TITLE: Agile Policy Development AUTHOR: Eugene Wallingford DATE: June 09, 2007 10:28 PM DESC: ----- BODY: I felt my agile tendencies come through yesterday. First of all, let me say that by and large I am a rule follower. If a rule exists, I like for it to be applied, and applied consistently. Because I tend to follow rules without much question, even when the rule is inconvenient, it often disturbs me to learn that someone else has gotten out of the rule, either because they asked for an exception or because the relevant authority chose not to enforce the rule when it was skirted. This tendency is almost certainly a direct result of my upbringing. That said, I also know that some rules simply get in the way of getting things done. Now for a story. Several years ago, we had a particular student in our undergraduate program. His work habits weren't very good, and his grades showed that. Before finishing his coursework, he left school and got a job. Fast forward a few years. Perhaps the student has grown up, but whatever the reason he is ready to finish his degree. He comes back to school part time and does well. He completes the coursework for his degree but comes up a bit short of the graduation requirements, due to the grades he earned during his first stint as a student. (The rule in question is rather picayune, but, hey ,it was written by faculty.) The students asks if the department will consider waiving the rule in question. His request doesn't seem like some students' requests, which are often demands masquerading as questions, or which presume that the department really should say yes. This request seems sincere and not to presume that the department owes him anything. If we say no, he will re-take another course in the fall and try to satisfy the requirement. However, a waiver would enable him to move on professionally with a sense of closure. Now, I am a rule follower, but... A colleague expressed well how I felt about this case: There is reason that we have this rule, but this case isn't that reason. Part of my job is to hear and decide on such requests, but I prefer not to make such decisions without input from the whole faculty. So I make it a practice to poll faculty whenever new requests come in. The purpose isn't to take a vote that decides for me but rather to get a sense of what the group thinks. I get to hear pros and cons, maybe learn to think about the request in a way I hadn't considered before. I still have to make the decision, but I do so in a state of greater knowledge -- and a state of transparency. I'm also willing to bow to a consensus that differs from my instinct, unless there is a really good reason not to. Faculty responses rolled in, without a consensus. More agreed with the idea of granting the request than disagreed, and a few offered suggestions for resolving the issue in another way. One person suggested that we should not waive the requirement without having a much more detailed policy in place. The more detailed policy would address the many dimensions of the case. His proposal was quite complete and well thought out. But it seemed like overkill. This is a one-off case. We haven't had a case like this in my memory, and I don't expect that we'll have many like it in the future. Perhaps students don't ask for this kind of waiver because they don't expect it to be granted, but I think it's simpler than that: there aren't many students in this situation. It seems unnecessary and perhaps even detrimental for us to specify rules that govern all possible combinations of features when we don't know the specifics of future cases yet -- and may never face them at all. In the realm of the policy, this feels like a prototypical case of YAGNI. If we see a second case (soon), we will at least have reason to believe that the effort designing a detailed exception policy will be worth it for both faculty and students. There is one other fact that makes this increasingly unlikely as time passes: the particular set of requirements under discussion is no longer in effect, and applies only to students who began their CS majors under an older catalog that has been superseded. In any case, I'd like to give us the opportunity to learn from the next request before we try to get a detailed policy just right. I do like to create policy where it is useful, such as for recurring decisions and for decisions that involve well-understood criteria. An example of this sort of policy is a set of guidelines for awarding a scholarship each year. I also think policy helps when it eliminate ambiguity that makes peoples' lives harder. An example here is a set of guidelines for faculty applying for and being awarded a course release for scholarly work. Without such a policy, the process looks like a free-for-all and is prone to unfairness, or appearance of the same; the result would be an inefficient market of projects that would hurt the faculty's collective work. Otherwise, I think that policy works best when it reflects practice, rather than prescribing practice. This reminds me of a discussion that America's founding fathers had in creating the U.S. Constitution, which I mentioned in an earlier entry on Ken Alder's The Measure of All Things. Ultimately, I trust my judgment and the judgment of my colleagues. In situations where I don't think we know enough to define good rules, I'd rather encourage a conversation that helps us reach a reasonable decision in light of current facts. When the conversation turns from giving value to wasting time, then a policy is in order. To me that is better at forecasting a policy for experiences we've not had and then facing the prospect of tinkering at its edges to get it right later. That creates just the sort of uncertainty, both in the folks applying the policy and in the folks to whom the policy applies, that a good policy should eliminate. Trusting judgment means trusting people. I'm comfortable with that, even as a law-and-order kind of guy. -----