Like many people, I dislike meetings. They are often a waste of time, although they sometimes generate
fodder for an interesting blog post.
But
Paul Graham's blog post on the cost of meetings enabled me to quantify some of my objections.
Graham calls programmers (and writers) "makers" because they are the ones who create the product at his company. He points out that a maker's schedule cannot be easily divided into hourly blocks. A programmer, he says, "prefers to use time in units of half a day at least. You can't write or program in units of an hour. That's barely enough time to get started."
He encourages companies to respect the maker's schedule:
For someone on the maker's schedule, having a meeting is like throwing an exception. It doesn't merely cause you to switch from one task to another; it changes the mode in which you work.
I think that the situation is even worse for statistical programmers. Like Graham, I need several uninterrupted hours in order to develop, debug, and test a complicated statistical analysis. For a particularly tough analysis, I close my door and shut down my email. A block of uninterrupted hours can make the difference between completing an analysis and leaving it for another day. Thus there is a real financial cost to having a programmer attend a meeting.
So close your door, shut down your email, stop reading this blog, and atack that project. And the next time you're invited to a meeting, ask if it can be rescheduled for 4:30 on Friday afternoon. That should keep it short.