Proposal for a Modest Staffing Increase in Computer Science

Last updated : Wednesday, October 27, 2004 at 09:29:43  EST

Overview

Computer Science is currently staffed at 1 2/3 FTEs. For 2004-2005 this is composed of Jim Rogers (full-time) and Charlie Peck (2/3 time). During the preceeding four years the staffing was Jim Rogers (full-time), Charlie Peck (1/2 time), and John Howell (1/6 time). John Howell is now serving as the teaching consultant and that is taking the time that he used to devote to CS.

We propose to increase this to 2 FTEs by increasing Charlie Peck's position to full-time and permanently removing the 1/6 FTE dependency on John Howell. This represents a gain of 1/3 FTE for Computer Science and a gain of 1/6 FTE for Physics, for a total addition of 1/2 FTE worth of salary to the operating budget. Since both of these additions are in the form of existing faculty members there are no additional costs for benefits, research equipment, support staff, etc.

A Brief History of Computer Science

During the 1980's Ray Hively and Hal Hanes made a strong case for building a computer science program at Earlham. As part of this they both volunteered to take a two summer intensive course at Clarkson University. This course was designed to bootstrap the teaching of computer science by faculty in related disciplines. For a time after this Ray and Hal covered the bulk of the courses offered in CS with occasional help from adjuncts and other faculty members (e.g. George Silver, Larry Fisher, Peter Suber, Ray Ontko, and Charlie Peck).

By the Fall of 1994 the strain of carrying the CS program was becoming too much for the limited resources allocated to it and the College undertook to review the place of, and support for, CS at Earlham. This review led to Charlie's teaching 3 courses per year (as a regular adjunct) and convening the program. The remaining courses in CS were covered in turn by Hal, and then Tim McLarnan and Hal, and then just Tim, along with a collection of adjuncts, through 2000. Physics has made a steady and important contribution to the support of CS in this period as well. John Howell (or his temporary replacement) has taught Programming and Problem Solving each year since the early 1990's and Lew Riley taught Algorithms and Data Structures, and Robotics during his tenure here.

During 1996 Charlie was reviewed as part of the normal course of events and one outcome of this process was the conversion of his position from regular adjunct to tenure track (1/2 time).

In 1999 the College sought to consolidate the 1 FTE contribution to CS being made by Tim, Hal, and others into a single tenure track position. With some trepidation we launched a search for a computer scientist who would strongly identify with Earlham's values and overall liberal arts emphasis. Jim Rogers found us and in him we have found someone who not only values Earlham's approach to education but who is also very strong in theoretical computer science. Charlie's strengths are in applied computer science and the combination of the two gives Earlham's CS department a unique combination of breadth and depth for a school our size.

Since Jim's arrival Computer Science has worked to regularize our offerings, further develop our applied groups, and bootstrap two research groups (one in theoretical computer science and one in computational science). We currently support four applied groups with the fifth, WebDB, largely in the hands of Computing Services now.

In 2002-2003 one of our 2002 graduates entered a Ph.D. program, the first of our students to move more or less directly into a Doctoral program. In 2003-2004 we placed three students, one 2003 graduate and two 2004 graduates, in Doctoral programs. This year we expect to place at least one current graduate and up to two 2004 graduates in Doctoral programs. Over the last four years we have moved from placing essentially none of our majors in graduate programs to placing better than 20% of our majors (100% of those who apply) in Doctoral programs.

In terms of off-campus activities our research students regularly present papers in regional undergraduate research conferences (Butler, MCURCSM). Both Jim and Charlie teach workshops and classes in their respective areas nationally and internationally. One of our research groups won "Best Poster" at the Society of Industrial and Applied Mathematics (SIAM) Parallel Processing conference in February of this year.

In general we feel as though we are developing computer science in ways that are congruent with the mission of the College.

The Current Problems

Computer science has proven to be a popular course of study for many students, here are the major/minor statistics for the past 14 years:

Going forward we anticipate graduating about 4 majors/minors in 2005, 6 in 2006, and 8 in 2008.

Clearly this is a growth area for the College. We now have the foundation for a program of national merit, but without staffing stability and a modest increase in resources it will be difficult for us to continue developing the program. Without the modest increase it is hard to see how we can to continue at this pace, even without growth in the number of majors.

As long as CS is dependent on other departments to staff some of its courses we will be faced with circumstances in which staffing difficulties in any of the relevant departments will result in CS being under staffed with no option of coverage from existing faculty. These difficulties will always be critical for us, given the relatively large number of courses we have to support in order to offer a genuine major (14 courses across 1 2/3 FTE works out to 8.4 courses/FTE). While the minimal requirements of the major do not require all of these courses, those of our students who are planning to go on to graduate school need to take essentially all of them. Our staffing level leaves us normally unable to offer any course more than once a year, and leaves us offering four courses every other year and three every three years. Consequently, the loss of any course offering is a significant detriment to our majors.

Short-term and Long-term Plans

Currently John Howell is serving as teaching consultant. The funds for his release time have enabled Charlie to teach an additional course this year (bringing him to 2/3 time). We expect to have to cover Jim's sabbatical in 2005-2006, the second of the two years in which John will be the TC. This will leave Charlie, at 2/3 FTE as the only regular CS faculty. While our experience with adjuncts last year has been considerably better than it might have been, the burden on Jim, our only regular CS faculty at 1 FTE, was considerable. We see no way that these duties could realistically be assigned to 2/3 FTE.

