Experiment: The open threads are always used a lot when they are first posted, but they are quickly forgotten, because they get hidden as the month progresses.
This month I will post two open threads. One for January 1-14, and another for January 15-31. I predict, at P=.8 that there will be significantly more open thread posts in the second half of the month using this 2-post method.
To test my hypothesis, I will average the number of open thread comment in the second half of the month in the past few open thread posts, and compare them to the amount of open thread comments on the January 15-31st post.
This test WILL be biased, as it is neither blind on my side, nor on the commenters' side (aka people who post will read this and see the test I am doing.). This will be somewhat ameliorated by me NOT posting this experiment note in the second thread, and in two weeks' time it will not be foremost in people's minds.
I predict, at P=.8 that there will be significantly more open thread posts in the second half of the month using this 2-post method.
Why would you test just the second half of the month, versus total posts in a month?
It seems that you would actually want to measure total participation in a month. I see no reason not to measure it directly, rather than indirectly.
If it's worth saying, but not worth its own post (even in Discussion), then it goes here.
If continuing the discussion becomes impractical, that means you win at open threads; a celebratory top-level post on the topic is traditional.
Poster's Note: omg, it felt so weird typing "2012" up there.