Just got from interlibrary loan today.
The theory that would not die by Sharon B. MaGrayne. Someone recommended it at exmo a couple of weeks ago. Bayesian Analysis. Basically an idea for research. If you have incomplete data, and a theory, and cannot yet prove it (enough to convince others

) the gaps can predict where you can find more data to complete the set and prove the theory.
It is well written, and the author has a sense of humor. I was laughing and read it to someone who was sitting next to me. She didn't get it, so I guess it is only funny to people who have a background in statistics.
She says that a lot of people instinctively use it without realizing that it dates back to the 1700's.