In the year following, 2006-2007, our expectation is that Charlie will be eligible for a sabbatical. While 1/2 FTE leaves fewer courses to cover, it also pretty much eliminates any possibility other than covering with adjuncts or not running courses. Our experience last year suggests that there is a high likelihood that we will need to cancel some courses. We did not offer our normally scheduled AI course last year, for instance, because the absence of his usual assistance in mentoring capstone projects forced us to split the Senior Seminar into two sections.

We propose that Charlie be increased an additional 1/3 FTE in the 2005-2006 academic year, bringing him to full-time. This will give us a chance to deal adequately with Jim's sabbatical in that year. Moreover, we would like the increase to full-time to be permanent, allowing us to search for a full-time replacement during his subsequent sabbatical. Together, this would have us looking to fill a two-year visiting position, which should be significantly more attractive to prospective visiting faculty. This will increase CS staffing from 1 2/3 to 2 FTE. Note, though, that as Charlie is already receiving full-time benefits, the added cost should be salary and FICA only.

This will eliminate the recurring difficulties in sharing staffing with Physics and will make CS responsible for all of our major courses. (Math will still be staffing Math Toolkit, a 2 CR course which covers some aspects of discrete mathematics that are peculiar to CS.) This move from being a department that is staffed ad hoc from other departments to one that stands on its own is an expected step in the maturation of the department.

We do not, on the other hand, want to lose John Howell as a regular member of the department. In addition, John is not happy about the prospect of never teaching CS-128 in the future. The way we propose to address both of these concerns is, in essence, to claim some portion of that additional 1/6 FTE by having John teach CS-128 in (more or less) alternate years. There would also be a possibility of either John or a new hire in Physics teaching a CS topics class occasionally. This would potentially free up one course every other Fall or so, allowing CS faculty to participate in things like Ford-Knight projects, interdisciplinary courses or Earlham Seminars. The difference between this model and the one in which we are dependent on Physics to staff CS-128 is that, in this model, the staffing is our normal responsibility, with the occasional coverage by Physics being a bonus.

Fully Staffed Computer Science

A staffing level of 2 FTEs will allow us to move those of our courses which are currently on a three-year rotation to a two-year rotation with room to add one more course to that rotation. Our proposal is to add a course in Computational Science. This is a course naturally growing out of Charlie's dissertation research and would likely be a course that we would have to add if we were hiring him as a new PhD. It is also an area that is emerging as a necessary component in CS education, one that will soon be a conspicuous gap in our curriculum. Moreover, it will be a course that serves a population of students across the Natural Science Division, potentially extending into the Social Sciences as well. In particular computational methods are now central to research in Physics, Chemistry, and Biology. In this way a portion of the added staffing will not serve the CS program alone, but rather a broad range of science students. Charlie is currently developing this course as part of his PhD work.

The second 1/6 FTE will be used either to add a second section of Senior Seminar, to move our applied groups into the regular curriculum or, most likely, a combination of the two. If our current projection for the number of majors we graduate per year is accurate, we will, as a matter of course, not be able to adequately supervise the capstone projects within a single section. Periodically, Charlie has been responsible for supervising half of the projects, without additional compensation. Moving to two sections would not only serve the students better but would recognize the additional load of supporting them.

By moving the applied groups to the regular curriculum, we would serve the students better by having their participation reflected in their academic transcript. We would also be formalizing the important role that we feel these groups play in our curriculum. Finally, we would also be recognizing some of the actual workload involved in maintaining these groups. Under the current arrangements we have doubled the number of applied groups and are splitting the responsibility for them between us without any recognition of the additional workload.

We might be able to cover both of these goals by splitting this additional 1/6 FTE between the two, keeping a single section of Senior Seminar, adding a single section of Applied CS, and team-teaching them both.

This modest addition in FTEs would allow us to manage our current, and to the extent that we can predict them, medium-term enrollments. If our expectations are realized and we grow, over the next five years, to the point of regularly graduating 10 or more majors we will need to revisit these issues again relatively soon. Our hope is that this addition will tide us over until we have fully addressed staffing crises elsewhere in the College.

Summary

We are proposing to increase the teaching FTEs held by CS faculty by 1/2 FTE, bringing it up to 2 FTE. Of that, 1/6 will primarily be increasing Physics staffing by relieving them of the need to staff CS-128 (although our hope is that they will share that 1/6 FTE with us by continuing to staff it in alternate years when possible). Another 1/6 FTE will go directly to addressing the additional load presented by our increasing number of majors. The remaining 1/6 will go to filling a gap in our program and will service not just CS majors but also majors across the breadth of the Sciences.

The budgetary impact of this increase would be relatively minor, given that Charlie is already receiving full benefits. The practical effect of the increase will be to eliminate the ongoing difficulties in sharing staffing with Physics, relieving Physics of the need to support the CS major and allowing CS to emerge as a self-sufficient department closer to par with the rest of the Natural Sciences Division and with the College as a whole.

In closing we would like to recognize other staffing priorities which have been recommended by FAC and CPC through the RATS process, particularly the Geosciences position. We don't presume that our request is more important than these other needs. Rather we feel that these address independent circumstances and they all deserve consideration, and approval, on their own merits.

John Howell, Charlie Peck, and Jim Rogers
October, 2004