And that's got to be part of the reason that procrastination wasn't studied during the heyday of process studies. If you're doing the cognitive protocol-based study, you're in a controlled environment watching the participant go through the whole process of writing. And while you can get useful information that way, you're missing a huge part of the picture of writing. The huge part that allows you to procrastinate for hours, days, weeks, or more. Time.
If you take writing out of a laboratory setting, a lot more goes on. You can't very well follow a freshman around for the whole semester to jot down every time she thinks about her term paper out loud. So while technically procrastination is a process issue, it didn't fit into the model for empirical studies that process theorists were using.
This is bad. What it did was allow our profession to have strong feelings about procrastination as an important issue since it apparently represented a dysfunctional writing process without research to support our beliefs. If we're teaching process, we're not going to just let our students wait until the last minute to start working on their papers--they need to be thinking and prewriting and drafting and revising over the course of time. Not just churning something out the night before the paper is due.
But none of this is based on empirical research. There weren't any studies done in our field to suggest that procrastination led to bad writing--and how do we know it's part of a bad process if we don't know it leads to a bad product?