciences, arts and other disciplines that serve to broaden the background of the student.
  The CS program only has 21 hours (TCC101, TCC2xx/3xx, TCC401, TCC402 and 9 credits HSS). There may be some room for interpretation in the Criteria that addresses programs in Engineering environments. I am investigating this, but if this really is a problem, our 9 credits of unrestricted electives may need to be used to fulfill this requirement.
2. Standard IV-8 states: Students must be exposed to a variety of programming languages and systems and must become proficient in at least one higher-level language.
  We have responded to this by requiring Java in CS302, and we plan to make sure that knowledge of UNIX is always covered in CS216, CS340 and CS414. This should not be a problem.
3. Standard IV-11 states: Course work in mathematics must include discrete mathematics, differential and integral calculus, and probability and statistics.
  We currently require APMA 310, Probability. We have discussed changing this to APMA311, but there seems to be opposition to this. We may require APMA312 and drop one of our Technical Electives, or we may try to argue that the other math classes beyond calculus that we require satisfy the intent behind this Standard. (We need feedback from our stakeholders on this issue.)
4. Standard IV-15 state: The oral communications skills of the student must be developed and applied in the program.
  We need to assess how this is meet in TCC courses and in ENGR162. Students give group presentations in CS340, but this topic may need to be addressed. It will be a challenge since our required courses for majors are typically at least 90 students.
5. Standard IV-17 states: There must be sufficient coverage of social and ethical implications of computing to give students an understanding of a broad range of issues in this area.
  Students report that this not happen in every section of TCC401 and TCC402. Also, since the courses are composed of a mixture of majors, they probably do not address computing-specific issues. We have had early discussions about adding a scattering of a few exercises etc. into courses like CS101, CS201, and CS340. We would also then add more of this to our one-credit seminar, CS390.
Laboratories and Computing Facilities: We should be able to meet the intent for this Category for ABET/CAC by showing that we satisfy each of the Standards listed in the ABET/CAC criteria.
Institutional Support and Financial Resources: We should be able to meet the intent for this Category for ABET/CAC by showing that we satisfy each of the Standards listed in the ABET/CAC criteria.
Institutional Facilities: We should be able to meet the intent for this Category for ABET/CAC by showing that we satisfy each of the Standards listed in the ABET/CAC criteria.

Major Things to Be Done in the Near Future

We need to reach consensus on what to do about our statement(s) of outcomes and objectives.
We need to being to document how we meet the ABET/CAC criteria by starting the self-study document, following the template / example provided by ABET.
We need to develop a time-line and work-plan that reflects what needs to be done and when before the ABET visit.
We need to address the issues listed in the previous section.

Report on ABET Activities Computer Science Program, University of Virginia