Saturday, April 12, 2008

Numb and number

Jobs & number games

Spring is the traditional hiring season for teachers and professors. Eager job-seekers file their applications, a lucky few are called in for interviews, and an even luckier fewer are called back for a second round. The sad condition of the state budget (and the governor's suggested cuts in education spending) puts a damper on the process this year, but our schools still set their hiring priorities and we work our way down the list as funds become available.

At my school, as at most colleges, the business of hiring faculty members is an intense multi-level survival-of-the-fittest contest. Each department carefully monitors the retirement plans of its members and crunches its enrollment numbers, ready at any opportunity to justify how many replacement or expansion slots it needs to fill. All sorts of factors come into play, although some of the most important cannot be voiced aloud (like the crying need to hire a tie-breaking third faculty member for the two-person department whose two professors hate each other).

Numbers play a huge role: enrollment, productivity, full-time equivalents (FTE), weekly student contact hours (WSCH), etc. It's never clear whether the less numerate departments are actually at a disadvantage in their quest for new faculty hires; their inability to crunch numbers sometimes produces spurious but impressive data that gives them the appearance of outpacing other academic areas. There is, naturally, constant second-guessing and auditing of each other's numbers. I suppose that's the main reason that departmental hiring request forms now require as much cross-checked information as your basic 1040 form.

Some of the numbers, however, are subjective in principle. These are the ranking scores from faculty members and school administrators who set priorities for the college. (It doesn't matter if your electronics department has the best productivity numbers if the campus as a whole elects to become a fashion and design school.) The ranking numbers have, on occasion, been used and abused to distort the process. One abuse was sheer innumeracy, which wiser heads (belonging to math faculty members) caught in time to avoid scandal. A different abuse was of the more numerate kind, and was perpetrated by a math professor with an impaired conscience.

Hurray for zero

The innumerate mistake was obvious once it was pointed out, but it came as quite a surprise to many participants in the ranking process. Each year the college's faculty senate meets to pass judgment on a stack of departmental petitions for faculty positions. The senators pore over the petitions, listen to brief presentations from each petitioning department, and then cast ballots that rank the petitions in priority order. The ranking numbers for each petition are added up and these totals are used to create an aggregate ranking by the faculty senate of all of the hiring requests. As in golf, a low score is better than a high score, since low scores indicate a lot of senators ranked you as 1, 2, or 3 rather than as 18, 19, or 20.

Several years ago, a few innumerate senators (most of whom didn't bother to attend the faculty senate at other times of the year) cast ballots to favor their own departments' requests (not an unknown phenomenon). Having accomplished their mission of ranking their parochial interests as number 1 (and 2 and 3 if multiple positions were requested), some of them got bored and left the rest of their ballots blank, as evidence they didn't care about the other positions. Just their own. By the logic of the aggregate ranking system, however, they thereby scored all the other positions with zeros ... and zeros outrank ones! In effect, all the positions they left unranked were treated as their top choices—beating out their own departments' positions.

They never made that mistake again, either because they learned better or because their departments chose to elect different senators the next year.

And the last shall be first

A canny member of the math department twisted the ranking system in the opposite direction, using his insight into the process to attempt to hijack it. He was an outspoken technophobe and dreaded the growth of our computer science department. One year the computer science people got ambitious and requested four new faculty positions for the next school year. While most of us chuckled at computer science's audacity (and agreed that the department's numbers justified two hires, but probably not three and certainly not four), my technophobic colleague was outraged. I think he took it as a personal affront.

As the senators filled out their ballots, most of us were ranking Computer Science #1 right near the top of the priority list and Computer Science #2 somewhere in the middle. Computer Science #3 and #4 were often relegated to ranks in the twenties (which was at or near the bottom because there were only 25 requested positions to rank in priority order). My sly colleague, however, voted to give Computer Science #1 last-place priority, writing a 25 on its ballot line. He similarly voted to give Computer Science #2 the next-to-last place by ranking it 24th. By giving the worst possible vote to computer science's first hiring request, my colleague had dramatically increased the score it would get when its rankings were totaled. He thus single-handedly pushed Computer Science #1 down several positions in our aggregate ranking, since most of us had given it 1's and 2's, so his 25 was like an elephant in the room.

Alas. His cleverness was for naught. That year our budget was good enough to cover most of the hiring requests and we worked our way further down the list than usual. Computer science got their two faculty positions. Another factor frustrating my colleague's machinations involved the actual disposition of the priority list generated by the faculty senate. It was submitted to the president's office, which then compared the faculty's priorities with a separate ranking created by the college's deans and vice presidents. A compromise priority list was then created by melding the faculty and management rankings, producing a list that looked the same as the management list.

They had already decided we were going to hire two computer science professors.


My clever colleague's ploy was banned the following year, when it was agreed by the faculty senators that no one could rearrange the rankings of multiple requests from a single department. Shortly after that the math department voted him off the senate.


Anonymous said...

Unfortunately in a voting or ranking system of this type, it's always possible for someone to skew the vote unfairly. Finding the fairest ranking system is a highly non-trivial problem!

The Ridger, FCD said...

Throw out the high and low score? That might work if just one guy is doing it.

Anonymous said...

"They had already decided we were going to hire two computer science professors."
Yeah, this I can relate to. Sic transit gloria mundi.

Anonymous said...

Spring is the traditional hiring season for teachers and professors.


Seriously, once I catch my breath you've got to tell me these schools that let you wait all the way until spring before applying for professorships.

Zeno said...

How about every community college in California? That's over a hundred schools right there. Several schools I know are hiring faculty for next fall and they're doing it right now. (And you might notice it's spring.) I know less about high schools, but one of my friends is a finalist for a position with one in southern California. A lot of the application deadlines are in February and March, so the candidates have to assemble their packets during the winter, but that doesn't contradict my simple statement about spring being the traditional hiring season for teachers and professors.

The universities and four-year colleges have more complex processes that may involve multiple campus visits and guest seminars or lectures, etc., so I'm sure they operate on a much longer timeline than we do, but I wasn't talking about them. I was speaking from my perspective as a community college faculty member.

Sorry if that was unclear to you.

Anonymous said...

California likes to leaves things until the last minute, evidently. The tenure-track process at some of the universities was accepting applications through late January. Meanwhile I know that community colleges in Maryland start looking around that time as well.

Sorry if I seemed a bit harsh.. I'm a bit edgy on the application cycle for the third year in a row.

Zeno said...

Yep, that's good old laid-back California for you. No harm done, although I admit I was a bit piqued that you thought I didn't know what I was talking about.

As for your third go-round in the application process, good luck! (I rode the merry-go-round for a few years before grabbing the brass ring. It's stressful, all right.)

Zeno said...

Throw out the high and low score?

We eventually did even better than that, Ridger. A senator from the math department eventually persuaded the faculty senate to rank the requests by median score rather than aggregate total. Even shenanigans like my manipulative colleague's would have had minimal impact under such a priority system.

Anonymous said...

So you threw out all but one score, but did it from the top and the bottom equally. I suppose I never thought of the median as the result of repeating "throw away the high and low score"...