What is the half-life of a tweet?

My tweets apparently have a half-life of about two hours, but I have no idea if that's unique to me.  My spouse is new to Twitter and as I was showing him how he could see some data about his tweets, I noticed that the graph of the data looked familiar.  Probably because I taught chemical kinetics twice last year (in pchem and general chemistry).

Over lunch today, while waiting for my car to be serviced, I decided to explore the kinetics of my tweets.  I used data from the first 10 hours after I posted a tweet, and used tweets that had several hundred total impressions and few retweets.  Using five data sets from the past month, I fit the tweets to linear models for 0th, 1st and 2nd order kinetics.  R² values suggest that a 1st order model is most appropriate, with a rate constant of 0.35/hour, which translates to a half-life of 2.0 ± 0.4 hours.  I'm curious if that's relatively constant for me, or whether it's characteristic of other parameters, but time is up.

---

Perhaps because I'm writing this outside in a park, I'm reminded of an infamous problem about the temperature dependence of the chirp rate of male snowy tree crickets in many general and physical chemistry texts.  A discussion of the phenomenon (first recorded in the late 19th century, and not true of cricket everywhere) can be found in Thomas Walker and Nancy Collins. “New World Thermometer Crickets: The Oecanthus Rileyi Species Group and a New Species from North America.” Journal of Orthoptera Research 19 (2010): 371–376.